Merge branch 'trunk' into dict-for-data

* and_condition_impl::combine_equal - Remove UB container modification.

The container is modified while iterating it.
Switched to indexed based iteration to avoid UB.

* Update condition.cpp

Sorry missed this line.

.github/workflows/build-documentation.yml

@@ -0,0 +1,65 @@
+# This starter workflow is for a CMake project running on multiple platforms. There is a different starter workflow if you just want a single platform.
+# See: https://github.com/actions/starter-workflows/blob/main/ci/cmake-single-platform.yml
+name: publish docs
+
+on:
+  push:
+    branches: [ "trunk" ]
+
+permissions:
+  contents: read
+  pages: write
+  id-token: write
+
+concurrency:
+  group: "pages"
+  cancel-in-progress: false
+
+jobs:
+  docs:
+    runs-on: ubuntu-latest
+    steps:
+    - uses: actions/checkout@v1
+
+    - name: Set reusable strings
+      # Turn repeated input strings (such as the build output directory) into step outputs. These step outputs can be used throughout the workflow file.
+      id: strings
+      shell: bash
+      run: |
+        echo "build-output-dir=${{ github.workspace }}/build" >> "$GITHUB_OUTPUT"
+
+    - name: Install dependencies Ubuntu
+      run: sudo apt-get update && sudo apt-get install cmake doxygen
+
+    - uses: actions/setup-python@v4
+      with:
+        python-version: '3.9'
+        cache: 'pip' # caching pip dependencies
+    - run: pip install -r docs/requirements.txt
+
+    - name: Configure CMake
+      run: cmake -S . -B ${{ steps.strings.outputs.build-output-dir }} -DBUILD_DOCUMENTATION=ON -DBUILD_TESTING=OFF
+
+    - name: Run Sphinx
+      run: |
+        cmake --build ${{ steps.strings.outputs.build-output-dir }} --target Sphinx-libcifpp
+        ls -l ${{ steps.strings.outputs.build-output-dir }}
+        ls -l ${{ steps.strings.outputs.build-output-dir }}/docs/sphinx
+
+    - name: Upload artifact
+      uses: actions/upload-pages-artifact@v2
+      with:
+        path: ${{ steps.strings.outputs.build-output-dir }}/docs/sphinx
+
+  deploy:
+    environment:
+      name: github-pages
+      url: ${{ steps.deployment.outputs.page_url }}
+
+    runs-on: ubuntu-latest
+    needs: docs
+
+    steps:
+      - name: Deploy to GitHub Pages
+        id: deployment
+        uses: actions/deploy-pages@v2

.github/workflows/cmake-multi-platform.yml

@@ -0,0 +1,59 @@
+name: multi platform test
+
+on:
+  push:
+    branches: [ "trunk", "develop" ]
+  pull_request:
+    branches: [ "trunk" ]
+
+jobs:
+  build:
+    runs-on: ${{ matrix.os }}
+
+    strategy:
+      fail-fast: false
+
+      matrix:
+        os: [ubuntu-latest, windows-latest, macos-latest]
+        include:
+          - os: windows-latest
+            cpp_compiler: cl
+          - os: ubuntu-latest
+            cpp_compiler: g++
+          - os: macos-latest
+            cpp_compiler: clang++
+
+    steps:
+    - uses: actions/checkout@v3
+
+    - name: Set reusable strings
+      id: strings
+      shell: bash
+      run: echo "build-output-dir=${{ github.workspace }}/build" >> "$GITHUB_OUTPUT"
+
+    - name: Install dependencies Ubuntu
+      if: matrix.os == 'ubuntu-latest'
+      run: sudo apt-get update && sudo apt-get install mrc
+
+    - name: Install dependencies Window
+      if: matrix.os == 'windows-latest'
+      run: ./tools/depends.cmd
+      shell: cmd
+
+    - name: Configure CMake
+      run: >
+        cmake -B ${{ steps.strings.outputs.build-output-dir }}
+        -DCMAKE_CXX_COMPILER=${{ matrix.cpp_compiler }}
+        -DCMAKE_BUILD_TYPE=Release
+        -S ${{ github.workspace }}
+        
+    - name: Build
+      run: cmake --build ${{ steps.strings.outputs.build-output-dir }} --config Release
+      
+    - name: Test
+      working-directory: ${{ steps.strings.outputs.build-output-dir }}
+      run: ctest --build-config Release --output-on-failure
+
+    - name: Install
+      if: matrix.os != 'windows-latest'
+      run: sudo cmake --install ${{ steps.strings.outputs.build-output-dir }} --config Release

.gitignore

@@ -2,9 +2,14 @@ build/
 .vscode/
 .vs/
 tools/update-libcifpp-data
-data/components.cif*
+rsrc/components.cif*
 CMakeSettings.json
 msvc/
 src/revision.hpp
 test/test-create_sugar_?.cif
 Testing/
+include/cif++/exports.hpp
+docs/api
+docs/conf.py
+build_ci/
+data/components.cif

.readthedocs.yaml

@@ -0,0 +1,22 @@
+version: 2
+
+build:
+  os: ubuntu-22.04
+  tools:
+    python: "3.11"
+  apt_packages:
+    - doxygen
+    - cmake
+  jobs:
+    pre_build:
+      - cmake -S . -B build -DBUILD_DOCUMENTATION=ON
+      - cmake --build build --target Doxygen
+
+# Build from the docs/ directory with Sphinx
+sphinx:
+  configuration: docs/conf.py
+
+# Explicitly set the version of Python and its requirements
+python:
+  install:
+    - requirements: docs/requirements.txt

CMakeLists.txt

@@ -11,86 +11,98 @@
 # this list of conditions and the following disclaimer in the documentation
 # and/or other materials provided with the distribution.
 
-# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
-# ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
-# WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
-# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
-# ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
-# (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
-# LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
-# ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
-# (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
-# SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+# THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS"
+# AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE
+# IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+# DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE
+# FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL
+# DAMAGES (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR
+# SERVICES; LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER
+# CAUSED AND ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY,
+# OR TORT (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE
+# OF THIS SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
-cmake_minimum_required(VERSION 3.16)
+cmake_minimum_required(VERSION 3.23)
 
 # set the project name
-project(cifpp VERSION 5.1.1 LANGUAGES CXX)
+project(
+	libcifpp
+	VERSION 8.0.0
+	LANGUAGES CXX)
 
 list(PREPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake")
 
-include(GNUInstallDirs)
+include(FindAtomic)
 include(CheckFunctionExists)
 include(CheckIncludeFiles)
 include(CheckLibraryExists)
 include(CMakePackageConfigHelpers)
 include(CheckCXXSourceCompiles)
-include(Dart)
 include(GenerateExportHeader)
+include(CTest)
+include(FetchContent)
+include(ExternalProject)
 
-set(CXX_EXTENSIONS OFF)
-set(CMAKE_CXX_STANDARD 20)
-set(CMAKE_CXX_STANDARD_REQUIRED ON)
+# FindBoost, take care of it now.
+if(CMAKE_VERSION VERSION_GREATER_EQUAL 3.30)
+	cmake_policy(SET CMP0167 NEW)
+endif()
+
+# When building with ninja-multiconfig, build both debug and release by default
+if(CMAKE_GENERATOR STREQUAL "Ninja Multi-Config")
+	set(CMAKE_CROSS_CONFIGS "Debug;Release")
+	set(CMAKE_DEFAULT_CONFIGS "Debug;Release")
+endif()
 
-# set(CMAKE_CXX_VISIBILITY_PRESET hidden)
-# set(CMAKE_VISIBILITY_INLINES_HIDDEN 1)
 if("${CMAKE_CXX_COMPILER_ID}" STREQUAL "GNU")
-	set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} -Wall -Wextra -Wno-unused-parameter -Wno-missing-field-initializers")
+	set(CMAKE_CXX_FLAGS
+		"${CMAKE_CXX_FLAGS} -Wall -Wextra -Wno-unused-parameter -Wno-missing-field-initializers"
+	)
 elseif(MSVC)
 	set(CMAKE_CXX_FLAGS "${CMAKE_CXX_FLAGS} /W4")
 endif()
 
-# Building shared libraries?
-option(BUILD_SHARED_LIBS "Build a shared library instead of a static one" OFF)
-
-# We do not want to write an export file for all our symbols...
-set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
+# Build documentation?
+option(BUILD_DOCUMENTATION "Build the documentation" OFF)
 
 # Optionally build a version to be installed inside CCP4
-option(BUILD_FOR_CCP4 "Build a version to be installed in CCP4" OFF)
+option(BUILD_FOR_CCP4 "Build a version to be installed in CCP4")
 
-# Lots of code depend on the availability of the components.cif file
-option(CIFPP_DOWNLOAD_CCD "Download the CCD file components.cif during installation" ON)
-
-# An optional cron script can be installed to keep the data files up-to-date
-if(UNIX)
-	option(CIFPP_INSTALL_UPDATE_SCRIPT "Install the script to update CCD and dictionary files" ON)
+# Building shared libraries?
+if(NOT(BUILD_FOR_CCP4 AND WIN32))
+	option(BUILD_SHARED_LIBS "Build a shared library instead of a static one" OFF)
 endif()
 
-# When CCP4 is sourced in the environment, we can recreate the symmetry operations table
-if(EXISTS "$ENV{CCP4}")
-	if(EXISTS "$ENV{CLIBD}/syminfo.lib")
-		option(CIFPP_RECREATE_SYMOP_DATA "Recreate SymOp data table in case it is out of date" ON)
-	else()
-		set(CIFPP_RECREATE_SYMOP_DATA OFF)
-		message(WARNING "Symop data table recreation requested, but file syminfo.lib was not found in $ENV{CLIBD}")
+if(PROJECT_IS_TOP_LEVEL AND NOT BUILD_FOR_CCP4)
+	# Lots of code depend on the availability of the components.cif file
+	option(CIFPP_DOWNLOAD_CCD
+		"Download the CCD file components.cif during installation" ON)
+
+	# An optional cron script can be installed to keep the data files up-to-date
+	if(UNIX AND NOT APPLE)
+		option(CIFPP_INSTALL_UPDATE_SCRIPT
+			"Install the script to update CCD and dictionary files" ON)
 	endif()
 else()
-	set(CIFPP_RECREATE_SYMOP_DATA OFF)
-	message("Not trying to recreate symop_table_data.hpp since CCP4 is not defined")
+	unset(CIFPP_DOWNLOAD_CCD)
+	unset(CIFPP_INSTALL_UPDATE_SCRIPT)
 endif()
 
-# Unit tests
-option(ENABLE_TESTING "Build test exectuables" OFF)
+# When CCP4 is sourced in the environment, we can recreate the symmetry
+# operations table
+if(EXISTS "$ENV{CCP4}/lib/data/syminfo.lib")
+	option(CIFPP_RECREATE_SYMOP_DATA
+		"Recreate SymOp data table in case it is out of date" ON)
+endif()
 
+# CCP4 build
 if(BUILD_FOR_CCP4)
 	if("$ENV{CCP4}" STREQUAL "" OR NOT EXISTS $ENV{CCP4})
 		message(FATAL_ERROR "A CCP4 built was requested but CCP4 was not sourced")
 	else()
-		list(APPEND CMAKE_MODULE_PATH "$ENV{CCP4}")
-		list(APPEND CMAKE_PREFIX_PATH "$ENV{CCP4}")
+		list(PREPEND CMAKE_MODULE_PATH "$ENV{CCP4}")
+		list(PREPEND CMAKE_PREFIX_PATH "$ENV{CCP4}")
 		set(CMAKE_INSTALL_PREFIX "$ENV{CCP4}")
-		set(CMAKE_INSTALL_FULL_DATADIR "${CMAKE_INSTALL_PREFIX}/share/libcifpp")
 
 		if(WIN32)
 			set(BUILD_SHARED_LIBS ON)
@@ -98,28 +110,37 @@ if(BUILD_FOR_CCP4)
 	endif()
 endif()
 
-if(WIN32)
-    if(${CMAKE_SYSTEM_VERSION} GREATER_EQUAL 10) # Windows 10
-        add_definitions(-D _WIN32_WINNT=0x0A00)
-    elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.3) # Windows 8.1
-        add_definitions(-D _WIN32_WINNT=0x0603)
-    elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.2) # Windows 8
-        add_definitions(-D _WIN32_WINNT=0x0602)
-    elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.1) # Windows 7
-        add_definitions(-D _WIN32_WINNT=0x0601)
-    elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.0) # Windows Vista
-        add_definitions(-D _WIN32_WINNT=0x0600)
-    else() # Windows XP (5.1)
-        add_definitions(-D _WIN32_WINNT=0x0501)
-    endif()
+# Now include the GNUInstallDirs module
+include(GNUInstallDirs)
 
+if(WIN32)
+	if(${CMAKE_SYSTEM_VERSION} GREATER_EQUAL 10) # Windows 10
+		add_definitions(-D _WIN32_WINNT=0x0A00)
+	elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.3) # Windows 8.1
+		add_definitions(-D _WIN32_WINNT=0x0603)
+	elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.2) # Windows 8
+		add_definitions(-D _WIN32_WINNT=0x0602)
+	elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.1) # Windows 7
+		add_definitions(-D _WIN32_WINNT=0x0601)
+	elseif(${CMAKE_SYSTEM_VERSION} EQUAL 6.0) # Windows Vista
+		add_definitions(-D _WIN32_WINNT=0x0600)
+	else() # Windows XP (5.1)
+		add_definitions(-D _WIN32_WINNT=0x0501)
+	endif()
+
+	# Man, this is 2024 we're living in...
 	add_definitions(-DNOMINMAX)
+
+	# We do not want to write an export file for all our symbols...
+	set(CMAKE_WINDOWS_EXPORT_ALL_SYMBOLS ON)
 endif()
 
 if(MSVC)
-    # make msvc standards compliant...
-    add_compile_options(/permissive-)
+	# make msvc standards compliant...
+	add_compile_options(/permissive- /bigobj)
+	add_link_options(/NODEFAULTLIB:library)
 
+	# This is dubious...
 	if(BUILD_SHARED_LIBS)
 		set(CMAKE_MSVC_RUNTIME_LIBRARY "MultiThreaded$<$<CONFIG:Debug>:Debug>DLL")
 	else()
@@ -130,36 +151,51 @@ endif()
 # Libraries
 
 # Start by finding out if std:regex is usable. Note that the current
-# implementation in GCC is not acceptable, it crashes on long lines.
-# The implementation in libc++ (clang) and MSVC seem to be OK.
-check_cxx_source_compiles("
+# implementation in GCC is not acceptable, it crashes on long lines. The
+# implementation in libc++ (clang) and MSVC seem to be OK.
+check_cxx_source_compiles(
+	"
 #include <iostream>
 #ifndef __GLIBCXX__
 #error
 #endif
-int main(int argc, char *argv[]) { return 0; }" GXX_LIBSTDCPP)
+int main(int argc, char *argv[]) { return 0; }"
+	GXX_LIBSTDCPP)
 
 if(GXX_LIBSTDCPP)
-	message(STATUS "Testing for known regex bug, since you're using GNU libstdc++")
+	message(
+		STATUS "Testing for known regex bug, since you're using GNU libstdc++")
 
 	try_run(STD_REGEX_RUNNING STD_REGEX_COMPILING
-		${CMAKE_CURRENT_BINARY_DIR}/test ${PROJECT_SOURCE_DIR}/cmake/test-rx.cpp)
+		${CMAKE_CURRENT_BINARY_DIR}/test
+		${CMAKE_CURRENT_SOURCE_DIR}/cmake/test-rx.cpp)
 
 	if(STD_REGEX_RUNNING STREQUAL FAILED_TO_RUN)
-		message(STATUS "You are probably trying to compile using the g++ standard library which contains a crashing std::regex implementation. Will use boost::regex instead")
+		message(
+			STATUS
+			"You are probably trying to compile using the g++ standard library which contains a crashing std::regex implementation. Will use boost::regex instead"
+		)
+
+		find_package(Boost 1.80 QUIET COMPONENTS regex)
+
+		if(NOT Boost_FOUND)
+			set(BOOST_REGEX_STANDALONE ON)
+
+			FetchContent_Declare(
+				boost-rx
+				GIT_REPOSITORY https://github.com/boostorg/regex
+				GIT_TAG boost-1.83.0)
+
+			FetchContent_MakeAvailable(boost-rx)
+		endif()
 
-		add_subdirectory(regex EXCLUDE_FROM_ALL)
 		set(BOOST_REGEX ON)
 	endif()
 endif()
 
-set(CMAKE_THREAD_PREFER_PTHREAD)
-set(THREADS_PREFER_PTHREAD_FLAG)
-find_package(Threads)
-
 if(MSVC)
-	# Avoid linking the shared library of zlib
-	# Search ZLIB_ROOT first if it is set.
+	# Avoid linking the shared library of zlib Search ZLIB_ROOT first if it is
+	# set.
 	if(ZLIB_ROOT)
 		set(_ZLIB_SEARCH_ROOT PATHS ${ZLIB_ROOT} NO_DEFAULT_PATH)
 		list(APPEND _ZLIB_SEARCHES _ZLIB_SEARCH_ROOT)
@@ -169,115 +205,163 @@ if(MSVC)
 	set(_ZLIB_x86 "(x86)")
 	set(_ZLIB_SEARCH_NORMAL
 		PATHS "[HKEY_LOCAL_MACHINE\\SOFTWARE\\GnuWin32\\Zlib;InstallPath]"
-		"$ENV{ProgramFiles}/zlib"
-		"$ENV{ProgramFiles${_ZLIB_x86}}/zlib")
+		"$ENV{ProgramFiles}/zlib" "$ENV{ProgramFiles${_ZLIB_x86}}/zlib")
 	unset(_ZLIB_x86)
 	list(APPEND _ZLIB_SEARCHES _ZLIB_SEARCH_NORMAL)
 
+	if(BUILD_FOR_CCP4)
+		list(PREPEND _ZLIB_SEARCHES "$ENV{CCP4}/lib")
+	endif()
+
 	foreach(search ${_ZLIB_SEARCHES})
-		find_library(ZLIB_LIBRARY NAMES zlibstatic NAMES_PER_DIR ${${search}} PATH_SUFFIXES lib)
+		find_library(
+			ZLIB_LIBRARY
+			NAMES zlibstatic NAMES_PER_DIR ${${search}}
+			PATH_SUFFIXES lib)
 	endforeach()
 endif()
 
-find_package(ZLIB REQUIRED)
+find_package(ZLIB QUIET)
+find_package(Threads)
 
-find_package(Eigen3 REQUIRED)
-
-include(FindFilesystem)
-list(APPEND CIFPP_REQUIRED_LIBRARIES ${STDCPPFS_LIBRARY})
-
-include(FindAtomic)
-list(APPEND CIFPP_REQUIRED_LIBRARIES ${STDCPPATOMIC_LIBRARY})
-
-if(MSVC)
-	# this dependency can go once MSVC supports std::experimental::is_detected
-	find_package(zeep 5.1.8 REQUIRED)
-	list(APPEND CIFPP_REQUIRED_LIBRARIES zeep::zeep)
+if(NOT ZLIB_FOUND)
+	message(FATAL_ERROR "The zlib development files were not found you this system, please install them and try again (hint: on debian/ubuntu use apt-get install zlib1g-dev)")
 endif()
 
+# Using Eigen3 is a bit of a thing. We don't want to build it completely since
+# we only need a couple of header files. Nothing special. But often, eigen3 is
+# already installed and then we prefer that.
+find_package(Eigen3 3.4 QUIET)
+
+if(Eigen3_FOUND AND TARGET Eigen3::Eigen)
+	get_target_property(EIGEN_INCLUDE_DIR Eigen3::Eigen
+		INTERFACE_INCLUDE_DIRECTORIES)
+else()
+	# Use ExternalProject since FetchContent always tries to install the result...
+	ExternalProject_Add(my-eigen3
+		GIT_REPOSITORY https://gitlab.com/libeigen/eigen.git
+		GIT_TAG 3.4.0
+		INSTALL_COMMAND "")
+	
+	ExternalProject_Get_Property(my-eigen3 SOURCE_DIR)
+	set(EIGEN_INCLUDE_DIR ${SOURCE_DIR})
+endif()
+
+message(STATUS "Eigen include dir is ${EIGEN_INCLUDE_DIR}")
+
 # Create a revision file, containing the current git version info
 include(VersionString)
-write_version_header(${PROJECT_SOURCE_DIR}/src/ "LibCIFPP")
+write_version_header(${CMAKE_CURRENT_SOURCE_DIR}/src/ LIB_NAME "LibCIFPP")
 
 # SymOp data table
 if(CIFPP_RECREATE_SYMOP_DATA)
 	# The tool to create the table
-	add_executable(symop-map-generator "${PROJECT_SOURCE_DIR}/src/symop-map-generator.cpp")
+	add_executable(symop-map-generator
+		"${CMAKE_CURRENT_SOURCE_DIR}/src/symop-map-generator.cpp")
+
+	target_compile_features(symop-map-generator PUBLIC cxx_std_20)
 
 	add_custom_command(
-		OUTPUT ${PROJECT_SOURCE_DIR}/src/symop_table_data.hpp
-		COMMAND $<TARGET_FILE:symop-map-generator> $ENV{CLIBD}/syminfo.lib $ENV{CLIBD}/symop.lib ${PROJECT_SOURCE_DIR}/src/symop_table_data.hpp
-	)
+		OUTPUT ${CMAKE_CURRENT_SOURCE_DIR}/src/symop_table_data.hpp
+		COMMAND
+		$<TARGET_FILE:symop-map-generator> $ENV{CLIBD}/syminfo.lib
+		$ENV{CLIBD}/symop.lib ${CMAKE_CURRENT_SOURCE_DIR}/src/symop_table_data.hpp)
 
 	add_custom_target(
-		OUTPUT ${PROJECT_SOURCE_DIR}/src/symop_table_data.hpp
-		DEPENDS symop-map-generator "$ENV{CLIBD}/syminfo.lib" "$ENV{CLIBD}/symop.lib"
-	)
+		OUTPUT
+		${CMAKE_CURRENT_SOURCE_DIR}/src/symop_table_data.hpp
+		DEPENDS symop-map-generator "$ENV{CLIBD}/syminfo.lib"
+		"$ENV{CLIBD}/symop.lib")
 endif()
 
 # Sources
 set(project_sources
-	${PROJECT_SOURCE_DIR}/src/category.cpp
-	${PROJECT_SOURCE_DIR}/src/condition.cpp
-	${PROJECT_SOURCE_DIR}/src/datablock.cpp
-	${PROJECT_SOURCE_DIR}/src/dictionary_parser.cpp
-	${PROJECT_SOURCE_DIR}/src/file.cpp
-	${PROJECT_SOURCE_DIR}/src/item.cpp
-	${PROJECT_SOURCE_DIR}/src/parser.cpp
-	${PROJECT_SOURCE_DIR}/src/row.cpp
-	${PROJECT_SOURCE_DIR}/src/validate.cpp
-	${PROJECT_SOURCE_DIR}/src/text.cpp
-	${PROJECT_SOURCE_DIR}/src/utilities.cpp
-
-	${PROJECT_SOURCE_DIR}/src/atom_type.cpp
-	${PROJECT_SOURCE_DIR}/src/compound.cpp
-	${PROJECT_SOURCE_DIR}/src/point.cpp
-	${PROJECT_SOURCE_DIR}/src/symmetry.cpp
-
-	${PROJECT_SOURCE_DIR}/src/model.cpp
-
-	${PROJECT_SOURCE_DIR}/src/pdb/cif2pdb.cpp
-	${PROJECT_SOURCE_DIR}/src/pdb/pdb2cif.cpp
-	${PROJECT_SOURCE_DIR}/src/pdb/pdb2cif_remark_3.cpp
-	${PROJECT_SOURCE_DIR}/src/pdb/tls.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/category.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/condition.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/datablock.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/dictionary_parser.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/file.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/item.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/parser.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/row.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/validate.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/text.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/utilities.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/atom_type.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/compound.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/point.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/symmetry.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/model.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/cif2pdb.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/pdb2cif.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/pdb_record.hpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/pdb2cif_remark_3.hpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/pdb2cif_remark_3.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/reconstruct.cpp
+	${CMAKE_CURRENT_SOURCE_DIR}/src/pdb/validate-pdbx.cpp
 )
 
 set(project_headers
-	${PROJECT_SOURCE_DIR}/include/cif++.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/utilities.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/item.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/datablock.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/file.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/validate.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/iterator.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/parser.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/forward_decl.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/dictionary_parser.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/condition.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/category.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/row.hpp
-
-	${PROJECT_SOURCE_DIR}/include/cif++/atom_type.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/compound.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/point.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/symmetry.hpp
-
-	${PROJECT_SOURCE_DIR}/include/cif++/model.hpp
-
-	${PROJECT_SOURCE_DIR}/include/cif++/pdb/cif2pdb.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/pdb/io.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/pdb/pdb2cif.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/pdb/pdb2cif_remark_3.hpp
-	${PROJECT_SOURCE_DIR}/include/cif++/pdb/tls.hpp
+	include/cif++.hpp
+	include/cif++/atom_type.hpp
+	include/cif++/category.hpp
+	include/cif++/compound.hpp
+	include/cif++/condition.hpp
+	include/cif++/datablock.hpp
+	include/cif++/dictionary_parser.hpp
+	include/cif++/exports.hpp
+	include/cif++/file.hpp
+	include/cif++/format.hpp
+	include/cif++/forward_decl.hpp
+	include/cif++/gzio.hpp
+	include/cif++/item.hpp
+	include/cif++/iterator.hpp
+	include/cif++/matrix.hpp
+	include/cif++/model.hpp
+	include/cif++/parser.hpp
+	include/cif++/pdb/cif2pdb.hpp
+	include/cif++/pdb.hpp
+	include/cif++/pdb/io.hpp
+	include/cif++/pdb/pdb2cif.hpp
+	include/cif++/pdb/tls.hpp
+	include/cif++/point.hpp
+	include/cif++/row.hpp
+	include/cif++/symmetry.hpp
+	include/cif++/text.hpp
+	include/cif++/utilities.hpp
+	include/cif++/validate.hpp
 )
 
-add_library(cifpp ${project_sources} ${project_headers} ${PROJECT_SOURCE_DIR}/src/symop_table_data.hpp)
+add_library(cifpp)
 add_library(cifpp::cifpp ALIAS cifpp)
-generate_export_header(cifpp EXPORT_FILE_NAME cif++/exports.hpp)
+
+if(TARGET my-eigen3)
+	add_dependencies(cifpp my-eigen3)
+endif()
+
+target_sources(cifpp
+	PRIVATE ${project_sources}
+	${CMAKE_CURRENT_SOURCE_DIR}/src/symop_table_data.hpp
+	PUBLIC
+	FILE_SET cifpp_headers TYPE HEADERS
+	BASE_DIRS ${CMAKE_CURRENT_SOURCE_DIR}/include
+	FILES ${project_headers}
+)
+
+# The code now really requires C++20
+target_compile_features(cifpp PUBLIC cxx_std_20)
+
+set(CMAKE_DEBUG_POSTFIX d)
+set_target_properties(cifpp PROPERTIES DEBUG_POSTFIX "d")
+
+generate_export_header(cifpp EXPORT_FILE_NAME
+	${CMAKE_CURRENT_SOURCE_DIR}/include/cif++/exports.hpp)
 
 if(BOOST_REGEX)
-	target_compile_definitions(cifpp PRIVATE USE_BOOST_REGEX=1 BOOST_REGEX_STANDALONE=1)
-	target_include_directories(cifpp PRIVATE regex/include)
+	target_compile_definitions(cifpp PRIVATE USE_BOOST_REGEX=1
+		BOOST_REGEX_STANDALONE=1)
+	get_target_property(BOOST_REGEX_INCLUDE_DIR Boost::regex
+		INTERFACE_INCLUDE_DIRECTORIES)
 endif()
 
 if(MSVC)
@@ -286,13 +370,13 @@ endif()
 
 set_target_properties(cifpp PROPERTIES POSITION_INDEPENDENT_CODE ON)
 
-target_include_directories(cifpp
-	PUBLIC
-	"$<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/include;${PROJECT_BINARY_DIR};${EIGEN3_INCLUDE_DIR}>"
+target_include_directories(
+	cifpp
+	PUBLIC "$<BUILD_INTERFACE:${CMAKE_CURRENT_SOURCE_DIR}/include>"
 	"$<INSTALL_INTERFACE:${CMAKE_INSTALL_INCLUDEDIR}>"
-)
+	PRIVATE "${BOOST_REGEX_INCLUDE_DIR}" "${EIGEN_INCLUDE_DIR}")
 
-target_link_libraries(cifpp PUBLIC Threads::Threads ZLIB::ZLIB Eigen3::Eigen ${CIFPP_REQUIRED_LIBRARIES})
+target_link_libraries(cifpp PUBLIC Threads::Threads ZLIB::ZLIB $<$<TARGET_EXISTS:std::atomic>:std::atomic>)
 
 if(CMAKE_CXX_COMPILER_ID STREQUAL "AppleClang")
 	target_link_options(cifpp PRIVATE -undefined dynamic_lookup)
@@ -300,200 +384,204 @@ endif(CMAKE_CXX_COMPILER_ID STREQUAL "AppleClang")
 
 if(CIFPP_DOWNLOAD_CCD)
 	# download the components.cif file from CCD
-	set(COMPONENTS_CIF ${PROJECT_SOURCE_DIR}/data/components.cif)
+	set(COMPONENTS_CIF ${CMAKE_CURRENT_SOURCE_DIR}/rsrc/components.cif)
 
-	if(NOT EXISTS ${COMPONENTS_CIF})
-		if(NOT EXISTS ${PROJECT_SOURCE_DIR}/data)
-			file(MAKE_DIRECTORY ${PROJECT_SOURCE_DIR}/data/)
+	if(EXISTS ${COMPONENTS_CIF})
+		file(SIZE ${COMPONENTS_CIF} CCD_FILE_SIZE)
+
+		if(CCD_FILE_SIZE EQUAL 0)
+			message(STATUS "Removing empty ${COMPONENTS_CIF} file")
+			file(REMOVE "${COMPONENTS_CIF}")
 		endif()
-
-		file(DOWNLOAD https://ftp.wwpdb.org/pub/pdb/data/monomers/components.cif ${COMPONENTS_CIF} SHOW_PROGRESS)
 	endif()
 
-	add_custom_target(COMPONENTS ALL DEPENDS ${COMPONENTS_CIF})
+	if(NOT EXISTS ${COMPONENTS_CIF})
+		# Since the file(DOWNLOAD) command in cmake does not use compression, we try
+		# to download the gzipped version and decompress it ourselves.
+		find_program(GUNZIP gunzip)
+
+		if(WIN32 OR GUNZIP STREQUAL "GUNZIP-NOTFOUND")
+			file(
+				DOWNLOAD https://files.wwpdb.org/pub/pdb/data/monomers/components.cif
+				${COMPONENTS_CIF}
+				SHOW_PROGRESS
+				STATUS CCD_FETCH_STATUS)
+		else()
+			if(NOT EXISTS "${COMPONENTS_CIF}.gz")
+				file(
+					DOWNLOAD
+					https://files.wwpdb.org/pub/pdb/data/monomers/components.cif.gz
+					${COMPONENTS_CIF}.gz
+					SHOW_PROGRESS
+					STATUS CCD_FETCH_STATUS)
+			endif()
+
+			add_custom_command(
+				OUTPUT ${COMPONENTS_CIF}
+				COMMAND "${GUNZIP}" ${COMPONENTS_CIF}.gz
+				WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/rsrc/)
+
+			add_custom_target(COMPONENTS ALL DEPENDS ${COMPONENTS_CIF})
+		endif()
+
+		# Do not continue if downloading went wrong
+		list(POP_FRONT CCD_FETCH_STATUS CCD_FETCH_STATUS_CODE)
+
+		if(CCD_FETCH_STATUS_CODE)
+			message(
+				FATAL_ERROR "Error trying to download CCD file: ${CCD_FETCH_STATUS}")
+		endif()
+	endif()
 endif()
 
 # Installation directories
-set(CIFPP_DATA_DIR "${CMAKE_INSTALL_FULL_DATADIR}/libcifpp")
-target_compile_definitions(cifpp PUBLIC DATA_DIR="${CIFPP_DATA_DIR}")
+if(BUILD_FOR_CCP4)
+	set(CIFPP_DATA_DIR
+		"$ENV{CCP4}/share/libcifpp"
+		CACHE PATH "Directory where dictionary and other static data is stored")
+else()
+	set(CIFPP_DATA_DIR
+		"${CMAKE_INSTALL_FULL_DATADIR}/libcifpp"
+		CACHE PATH "Directory where dictionary and other static data is stored")
+endif()
+
+if(CIFPP_DATA_DIR)
+	target_compile_definitions(cifpp PUBLIC DATA_DIR="${CIFPP_DATA_DIR}")
+	set_target_properties(cifpp PROPERTIES CIFPP_DATA_DIR ${CIFPP_DATA_DIR})
+endif()
+
+if(NOT PROJECT_IS_TOP_LEVEL)
+	set(CIFPP_SHARE_DIR ${CIFPP_DATA_DIR} PARENT_SCOPE)
+endif()
+
+if(UNIX AND NOT BUILD_FOR_CCP4)
+	if("${CMAKE_INSTALL_PREFIX}" STREQUAL "/usr/local")
+		set(CIFPP_CACHE_DIR
+			"/var/cache/libcifpp"
+			CACHE PATH "The directory where downloaded data files are stored")
+	else()
+		set(CIFPP_CACHE_DIR
+			"${CMAKE_INSTALL_FULL_LOCALSTATEDIR}/cache/libcifpp"
+			CACHE PATH "The directory where downloaded data files are stored")
+	endif()
 
-if(UNIX)
-	set(CIFPP_CACHE_DIR "${CMAKE_INSTALL_FULL_LOCALSTATEDIR}/cache/libcifpp")
 	target_compile_definitions(cifpp PUBLIC CACHE_DIR="${CIFPP_CACHE_DIR}")
 
-	set(CIFPP_ETC_DIR "${CMAKE_INSTALL_FULL_SYSCONFDIR}")
+	set(CIFPP_ETC_DIR
+		"${CMAKE_INSTALL_FULL_SYSCONFDIR}"
+		CACHE PATH "The directory where the update configuration file is stored")
+else()
+	unset(CIFPP_CACHE_DIR)
 endif()
 
 # Install rules
 install(TARGETS cifpp
-	EXPORT cifppTargets
-	ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR}
-	LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
-	RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
-	INCLUDES DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+	EXPORT cifpp
+	FILE_SET cifpp_headers DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
 
 if(MSVC AND BUILD_SHARED_LIBS)
 	install(
-		FILES $<TARGET_PDB_FILE:${PROJECT_NAME}>
+		FILES $<TARGET_PDB_FILE:cifpp>
 		DESTINATION ${CMAKE_INSTALL_LIBDIR}
 		OPTIONAL)
 endif()
 
-install(EXPORT cifppTargets
-	FILE "cifppTargets.cmake"
-	NAMESPACE cifpp::
-	DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp
-)
+# Clean up old config files (with old names)
+file(GLOB OLD_CONFIG_FILES
+	${CMAKE_INSTALL_FULL_LIBDIR}/cmake/cifpp/cifppConfig*.cmake
+	${CMAKE_INSTALL_FULL_LIBDIR}/cmake/cifpp/cifppTargets*.cmake)
 
-install(
-	DIRECTORY include/cif++
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
-	COMPONENT Devel
-)
-
-install(
-	FILES include/cif++.hpp
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
-	COMPONENT Devel
-)
-
-install(
-	FILES ${PROJECT_BINARY_DIR}/cif++/exports.hpp
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}/cif++
-	COMPONENT Devel
-)
-
-install(FILES
-	${PROJECT_SOURCE_DIR}/rsrc/mmcif_ddl.dic
-	${PROJECT_SOURCE_DIR}/rsrc/mmcif_pdbx.dic
-	${PROJECT_SOURCE_DIR}/rsrc/mmcif_ma.dic
-	${COMPONENTS_CIF}
-	DESTINATION ${CIFPP_DATA_DIR}
-)
-
-if(${CIFPP_CACHE_DIR})
-	install(FILES
-		${PROJECT_SOURCE_DIR}/rsrc/mmcif_ddl.dic
-		${PROJECT_SOURCE_DIR}/rsrc/mmcif_pdbx.dic
-		${PROJECT_SOURCE_DIR}/rsrc/mmcif_ma.dic
-		${COMPONENTS_CIF}
-		DESTINATION ${CIFPP_CACHE_DIR}
-	)
+if(OLD_CONFIG_FILES)
+	message(
+		STATUS "Installation will remove old config files: ${OLD_CONFIG_FILES}")
+	install(CODE "file(REMOVE ${OLD_CONFIG_FILES})")
 endif()
 
-set(CONFIG_TEMPLATE_FILE ${PROJECT_SOURCE_DIR}/cmake/cifppConfig.cmake.in)
+install(EXPORT cifpp
+	NAMESPACE cifpp::
+	FILE "cifpp-targets.cmake"
+	DESTINATION lib/cmake/cifpp)
+
+install(
+	FILES ${CMAKE_CURRENT_SOURCE_DIR}/rsrc/mmcif_ddl.dic
+	${CMAKE_CURRENT_SOURCE_DIR}/rsrc/mmcif_pdbx.dic
+	${CMAKE_CURRENT_SOURCE_DIR}/rsrc/mmcif_ma.dic
+	DESTINATION ${CMAKE_INSTALL_DATADIR}/libcifpp)
+
+if(CIFPP_DATA_DIR AND CIFPP_DOWNLOAD_CCD)
+	install(FILES ${COMPONENTS_CIF}
+		DESTINATION ${CMAKE_INSTALL_DATADIR}/libcifpp)
+endif()
+
+set(CONFIG_TEMPLATE_FILE ${CMAKE_CURRENT_SOURCE_DIR}/cmake/cifpp-config.cmake.in)
 
 configure_package_config_file(
-	${CONFIG_TEMPLATE_FILE}
-	${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifppConfig.cmake
-	INSTALL_DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp
-	PATH_VARS CIFPP_DATA_DIR
-)
+	${CONFIG_TEMPLATE_FILE} ${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config.cmake
+	INSTALL_DESTINATION lib/cmake/cifpp
+	PATH_VARS CIFPP_DATA_DIR)
 
-install(FILES
-	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifppConfig.cmake"
-	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifppConfigVersion.cmake"
-	DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp
-	COMPONENT Devel
-)
+install(
+	FILES "${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config.cmake"
+	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config-version.cmake"
+	DESTINATION lib/cmake/cifpp)
 
-set(cifpp_MAJOR_VERSION ${CMAKE_PROJECT_VERSION_MAJOR})
-set_target_properties(cifpp PROPERTIES
-	VERSION ${PROJECT_VERSION}
-	SOVERSION ${cifpp_MAJOR_VERSION}
-	INTERFACE_cifpp_MAJOR_VERSION ${cifpp_MAJOR_VERSION})
+set_target_properties(
+	cifpp
+	PROPERTIES VERSION ${PROJECT_VERSION}
+	SOVERSION "${PROJECT_VERSION_MAJOR}.${PROJECT_VERSION_MINOR}"
+	INTERFACE_cifpp_MAJOR_VERSION ${PROJECT_VERSION_MAJOR})
 
-set_property(TARGET cifpp APPEND PROPERTY
-	COMPATIBLE_INTERFACE_STRING cifpp_MAJOR_VERSION
-)
+set_property(
+	TARGET cifpp
+	APPEND
+	PROPERTY COMPATIBLE_INTERFACE_STRING cifpp_MAJOR_VERSION)
 
 write_basic_package_version_file(
-	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifppConfigVersion.cmake"
+	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config-version.cmake"
 	VERSION ${PROJECT_VERSION}
-	COMPATIBILITY AnyNewerVersion
-)
+	COMPATIBILITY AnyNewerVersion)
 
-# pkgconfig support
-set(prefix ${CMAKE_INSTALL_PREFIX})
-set(exec_prefix ${CMAKE_INSTALL_PREFIX})
-set(libdir ${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_LIBDIR})
-set(includedir ${CMAKE_INSTALL_PREFIX}/${CMAKE_INSTALL_INCLUDEDIR})
-
-configure_file(${CMAKE_CURRENT_SOURCE_DIR}/libcifpp.pc.in
-	${CMAKE_CURRENT_BINARY_DIR}/libcifpp.pc.in @ONLY)
-file(GENERATE OUTPUT ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.pc
-	INPUT ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.pc.in)
-install(FILES ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.pc DESTINATION ${CMAKE_INSTALL_LIBDIR}/pkgconfig)
-
-if(ENABLE_TESTING)
-	enable_testing()
-
-	find_package(Boost REQUIRED)
-
-	list(APPEND CIFPP_tests unit-v2 unit-3d format model rename-compound sugar spinner)
-
-	foreach(CIFPP_TEST IN LISTS CIFPP_tests)
-		set(CIFPP_TEST "${CIFPP_TEST}-test")
-		set(CIFPP_TEST_SOURCE "${CMAKE_CURRENT_SOURCE_DIR}/test/${CIFPP_TEST}.cpp")
-
-		add_executable(${CIFPP_TEST} ${CIFPP_TEST_SOURCE})
-
-		target_link_libraries(${CIFPP_TEST} PRIVATE Threads::Threads cifpp::cifpp Boost::boost)
-
-		if(MSVC)
-			# Specify unwind semantics so that MSVC knowns how to handle exceptions
-			target_compile_options(${CIFPP_TEST} PRIVATE /EHsc)
-		endif()
-
-		add_custom_target("run-${CIFPP_TEST}" DEPENDS ${CMAKE_CURRENT_BINARY_DIR}/Run${CIFPP_TEST}.touch ${CIFPP_TEST})
-
-		add_custom_command(
-			OUTPUT ${CMAKE_CURRENT_BINARY_DIR}/Run${CIFPP_TEST}.touch
-			COMMAND $<TARGET_FILE:${CIFPP_TEST}> -- ${CMAKE_CURRENT_SOURCE_DIR}/test)
-
-		add_test(NAME ${CIFPP_TEST}
-			COMMAND $<TARGET_FILE:${CIFPP_TEST}> -- ${CMAKE_CURRENT_SOURCE_DIR}/test)
-	endforeach()
+if(BUILD_TESTING AND PROJECT_IS_TOP_LEVEL)
+	add_subdirectory(test)
 endif()
 
 # Optionally install the update scripts for CCD and dictionary files
 if(CIFPP_INSTALL_UPDATE_SCRIPT)
-	if(${CMAKE_SYSTEM_NAME} STREQUAL "Linux" OR ${CMAKE_SYSTEM_NAME} STREQUAL "GNU")
-		set(CIFPP_CRON_DIR "${CIFPP_ETC_DIR}/cron.weekly")
-	elseif(UNIX) # assume all others are like FreeBSD...
-		set(CIFPP_CRON_DIR "${CIFPP_ETC_DIR}/periodic/weekly")
+	configure_file(${CMAKE_CURRENT_SOURCE_DIR}/tools/update-libcifpp-data.in
+		update-libcifpp-data @ONLY)
+
+	if(${CMAKE_SYSTEM_NAME} STREQUAL "Linux" OR
+		${CMAKE_SYSTEM_NAME} STREQUAL "GNU" OR
+		${CMAKE_SYSTEM_NAME} STREQUAL "FreeBSD")
+		install(
+			FILES ${CMAKE_CURRENT_BINARY_DIR}/update-libcifpp-data
+			DESTINATION ${CMAKE_INSTALL_SYSCONFDIR}/cron.weekly
+			PERMISSIONS OWNER_EXECUTE OWNER_READ GROUP_EXECUTE GROUP_READ WORLD_EXECUTE
+			WORLD_READ)
 	else()
 		message(FATAL_ERROR "Don't know where to install the update script")
 	endif()
 
-	configure_file(${PROJECT_SOURCE_DIR}/tools/update-libcifpp-data.in update-libcifpp-data @ONLY)
-	install(
-		FILES ${CMAKE_CURRENT_BINARY_DIR}/update-libcifpp-data
-		DESTINATION ${CIFPP_CRON_DIR}
-		PERMISSIONS OWNER_EXECUTE OWNER_READ GROUP_EXECUTE GROUP_READ WORLD_EXECUTE WORLD_READ
-	)
-
-	install(DIRECTORY DESTINATION ${CIFPP_CACHE_DIR})
-	install(DIRECTORY DESTINATION "${CIFPP_ETC_DIR}/libcifpp/cache-update.d")
-
-	# a config to, to make it complete
-	if(NOT EXISTS "${CIFPP_ETC_DIR}/libcifpp.conf")
-		file(WRITE ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.conf [[# Uncomment the next line to enable automatic updates
+	# a config file, to make it complete
+	# install(DIRECTORY DESTINATION "${CMAKE_INSTALL_LOCALSTATEDIR}/libcifpp")
+	if(NOT EXISTS "${CMAKE_INSTALL_SYSCONFDIR}/libcifpp.conf")
+		file(
+			WRITE ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.conf
+			[[# Uncomment the next line to enable automatic updates
 # update=true
 ]])
-		install(FILES ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.conf DESTINATION "${CIFPP_ETC_DIR}")
-		install(CODE "message(\"A configuration file has been written to ${CIFPP_ETC_DIR}/libcifpp.conf, please edit this file to enable automatic updates\")")
+		install(FILES ${CMAKE_CURRENT_BINARY_DIR}/libcifpp.conf
+			DESTINATION ${CMAKE_INSTALL_SYSCONFDIR})
+		install(
+			CODE "message(\"A configuration file has been written to ${CIFPP_ETC_DIR}/libcifpp.conf, please edit this file to enable automatic updates\")"
+		)
+
+		install(DIRECTORY DESTINATION ${CMAKE_INSTALL_SYSCONFDIR}/libcifpp/cache-update.d)
 	endif()
 
 	target_compile_definitions(cifpp PUBLIC CACHE_DIR="${CIFPP_CACHE_DIR}")
 endif()
 
-set(CPACK_RESOURCE_FILE_LICENSE "${CMAKE_CURRENT_SOURCE_DIR}/LICENSE")
-set(CPACK_SOURCE_TGZ ON)
-set(CPACK_SOURCE_TBZ2 OFF)
-set(CPACK_SOURCE_TXZ OFF)
-set(CPACK_SOURCE_TZ OFF)
-set(CPACK_SOURCE_IGNORE_FILES "/data/components.cif;/build;/.vscode;/.git")
-set(CPACK_PACKAGE_FILE_NAME "${PROJECT_NAME}-${PROJECT_VERSION}")
-set(CPACK_SOURCE_PACKAGE_FILE_NAME ${CPACK_PACKAGE_FILE_NAME})
-include(CPack)
+if(BUILD_DOCUMENTATION)
+	add_subdirectory(docs)
+endif()

README.md

@@ -1,12 +1,38 @@
-libcifpp
-========
+[![github CI](https://github.com/pdb-redo/libcifpp/actions/workflows/cmake-multi-platform.yml/badge.svg)](https://github.com/pdb-redo/libcifpp/actions)
+[![GitHub License](https://img.shields.io/github/license/pdb-redo/libcifpp)](https://github.com/pdb-redo/libcifpp/LICENSE)
 
-This library contains code to work with mmCIF and PDB files.
+# libcifpp
 
-Synopsis
---------
+As the name implies, this library was originally written to work with mmCIF files
+using C++ as programming language. The design of this library leanes heavily on
+the structure of CIF files. These files can be thought of as a text dump of a
+relational databank with, often but not always, a very strict schema describing
+the data. These schema's are called dictionaries.
 
-```c++
+Using information from the content of a mmCIF file and an optional schema,
+libcifpp allows you to access the data in the file as a collection of datablock
+each containing a collection of categories with rows of data. The categories can
+be searched for data using queries written in regular C++ syntax. When a dictionary
+was specified, inserted data is checked for validity. Likewise removal of data
+may result in cascaded removal of linked data in other categories using
+parent/child relationship information.
+
+Since there were still many programs using the legacy PDB format at the time
+development started, a layer was added that converts data to and from PDB format
+into mmCIF format. This means you can manipulate PDB files as if they were
+normal mmCIF files.
+
+Apart from this basic functionality, libcifpp also offers code to help with
+symmetry calculations, 3d manipulations and obtaining information from the CCD
+[Chemical Component Dictionary](https://www.wwpdb.org/data/ccd).
+
+## Documentation
+
+The documentation can be found at [github.io](https://pdb-redo.github.io/libcifpp/)
+
+## Synopsis
+
+```cpp
 // A simple program counting residues with an OXT atom
 
 #include <filesystem>
@@ -26,7 +52,7 @@ int main(int argc, char *argv[])
 
     if (file.empty())
     {
-        std::cerr << "Empty file" << std::endl;
+        std::cerr << "Empty file\n";
         exit(1);
     }
 
@@ -40,8 +66,8 @@ int main(int argc, char *argv[])
     auto n = atom_site.count(cif::key("label_atom_id") == "OXT");
 
     std::cout << "File contains " << atom_site.size() << " atoms of which "
-              << n << (n == 1 ? " is" : " are") << " OXT" << std::endl
-              << "residues with an OXT are:" << std::endl;
+              << n << (n == 1 ? " is" : " are") << " OXT\n"
+              << "residues with an OXT are:\n";
 
     // Loop over all atoms with atom-id "OXT" and print out some info.
     // That info is extracted using structured binding in C++
@@ -50,59 +76,150 @@ int main(int argc, char *argv[])
                 cif::key("label_atom_id") == "OXT",
                 "label_asym_id", "label_comp_id", "label_seq_id"))
     {
-        std::cout << asym << ' ' << comp << ' ' << seqnr << std::endl;
+        std::cout << asym << ' ' << comp << ' ' << seqnr << '\n';
     }
 
     return 0;
 }
-
 ```
 
-Requirements
-------------
+## Installation
+
+You might be able to use libcifpp from a package manager used by your
+OS distribution. But most likely this package will be out-of-date.
+Therefore it is recommended to build *libcifpp* from code. It is not
+hard to do. But it is recommended to read the following instructions
+carefully.
+
+### Requirements
 
 The code for this library was written in C++17. You therefore need a
-recent compiler to build it. For the development gcc 9.4 and clang 9.0
+recent compiler to build it. For the development gcc >= 9.4 and clang >= 9.0
 have been used as well as MSVC version 2019.
 
-Other requirements are:
+The other requirement you really need to have installed on your computer
+is a version of [CMake](https://cmake.org). For now the minimum version
+is 3.16 but that may soon change into a higher version. You should also
+install the gui version of CMake to set build options easily, on Debian
+I prefer to use the curses version installed with `cmake-curses-gui`.
+
+It is very useful to have [mrc](https://github.com/mhekkel/mrc) available.
+However, this is only an option if you use Windows or an operating system
+using the ELF executable format (i.e. Linux or FreeBSD). MRC is a resource
+compiler that allows including data files into the executable making them
+easier to install.
+
+Other libraries you might want to install beforehand are:
 
-- [mrc](https://github.com/mhekkel/mrc), a resource compiler that
-  allows including data files into the executable making them easier to
-  install. Strictly speaking this is optional, but at the expense of
-  functionality.
 - [libeigen](https://eigen.tuxfamily.org/index.php?title=Main_Page), a
   library to do amongst others matrix calculations. This usually can be
   installed using your package manager, in Debian/Ubuntu it is called
   `libeigen3-dev`
-- zlib, the development version of this library. On Debian/Ubuntu this
-  is the package `zlib1g-dev`.
-- [boost](https://www.boost.org). The boost libraries are only needed if
-  you want to build the testing code.
+- [zlib](https://github.com/madler/zlib), the development version of this
+  library. On Debian/Ubuntu this is the package `zlib1g-dev`.
+- [boost](https://www.boost.org), in Debian/Ubuntu this is `libboost-dev`.
+  
+  The Boost libraries are only needed in case you are using GCC due to a long
+  standing bug in GNU's implementation of std::regex. It simply crashes
+  on the regular expressions used in the mmcif_pdbx dictionary and so
+  we use the boost regex implementation instead.
 
-When building using MS Visual Studio, you will also need [libzeep](https://github.com/mhekkel/libzeep)
-since MSVC does not yet provide a C++ template required by libcifpp.
+### Building
 
-Building
---------
+First you need to download the code:
 
-This library uses [cmake](https://cmake.org). The usual way of building
-and installing is to create a `build` directory and run cmake there.
-
-On linux e.g. you would issue the following commands to build and install
-libcifpp in your `$HOME/.local` folder:
-
-```bash
- git clone https://github.com/PDB-REDO/libcifpp.git --recurse-submodules
+```console
+ git clone https://github.com/PDB-REDO/libcifpp.git
  cd libcifpp
- cmake -S . -B build -DCMAKE_INSTALL_PREFIX=$HOME/.local -DCMAKE_BUILD_TYPE=Release
- cmake --build build
- cmake --install build
 ```
 
-This checks out the source code from github, creates a new directory
-where cmake stores its files. Run a configure, build the code and then
-it installs the library and auxiliary files.
+You should start by considering where to install libcifpp. If you have
+sufficient permissions on your computer you perhaps should use the
+default but libcifpp can be configured to be installed anywhere
+including e.g. *$HOME/.local*.
 
-If you want to run the tests before installing, you should add `-DENABLE_TESTING=ON`
-to the first cmake command.
+Next step is to configure, for this use the CMake gui application. If you
+installed the curses version of cmake you can type `ccmake`. On Windows
+you can use `cmake-gui.exe`.
+
+To install in the default location:
+
+```console
+ccmake -S . -B build
+```
+
+To install elsewhere, e.g. *$HOME/.local*:
+
+```console
+ccmake -S . -B build -DCMAKE_INSTALL_PREFIX=$HOME/.local
+```
+
+In the cmake window, start the configure command (use button or press 'c').
+After the first configure step you will see a list of settable options.
+Alter these to match your preferences. Most options are self explaining
+and contain a description. Some may need a bit more explanation:
+
+- CIFPP_DATA_DIR, this directory will be used to store initial versions
+  of the mmcif_pdbx dictionary as well as the optional CCD file.
+
+- CIFPP_DOWNLOAD_CCD
+
+  The CCD file is huge and perhaps you think you don't
+  need it. In that case you can leave this OFF. But that will limit the
+  use cases.
+
+- CIFPP_INSTALL_UPDATE_SCRIPT
+  
+  The files in CIFPP_DATA_DIR are quickly becoming out of date. On
+  FreeBSD and Linux you can install a script that updates these files
+  on a weekly basis.
+
+- CIFPP_CRON_DIR
+  
+  The directory where the update script is to be installed.
+
+- CIFPP_ETC_DIR
+  
+  The update script will only work if the file called *libcifpp.conf*
+  in this *etc* directory will contain an uncommented line with
+
+```console
+update=true
+```
+
+- CIFPP_CACHE_DIR
+
+  When you installed and enabled the update script, new files are
+  written to this directory.
+
+- CIFPP_RECREATE_SYMOP_DATA
+  
+  If you had CCP4 sourced into your environment, this option allows
+  you to recreate the symop data file.
+
+- BUILD_FOR_CCP4
+  
+  Build a special version of libcifpp to be installed in the CCP4
+  environment.
+
+After setting these options you can run the configure step again and
+then use generate to create the makefiles.
+
+Building and installing is then as simple as:
+
+```console
+cmake --build build
+cmake --install build
+```
+
+If this fails due to lack of permissions, you can try:
+
+```console
+sudo cmake --install build
+```
+
+Tests are created by default, and to test the code you can run:
+
+```console
+ctest --test-dir build
+```

changelog

@@ -1,8 +1,112 @@
+Version 8.0.0
+- A dictionary is for a datablock and a file can have
+  datablocks with differing dictionaries.
+
+Version 7.0.10
+- Deal with missing _entity.type in reconstructing mmCIF files
+- Replace code creating quaternions from rotation matrices
+  that might sometimes give incorrect results. Or at least,
+  the test code failed on this particular kind of code. Sometimes.
+- Fix reconstruction to build pdbx_nonpoly_scheme
+
+Version 7.0.9
+- Using cif::file::load_dictionary it is now possible to
+  load a dictionary along with its extensions in one go.
+  E.g. file.load_dictionary("mmcif_pdbx;dssp-extension")
+- Fix in compound factory to avoid errors with lower case
+  compound id's
+- Fix sac_parser's index to be case insensitive
+
+Version 7.0.8
+- Fix PDB Remark 3 parser
+- Added three way comparison for point
+
+Version 7.0.7
+- Set CIFPP_DATA_DIR on target cifpp for use in projects that include
+  libcifpp directly
+
+Version 7.0.6
+- Fix linking to std::atomic
+
+Version 7.0.5
+- Fix case where category index was not updated for updated value
+
+Version 7.0.4
+- Do not install headers and library in case we're not the top project
+
+Version 7.0.3
+- Fix installation, write exports.hpp again
+
+Version 7.0.2
+- Fix in testing error_code results.
+
+Version 7.0.1
+- Various reconstruction fixes
+- category order in output fixed
+- better implementation of constructors for file, datablock and category
+- small optimisation in iterator
+
+Version 7.0.0
+- Renaming many methods and parameters to be more
+  consistent with the mmCIF dictionaries.
+  (Most notably, item used to be called column or
+  tag sometimes).
+- validation_error is now a std::system_error error
+  value. The exception is gone.
+- Added repairSequenceInfo to repair invalid files
+
+Version 6.1.0
+- Add formula weight to entity in pdb2cif
+- Change order of categories inside a datablock to match order in file
+- Change default order to write out categories in a file based on
+  parent/child relationship
+- Added validate_pdbx and recover_pdbx
+- Fixed a serious bug in category_index when moving categories
+
+Version 6.0.0
+- Drop the use of CCP4's monomer library for compound information
+
+Version 5.2.5
+- Correctly import the Eigen3 library
+
+Version 5.2.4
+- Changes required to build on Windows
+
+Version 5.2.3
+- New constructors for cif::item, one taking std::optional values
+  and another taking only a name resulting in a value '.' (i.e. inapplicable).
+- added cif::cell::get_volume
+
+Version 5.2.2
+- Remove dependency on Eigen3 for users of libcifpp
+- Fix typos in documentation
+- Do not build latex files in documentation
+- Fixed conversion from string to integer, would fail on +2 e.g.
+- sqrt is not constexpr, thus kGoldenRatio should be const, not constexpr
+
+Version 5.2.1
+- New versionstring module
+- small fixes for generating documentation
+- correctly setting SONAME
+
+Version 5.2.0
+- With lots of documentation
+- Refactored coloured text output
+- Removed the subdirectory cif++/pdb, there now is a single
+  header file pdb.hpp for I/O of legacy PDB files.
+
+Version 5.1.3
+- Dropped pkgconfig support
+
+Version 5.1.2
+- New version string code
+- Added check for Eigen3 in cifppConfig.cmake
+
 Version 5.1.1
 - Added missing include <compare> in symmetry.hpp
 - Added empty() to matrix
-- Fix for parsing PDB files with a last line that does not end with
-  a new line character.
+- Fix for parsing legacy PDB files with a last line that does
+  not end with a new line character.
 
 Version 5.1
 - New parser, optimised for speed

cmake/FindAtomic.cmake

@@ -56,6 +56,7 @@ endif()
 cmake_pop_check_state()
 
 set(Atomic_FOUND ${_found} CACHE BOOL "TRUE if we can run a program using std::atomic" FORCE)
+mark_as_advanced(Atomic_FOUND)
 
 if(Atomic_FIND_REQUIRED AND NOT Atomic_FOUND)
     message(FATAL_ERROR "Cannot run simple program using std::atomic")

cmake/FindFilesystem.cmake

@@ -1,74 +0,0 @@
-# Simplistic reimplementation of https://github.com/vector-of-bool/CMakeCM/blob/master/modules/FindFilesystem.cmake
-
-if(TARGET std::filesystem)
-	return()
-endif()
-
-cmake_minimum_required(VERSION 3.10)
-
-include(CMakePushCheckState)
-include(CheckIncludeFileCXX)
-include(CheckCXXSourceCompiles)
-
-cmake_push_check_state()
-
-check_include_file_cxx("filesystem" _CXX_FILESYSTEM_HAVE_HEADER)
-mark_as_advanced(_CXX_FILESYSTEM_HAVE_HEADER)
-
-set(code [[
-#include <cstdlib>
-#include <filesystem>
-
-int main() {
-	auto cwd = std::filesystem::current_path();
-	return EXIT_SUCCESS;
-}
-]])
-
-if(CMAKE_CXX_COMPILER_ID STREQUAL "GNU" AND CMAKE_CXX_COMPILER_VERSION VERSION_LESS_EQUAL 8.4.0)
-	# >> https://stackoverflow.com/questions/63902528/program-crashes-when-filesystempath-is-destroyed
-	set(CXX_FILESYSTEM_NO_LINK_NEEDED 0)
-else()
-	# Check a simple filesystem program without any linker flags
-	check_cxx_source_compiles("${code}" CXX_FILESYSTEM_NO_LINK_NEEDED)
-endif()
-
-if(CXX_FILESYSTEM_NO_LINK_NEEDED)
-	set(_found 1)
-else()
-	set(prev_libraries ${CMAKE_REQUIRED_LIBRARIES})
-	# Add the libstdc++ flag
-	set(CMAKE_REQUIRED_LIBRARIES ${prev_libraries} -lstdc++fs)
-	check_cxx_source_compiles("${code}" CXX_FILESYSTEM_STDCPPFS_NEEDED)
-	set(_found ${CXX_FILESYSTEM_STDCPPFS_NEEDED})
-	if(NOT CXX_FILESYSTEM_STDCPPFS_NEEDED)
-		# Try the libc++ flag
-		set(CMAKE_REQUIRED_LIBRARIES ${prev_libraries} -lc++fs)
-		check_cxx_source_compiles("${code}" CXX_FILESYSTEM_CPPFS_NEEDED)
-		set(_found ${CXX_FILESYSTEM_CPPFS_NEEDED})
-	endif()
-endif()
-
-if(_found)
-	add_library(std::filesystem INTERFACE IMPORTED)
-	set_property(TARGET std::filesystem APPEND PROPERTY INTERFACE_COMPILE_FEATURES cxx_std_17)
-
-	if(CXX_FILESYSTEM_NO_LINK_NEEDED)
-		# Nothing to add...
-	elseif(CXX_FILESYSTEM_STDCPPFS_NEEDED)
-		set_target_properties(std::filesystem PROPERTIES IMPORTED_LIBNAME stdc++fs)
-		set(STDCPPFS_LIBRARY stdc++fs)
-	elseif(CXX_FILESYSTEM_CPPFS_NEEDED)
-		set_target_properties(std::filesystem PROPERTIES IMPORTED_LIBNAME c++fs)
-		set(STDCPPFS_LIBRARY c++fs)
-	endif()
-endif()
-
-cmake_pop_check_state()
-
-set(Filesystem_FOUND ${_found} CACHE BOOL "TRUE if we can run a program using std::filesystem" FORCE)
-
-if(Filesystem_FIND_REQUIRED AND NOT Filesystem_FOUND)
-    message(FATAL_ERROR "Cannot run simple program using std::filesystem")
-endif()
-

cmake/FindSphinx.cmake

@@ -0,0 +1,11 @@
+#Look for an executable called sphinx-build
+find_program(SPHINX_EXECUTABLE
+             NAMES sphinx-build
+             DOC "Path to sphinx-build executable")
+
+include(FindPackageHandleStandardArgs)
+
+#Handle standard arguments to find_package like REQUIRED and QUIET
+find_package_handle_standard_args(Sphinx
+                                  "Failed to find sphinx-build executable"
+                                  SPHINX_EXECUTABLE)

cmake/GetGitRevisionDescription.cmake

@@ -1,284 +0,0 @@
-# - Returns a version string from Git
-#
-# These functions force a re-configure on each git commit so that you can
-# trust the values of the variables in your build system.
-#
-#  get_git_head_revision(<refspecvar> <hashvar> [ALLOW_LOOKING_ABOVE_CMAKE_SOURCE_DIR])
-#
-# Returns the refspec and sha hash of the current head revision
-#
-#  git_describe(<var> [<additional arguments to git describe> ...])
-#
-# Returns the results of git describe on the source tree, and adjusting
-# the output so that it tests false if an error occurs.
-#
-#  git_describe_working_tree(<var> [<additional arguments to git describe> ...])
-#
-# Returns the results of git describe on the working tree (--dirty option),
-# and adjusting the output so that it tests false if an error occurs.
-#
-#  git_get_exact_tag(<var> [<additional arguments to git describe> ...])
-#
-# Returns the results of git describe --exact-match on the source tree,
-# and adjusting the output so that it tests false if there was no exact
-# matching tag.
-#
-#  git_local_changes(<var>)
-#
-# Returns either "CLEAN" or "DIRTY" with respect to uncommitted changes.
-# Uses the return code of "git diff-index --quiet HEAD --".
-# Does not regard untracked files.
-#
-# Requires CMake 2.6 or newer (uses the 'function' command)
-#
-# Original Author:
-# 2009-2020 Ryan Pavlik <ryan.pavlik@gmail.com> <abiryan@ryand.net>
-# http://academic.cleardefinition.com
-#
-# Copyright 2009-2013, Iowa State University.
-# Copyright 2013-2020, Ryan Pavlik
-# Copyright 2013-2020, Contributors
-# SPDX-License-Identifier: BSL-1.0
-# Distributed under the Boost Software License, Version 1.0.
-# (See accompanying file LICENSE_1_0.txt or copy at
-# http://www.boost.org/LICENSE_1_0.txt)
-
-if(__get_git_revision_description)
-    return()
-endif()
-set(__get_git_revision_description YES)
-
-# We must run the following at "include" time, not at function call time,
-# to find the path to this module rather than the path to a calling list file
-get_filename_component(_gitdescmoddir ${CMAKE_CURRENT_LIST_FILE} PATH)
-
-# Function _git_find_closest_git_dir finds the next closest .git directory
-# that is part of any directory in the path defined by _start_dir.
-# The result is returned in the parent scope variable whose name is passed
-# as variable _git_dir_var. If no .git directory can be found, the
-# function returns an empty string via _git_dir_var.
-#
-# Example: Given a path C:/bla/foo/bar and assuming C:/bla/.git exists and
-# neither foo nor bar contain a file/directory .git. This wil return
-# C:/bla/.git
-#
-function(_git_find_closest_git_dir _start_dir _git_dir_var)
-    set(cur_dir "${_start_dir}")
-    set(git_dir "${_start_dir}/.git")
-    while(NOT EXISTS "${git_dir}")
-        # .git dir not found, search parent directories
-        set(git_previous_parent "${cur_dir}")
-        get_filename_component(cur_dir "${cur_dir}" DIRECTORY)
-        if(cur_dir STREQUAL git_previous_parent)
-            # We have reached the root directory, we are not in git
-            set(${_git_dir_var}
-                ""
-                PARENT_SCOPE)
-            return()
-        endif()
-        set(git_dir "${cur_dir}/.git")
-    endwhile()
-    set(${_git_dir_var}
-        "${git_dir}"
-        PARENT_SCOPE)
-endfunction()
-
-function(get_git_head_revision _refspecvar _hashvar)
-    _git_find_closest_git_dir("${CMAKE_CURRENT_SOURCE_DIR}" GIT_DIR)
-
-    if("${ARGN}" STREQUAL "ALLOW_LOOKING_ABOVE_CMAKE_SOURCE_DIR")
-        set(ALLOW_LOOKING_ABOVE_CMAKE_SOURCE_DIR TRUE)
-    else()
-        set(ALLOW_LOOKING_ABOVE_CMAKE_SOURCE_DIR FALSE)
-    endif()
-    if(NOT "${GIT_DIR}" STREQUAL "")
-        file(RELATIVE_PATH _relative_to_source_dir "${CMAKE_SOURCE_DIR}"
-             "${GIT_DIR}")
-        if("${_relative_to_source_dir}" MATCHES "[.][.]" AND NOT ALLOW_LOOKING_ABOVE_CMAKE_SOURCE_DIR)
-            # We've gone above the CMake root dir.
-            set(GIT_DIR "")
-        endif()
-    endif()
-    if("${GIT_DIR}" STREQUAL "")
-        set(${_refspecvar}
-            "GITDIR-NOTFOUND"
-            PARENT_SCOPE)
-        set(${_hashvar}
-            "GITDIR-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-
-    # Check if the current source dir is a git submodule or a worktree.
-    # In both cases .git is a file instead of a directory.
-    #
-    if(NOT IS_DIRECTORY ${GIT_DIR})
-        # The following git command will return a non empty string that
-        # points to the super project working tree if the current
-        # source dir is inside a git submodule.
-        # Otherwise the command will return an empty string.
-        #
-        execute_process(
-            COMMAND "${GIT_EXECUTABLE}" rev-parse
-                    --show-superproject-working-tree
-            WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
-            OUTPUT_VARIABLE out
-            ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
-        if(NOT "${out}" STREQUAL "")
-            # If out is empty, GIT_DIR/CMAKE_CURRENT_SOURCE_DIR is in a submodule
-            file(READ ${GIT_DIR} submodule)
-            string(REGEX REPLACE "gitdir: (.*)$" "\\1" GIT_DIR_RELATIVE
-                                 ${submodule})
-            string(STRIP ${GIT_DIR_RELATIVE} GIT_DIR_RELATIVE)
-            get_filename_component(SUBMODULE_DIR ${GIT_DIR} PATH)
-            get_filename_component(GIT_DIR ${SUBMODULE_DIR}/${GIT_DIR_RELATIVE}
-                                   ABSOLUTE)
-            set(HEAD_SOURCE_FILE "${GIT_DIR}/HEAD")
-        else()
-            # GIT_DIR/CMAKE_CURRENT_SOURCE_DIR is in a worktree
-            file(READ ${GIT_DIR} worktree_ref)
-            # The .git directory contains a path to the worktree information directory
-            # inside the parent git repo of the worktree.
-            #
-            string(REGEX REPLACE "gitdir: (.*)$" "\\1" git_worktree_dir
-                                 ${worktree_ref})
-            string(STRIP ${git_worktree_dir} git_worktree_dir)
-            _git_find_closest_git_dir("${git_worktree_dir}" GIT_DIR)
-            set(HEAD_SOURCE_FILE "${git_worktree_dir}/HEAD")
-        endif()
-    else()
-        set(HEAD_SOURCE_FILE "${GIT_DIR}/HEAD")
-    endif()
-    set(GIT_DATA "${CMAKE_CURRENT_BINARY_DIR}/CMakeFiles/git-data")
-    if(NOT EXISTS "${GIT_DATA}")
-        file(MAKE_DIRECTORY "${GIT_DATA}")
-    endif()
-
-    if(NOT EXISTS "${HEAD_SOURCE_FILE}")
-        return()
-    endif()
-    set(HEAD_FILE "${GIT_DATA}/HEAD")
-    configure_file("${HEAD_SOURCE_FILE}" "${HEAD_FILE}" COPYONLY)
-
-    configure_file("${_gitdescmoddir}/GetGitRevisionDescription.cmake.in"
-                   "${GIT_DATA}/grabRef.cmake" @ONLY)
-    include("${GIT_DATA}/grabRef.cmake")
-
-    set(${_refspecvar}
-        "${HEAD_REF}"
-        PARENT_SCOPE)
-    set(${_hashvar}
-        "${HEAD_HASH}"
-        PARENT_SCOPE)
-endfunction()
-
-function(git_describe _var)
-    if(NOT GIT_FOUND)
-        find_package(Git QUIET)
-    endif()
-    get_git_head_revision(refspec hash)
-    if(NOT GIT_FOUND)
-        set(${_var}
-            "GIT-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-    if(NOT hash)
-        set(${_var}
-            "HEAD-HASH-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-
-    # TODO sanitize
-    #if((${ARGN}" MATCHES "&&") OR
-    #	(ARGN MATCHES "||") OR
-    #	(ARGN MATCHES "\\;"))
-    #	message("Please report the following error to the project!")
-    #	message(FATAL_ERROR "Looks like someone's doing something nefarious with git_describe! Passed arguments ${ARGN}")
-    #endif()
-
-    #message(STATUS "Arguments to execute_process: ${ARGN}")
-
-    execute_process(
-        COMMAND "${GIT_EXECUTABLE}" describe --tags --always ${hash} ${ARGN}
-        WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
-        RESULT_VARIABLE res
-        OUTPUT_VARIABLE out
-        ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
-    if(NOT res EQUAL 0)
-        set(out "${out}-${res}-NOTFOUND")
-    endif()
-
-    set(${_var}
-        "${out}"
-        PARENT_SCOPE)
-endfunction()
-
-function(git_describe_working_tree _var)
-    if(NOT GIT_FOUND)
-        find_package(Git QUIET)
-    endif()
-    if(NOT GIT_FOUND)
-        set(${_var}
-            "GIT-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-
-    execute_process(
-        COMMAND "${GIT_EXECUTABLE}" describe --dirty ${ARGN}
-        WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
-        RESULT_VARIABLE res
-        OUTPUT_VARIABLE out
-        ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
-    if(NOT res EQUAL 0)
-        set(out "${out}-${res}-NOTFOUND")
-    endif()
-
-    set(${_var}
-        "${out}"
-        PARENT_SCOPE)
-endfunction()
-
-function(git_get_exact_tag _var)
-    git_describe(out --exact-match ${ARGN})
-    set(${_var}
-        "${out}"
-        PARENT_SCOPE)
-endfunction()
-
-function(git_local_changes _var)
-    if(NOT GIT_FOUND)
-        find_package(Git QUIET)
-    endif()
-    get_git_head_revision(refspec hash)
-    if(NOT GIT_FOUND)
-        set(${_var}
-            "GIT-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-    if(NOT hash)
-        set(${_var}
-            "HEAD-HASH-NOTFOUND"
-            PARENT_SCOPE)
-        return()
-    endif()
-
-    execute_process(
-        COMMAND "${GIT_EXECUTABLE}" diff-index --quiet HEAD --
-        WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
-        RESULT_VARIABLE res
-        OUTPUT_VARIABLE out
-        ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
-    if(res EQUAL 0)
-        set(${_var}
-            "CLEAN"
-            PARENT_SCOPE)
-    else()
-        set(${_var}
-            "DIRTY"
-            PARENT_SCOPE)
-    endif()
-endfunction()

cmake/GetGitRevisionDescription.cmake.in

@@ -1,43 +0,0 @@
-#
-# Internal file for GetGitRevisionDescription.cmake
-#
-# Requires CMake 2.6 or newer (uses the 'function' command)
-#
-# Original Author:
-# 2009-2010 Ryan Pavlik <rpavlik@iastate.edu> <abiryan@ryand.net>
-# http://academic.cleardefinition.com
-# Iowa State University HCI Graduate Program/VRAC
-#
-# Copyright 2009-2012, Iowa State University
-# Copyright 2011-2015, Contributors
-# Distributed under the Boost Software License, Version 1.0.
-# (See accompanying file LICENSE_1_0.txt or copy at
-# http://www.boost.org/LICENSE_1_0.txt)
-# SPDX-License-Identifier: BSL-1.0
-
-set(HEAD_HASH)
-
-file(READ "@HEAD_FILE@" HEAD_CONTENTS LIMIT 1024)
-
-string(STRIP "${HEAD_CONTENTS}" HEAD_CONTENTS)
-if(HEAD_CONTENTS MATCHES "ref")
-	# named branch
-	string(REPLACE "ref: " "" HEAD_REF "${HEAD_CONTENTS}")
-	if(EXISTS "@GIT_DIR@/${HEAD_REF}")
-		configure_file("@GIT_DIR@/${HEAD_REF}" "@GIT_DATA@/head-ref" COPYONLY)
-	else()
-		configure_file("@GIT_DIR@/packed-refs" "@GIT_DATA@/packed-refs" COPYONLY)
-		file(READ "@GIT_DATA@/packed-refs" PACKED_REFS)
-		if(${PACKED_REFS} MATCHES "([0-9a-z]*) ${HEAD_REF}")
-			set(HEAD_HASH "${CMAKE_MATCH_1}")
-		endif()
-	endif()
-else()
-	# detached HEAD
-	configure_file("@GIT_DIR@/HEAD" "@GIT_DATA@/head-ref" COPYONLY)
-endif()
-
-if(NOT HEAD_HASH)
-	file(READ "@GIT_DATA@/head-ref" HEAD_HASH LIMIT 1024)
-	string(STRIP "${HEAD_HASH}" HEAD_HASH)
-endif()

cmake/VersionString.cmake

@@ -1,6 +1,6 @@
 # SPDX-License-Identifier: BSD-2-Clause
 
-# Copyright (c) 2021 NKI/AVL, Netherlands Cancer Institute
+# Copyright (c) 2021-2023 Maarten L. Hekkelman
 
 # Redistribution and use in source and binary forms, with or without
 # modification, are permitted provided that the following conditions are met:
@@ -22,60 +22,254 @@
 # (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
 # SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
 
+# This cmake extension writes out a revision.hpp file in a specified directory.
+# The file will contain a C++ inline function that can be used to write out
+# version information.
+
 cmake_minimum_required(VERSION 3.15)
 
+# We want the revision.hpp file to be updated whenever the status of the
+# git repository changes. Use the same technique as in GetGitRevisionDescription.cmake
+# from https://github.com/rpavlik/cmake-modules
+
+
+#[=======================================================================[.rst:
+.. command:: write_version_header
+
+  Write a file named revision.hpp containing version info::
+
+	write_version_header(<destdir>
+	                     [FILE_NAME <file-name>]
+						 [LIB_NAME <library-name>]
+	                    )
+  
+  This command will generate the code to write a file name
+  revision.hpp in the directory ``<destdir>``.
+  
+  ``FILE_NAME``
+	Specify the name of the file to create, default is ``revision.hpp``.
+
+  ``LIB_NAME``
+	Specify the library name which will be used as a prefix part for the
+	variables contained in the revision file.
+#]=======================================================================]
+
+# Record the location of this module now, not at the time the CMakeLists.txt
+# is being processed
+get_filename_component(_current_cmake_module_dir ${CMAKE_CURRENT_LIST_FILE} PATH)
+
+# First locate a .git file or directory.
+function(_get_git_dir _start_dir _variable)
+
+	set(cur_dir "${_start_dir}")
+	set(git_dir "${_start_dir}/.git")
+
+	while(NOT EXISTS "${git_dir}")
+		# .git dir not found, search parent directories
+		set(prev_dir "${cur_dir}")
+		get_filename_component(cur_dir "${cur_dir}" DIRECTORY)
+		if(cur_dir STREQUAL prev_dir OR cur_dir STREQUAL ${_start_dir})
+			# we are not in git since we either hit root or
+			# the ${_start_dir} which should be the top
+			set(${_variable} "" PARENT_SCOPE)
+			return()
+		endif()
+		set(git_dir "${cur_dir}/.git")
+	endwhile()
+
+	set(${_variable} "${git_dir}" PARENT_SCOPE)
+endfunction()
+
+# Locate the git refspec hash and load the hash
+# This code locates the file containing the git refspec/hash
+# and loads it. Doing it this way assures that each time the git
+# repository changes the revision.hpp file gets out of date.
+function(_get_git_hash _data_dir _variable)
+
+	# Be pessimistic
+	set(_variable "" PARENT_SCOPE)
+
+	# Load git package if needed
+	if(NOT GIT_FOUND)
+		find_package(Git QUIET)
+	endif()
+
+	# And fail if not found
+	if(NOT GIT_FOUND)
+		return()
+	endif()
+
+	# Locate the nearest .git file or directory
+	_get_git_dir(${CMAKE_CURRENT_SOURCE_DIR} GIT_DIR)
+
+	# And fail if not found
+	if("${GIT_DIR}" STREQUAL "")
+        return()
+    endif()
+
+    # Check if the current source dir is a git submodule or a worktree.
+    # In both cases .git is a file instead of a directory.
+    #
+    if(IS_DIRECTORY ${GIT_DIR})
+		set(HEAD_SOURCE_FILE "${GIT_DIR}/HEAD")
+	else()
+		# The following git command will return a non empty string that
+        # points to the super project working tree if the current
+        # source dir is inside a git submodule.
+        # Otherwise the command will return an empty string.
+        #
+        execute_process(
+            COMMAND "${GIT_EXECUTABLE}" rev-parse
+                    --show-superproject-working-tree
+            WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
+            OUTPUT_VARIABLE out
+            ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
+        if(NOT "${out}" STREQUAL "")
+            # If out is not empty, GIT_DIR/CMAKE_CURRENT_SOURCE_DIR is in a submodule
+            file(READ ${GIT_DIR} submodule)
+            string(REGEX REPLACE "gitdir: (.*)$" "\\1" GIT_DIR_RELATIVE
+                                 ${submodule})
+            string(STRIP ${GIT_DIR_RELATIVE} GIT_DIR_RELATIVE)
+            get_filename_component(SUBMODULE_DIR ${GIT_DIR} PATH)
+            get_filename_component(GIT_DIR ${SUBMODULE_DIR}/${GIT_DIR_RELATIVE}
+                                   ABSOLUTE)
+            set(HEAD_SOURCE_FILE "${GIT_DIR}/HEAD")
+        else()
+            # GIT_DIR/CMAKE_CURRENT_SOURCE_DIR is in a worktree
+            file(READ ${GIT_DIR} worktree_ref)
+            # The .git directory contains a path to the worktree information directory
+            # inside the parent git repo of the worktree.
+            #
+            string(REGEX REPLACE "gitdir: (.*)$" "\\1" git_worktree_dir
+                                 ${worktree_ref})
+            string(STRIP ${git_worktree_dir} git_worktree_dir)
+            _get_git_dir("${git_worktree_dir}" GIT_DIR)
+            set(HEAD_SOURCE_FILE "${git_worktree_dir}/HEAD")
+        endif()
+	endif()
+
+	# Fail if the 'head' file was not found
+    if(NOT EXISTS "${HEAD_SOURCE_FILE}")
+        return()
+    endif()
+
+	# Make a copy of the head file
+    set(HEAD_FILE "${_data_dir}/HEAD")
+    configure_file("${HEAD_SOURCE_FILE}" "${HEAD_FILE}" COPYONLY)
+
+	# Now we create a cmake file that will read the contents of this
+	# head file in the appropriate way
+	file(WRITE "${_data_dir}/grab-ref.cmake.in" [[
+set(HEAD_HASH)
+
+file(READ "@HEAD_FILE@" HEAD_CONTENTS LIMIT 1024)
+
+string(STRIP "${HEAD_CONTENTS}" HEAD_CONTENTS)
+if(HEAD_CONTENTS MATCHES "ref")
+	# named branch
+	string(REPLACE "ref: " "" HEAD_REF "${HEAD_CONTENTS}")
+	if(EXISTS "@GIT_DIR@/${HEAD_REF}")
+		configure_file("@GIT_DIR@/${HEAD_REF}" "@VERSION_STRING_DATA@/head-ref" COPYONLY)
+	else()
+		configure_file("@GIT_DIR@/packed-refs" "@VERSION_STRING_DATA@/packed-refs" COPYONLY)
+		file(READ "@VERSION_STRING_DATA@/packed-refs" PACKED_REFS)
+		if(${PACKED_REFS} MATCHES "([0-9a-z]*) ${HEAD_REF}")
+			set(HEAD_HASH "${CMAKE_MATCH_1}")
+		endif()
+	endif()
+else()
+	# detached HEAD
+	configure_file("@GIT_DIR@/HEAD" "@VERSION_STRING_DATA@/head-ref" COPYONLY)
+endif()
+
+if(NOT HEAD_HASH)
+	file(READ "@VERSION_STRING_DATA@/head-ref" HEAD_HASH LIMIT 1024)
+	string(STRIP "${HEAD_HASH}" HEAD_HASH)
+endif()
+]])
+
+    configure_file("${VERSION_STRING_DATA}/grab-ref.cmake.in"
+                   "${VERSION_STRING_DATA}/grab-ref.cmake" @ONLY)
+    
+	# Include the aforementioned file, this will define
+	# the HEAD_HASH variable we're looking for
+	include("${VERSION_STRING_DATA}/grab-ref.cmake")
+
+    set(${_variable} "${HEAD_HASH}" PARENT_SCOPE)
+endfunction()
+
 # Create a revision file, containing the current git version info, if any
 function(write_version_header dir)
+
+	set(flags )
+	set(options LIB_NAME FILE_NAME)
+	set(sources )
+	cmake_parse_arguments(VERSION_STRING_OPTION "${flags}" "${options}" "${sources}" ${ARGN})
+
 	# parameter check
 	if(NOT IS_DIRECTORY ${dir})
 		message(FATAL_ERROR "First parameter to write_version_header should be a directory where the final revision.hpp file will be placed")
 	endif()
 
-	include(GetGitRevisionDescription)
-	if(NOT(GIT-NOTFOUND OR HEAD-HASH-NOTFOUND))
-		git_describe_working_tree(BUILD_VERSION_STRING --match=build --dirty)
+	if(VERSION_STRING_OPTION_FILE_NAME)
+		set(file_name "${VERSION_STRING_OPTION_FILE_NAME}")
+	else()
+		set(file_name "revision.hpp")
+	endif()
 
-		if(BUILD_VERSION_STRING MATCHES "build-([0-9]+)-g([0-9a-f]+)(-dirty)?")
-			set(BUILD_GIT_TAGREF "${CMAKE_MATCH_2}")
-			if(CMAKE_MATCH_3)
-				set(BUILD_VERSION_STRING "${CMAKE_MATCH_1}*")
-			else()
-				set(BUILD_VERSION_STRING "${CMAKE_MATCH_1}")
-			endif()
+	# Where to store intermediate files
+	set(VERSION_STRING_DATA "${CMAKE_CURRENT_BINARY_DIR}/CMakeFiles/VersionString")
+	if(NOT EXISTS "${VERSION_STRING_DATA}")
+        file(MAKE_DIRECTORY "${VERSION_STRING_DATA}")
+    endif()
+
+	# Load the git hash using the wizzard-like code above.
+	_get_git_hash("${VERSION_STRING_DATA}" GIT_HASH)
+
+	# If git was found, fetch the git description string
+	if(GIT_HASH)
+		execute_process(
+			COMMAND "${GIT_EXECUTABLE}" describe --dirty --match=build
+			WORKING_DIRECTORY "${CMAKE_CURRENT_SOURCE_DIR}"
+			RESULT_VARIABLE res
+			OUTPUT_VARIABLE out
+			ERROR_QUIET OUTPUT_STRIP_TRAILING_WHITESPACE)
+
+		if(res EQUAL 0)
+			set(REVISION_STRING "${out}")
+		else()
+			message(STATUS "Git hash not found, does this project has a 'build' tag?")
 		endif()
 	else()
-		message(WARNING "no git info available, cannot update version string")
+		message(STATUS "Git hash not found")
 	endif()
 
-	string(TIMESTAMP BUILD_DATE_TIME "%Y-%m-%dT%H:%M:%SZ" UTC)
+	# Check the revision string, if it matches we fill in the required info
+	if(REVISION_STRING MATCHES "build-([0-9]+)-g([0-9a-f]+)(-dirty)?")
+		set(BUILD_NUMBER ${CMAKE_MATCH_1})
+		if(CMAKE_MATCH_3)
+			set(REVISION_GIT_TAGREF "${CMAKE_MATCH_2}*")
+		else()
+			set(REVISION_GIT_TAGREF "${CMAKE_MATCH_2}")
+		endif()
 
-	if(ARGC GREATER 1)
-		set(VAR_PREFIX "${ARGV1}")
+		string(TIMESTAMP REVISION_DATE_TIME "%Y-%m-%dT%H:%M:%SZ" UTC)
+	else()
+		set(REVISION_GIT_TAGREF "")
+		set(BUILD_NUMBER 0)
+		set(REVISION_DATE_TIME "")
 	endif()
 
-	file(WRITE "${PROJECT_BINARY_DIR}/revision.hpp.in" [[// Generated revision file
+	if(VERSION_STRING_OPTION_LIB_NAME)
+		set(VAR_PREFIX "${VERSION_STRING_OPTION_LIB_NAME}")
+		set(IDENT_PREFIX "${VERSION_STRING_OPTION_LIB_NAME}_")
+		set(BOOL_IS_MAIN "false")
+	else()
+		set(VAR_PREFIX "")
+		set(IDENT_PREFIX "")
+		set(BOOL_IS_MAIN "true")
+	endif()
 
-#pragma once
-
-#include <ostream>
-
-const char k@VAR_PREFIX@ProjectName[] = "@PROJECT_NAME@";
-const char k@VAR_PREFIX@VersionNumber[] = "@PROJECT_VERSION@";
-const char k@VAR_PREFIX@VersionGitTag[] = "@BUILD_GIT_TAGREF@";
-const char k@VAR_PREFIX@BuildInfo[] = "@BUILD_VERSION_STRING@";
-const char k@VAR_PREFIX@BuildDate[] = "@BUILD_DATE_TIME@";
-
-inline void write_version_string(std::ostream &os, bool verbose)
-{
-	os << k@VAR_PREFIX@ProjectName << " version " << k@VAR_PREFIX@VersionNumber << std::endl;
-	if (verbose)
-	{
-		os << "build: " << k@VAR_PREFIX@BuildInfo << ' ' << k@VAR_PREFIX@BuildDate << std::endl;
-		if (k@VAR_PREFIX@VersionGitTag[0] != 0)
-			os << "git tag: " << k@VAR_PREFIX@VersionGitTag << std::endl;
-	}
-}
-]])
-	configure_file("${PROJECT_BINARY_DIR}/revision.hpp.in" "${dir}/revision.hpp" @ONLY)
+	configure_file("${_current_cmake_module_dir}/revision.hpp.in" "${dir}/${file_name}" @ONLY)
 endfunction()
 

cmake/cifpp-config.cmake.in

@@ -1,17 +1,12 @@
 @PACKAGE_INIT@
 
-include(CMakeFindDependencyMacro)
-find_dependency(Threads)
-
-find_dependency(ZLIB REQUIRED)
-find_dependency(Eigen3 REQUIRED)
-
-if(MSVC)
-	find_dependency(zeep REQUIRED)
-endif()
-
-INCLUDE("${CMAKE_CURRENT_LIST_DIR}/cifppTargets.cmake")
+include("${CMAKE_CURRENT_LIST_DIR}/cifpp-targets.cmake")
 
 set_and_check(CIFPP_SHARE_DIR "@PACKAGE_CIFPP_DATA_DIR@")
 
+include(CMakeFindDependencyMacro)
+
+find_dependency(Threads)
+find_dependency(ZLIB REQUIRED)
+
 check_required_components(cifpp)

cmake/revision.hpp.in

@@ -0,0 +1,121 @@
+// This file was generated by VersionString.cmake
+
+#pragma once
+
+#include <ostream>
+
+constexpr const char k@VAR_PREFIX@ProjectName[] = "@PROJECT_NAME@";
+constexpr const char k@VAR_PREFIX@VersionNumber[] = "@PROJECT_VERSION@";
+constexpr int k@VAR_PREFIX@BuildNumber = @BUILD_NUMBER@;
+constexpr const char k@VAR_PREFIX@RevisionGitTag[] = "@REVISION_GIT_TAGREF@";
+constexpr const char k@VAR_PREFIX@RevisionDate[] = "@REVISION_DATE_TIME@";
+
+#ifndef VERSION_INFO_DEFINED
+#define VERSION_INFO_DEFINED 1
+
+namespace version_info_v1_1
+{
+
+class version_info_base
+{
+  public:
+	static void write_version_string(std::ostream &os, bool verbose)
+	{
+		auto s_main = registered_main();
+		if (s_main != nullptr)
+			s_main->write(os, verbose);
+
+		if (verbose)
+		{
+			for (auto lib = registered_libraries(); lib != nullptr; lib = lib->m_next)
+			{
+				os << "-\n";
+				lib->write(os, verbose);
+			}
+		}
+	}
+
+  protected:
+	version_info_base(const char *name, const char *version, int build_number, const char *git_tag, const char *revision_date, bool is_main)
+		: m_name(name)
+		, m_version(version)
+		, m_build_number(build_number)
+		, m_git_tag(git_tag)
+		, m_revision_date(revision_date)
+	{
+		if (is_main)
+			registered_main() = this;
+		else
+		{
+			auto &s_head = registered_libraries();
+			m_next = s_head;
+			s_head = this;
+		}
+	}
+
+	void write(std::ostream &os, bool verbose)
+	{
+		os << m_name << " version " << m_version << '\n';
+
+		if (verbose)
+		{
+			if (m_build_number != 0)
+			{
+				os << "build: " << m_build_number << ' ' << m_revision_date << '\n';
+				if (m_git_tag[0] != 0)
+					os << "git tag: " << m_git_tag << '\n';
+			}
+		}
+	}
+
+	using version_info_ptr = version_info_base *;
+
+	static version_info_ptr &registered_main()
+	{
+		static version_info_ptr s_main = nullptr;
+		return s_main;
+	}
+
+	static version_info_ptr &registered_libraries()
+	{
+		static version_info_ptr s_head = nullptr;
+		return s_head;
+	}
+
+	const char *m_name;
+	const char *m_version;
+	int m_build_number;
+	const char *m_git_tag;
+	const char *m_revision_date;
+	version_info_base *m_next = nullptr;
+};
+
+template <typename T>
+class version_info : public version_info_base
+{
+  public:
+	using implementation_type = T;
+
+	version_info(const char *name, const char *version, int build_number, const char *git_tag, const char *revision_date, bool is_main)
+		: version_info_base(name, version, build_number, git_tag, revision_date, is_main)
+	{
+	}
+};
+
+} // namespace version_info_v1_1
+
+inline void write_version_string(std::ostream &os, bool verbose)
+{
+	version_info_v1_1::version_info_base::write_version_string(os, verbose);
+}
+
+#endif
+
+const class version_info_@IDENT_PREFIX@impl : public version_info_v1_1::version_info<version_info_@IDENT_PREFIX@impl>
+{
+  public:
+	version_info_@IDENT_PREFIX@impl()
+		: version_info(k@VAR_PREFIX@ProjectName, k@VAR_PREFIX@VersionNumber, k@VAR_PREFIX@BuildNumber, k@VAR_PREFIX@RevisionGitTag, k@VAR_PREFIX@RevisionDate, @BOOL_IS_MAIN@)
+	{
+	}
+} s_version_info_@IDENT_PREFIX@instance;

cmake/test-rx.cpp

@@ -11,8 +11,8 @@ int main()
 
 	std::regex_search(s, m, r);
 
-	std::cout << s.substr(0, 10) << std::endl;
-	std::cout << m.str(1).substr(0, 10) << std::endl;
+	std::cout << s.substr(0, 10) << '\n';
+	std::cout << m.str(1).substr(0, 10) << '\n';
 
 	return 0;
 }

docs/CMakeLists.txt

@@ -0,0 +1,48 @@
+find_package(Doxygen REQUIRED)
+find_package(Sphinx REQUIRED)
+
+# Find all the public headers
+# get_target_property(CIFPP_PUBLIC_HEADER_DIR libCIFPP INTERFACE_INCLUDE_DIRECTORIES)
+set(CIFPP_PUBLIC_HEADER_DIR ${PROJECT_SOURCE_DIR}/include)
+file(GLOB_RECURSE CIFPP_PUBLIC_HEADERS ${CIFPP_PUBLIC_HEADER_DIR}/*.hpp)
+
+set(DOXYGEN_INPUT_DIR ${CIFPP_PUBLIC_HEADER_DIR})
+set(DOXYGEN_OUTPUT_DIR ${CMAKE_CURRENT_BINARY_DIR}/xml)
+set(DOXYGEN_INDEX_FILE ${DOXYGEN_OUTPUT_DIR}/index.xml)
+set(DOXYFILE_IN ${CMAKE_CURRENT_SOURCE_DIR}/Doxyfile.in)
+set(DOXYFILE_OUT ${CMAKE_CURRENT_BINARY_DIR}/Doxyfile)
+
+# Replace variables inside @@ with the current values
+configure_file(${DOXYFILE_IN} ${DOXYFILE_OUT} @ONLY)
+
+add_custom_command(
+	OUTPUT ${DOXYGEN_OUTPUT_DIR}
+	COMMAND ${CMAKE_COMMAND} -E make_directory ${DOXYGEN_OUTPUT_DIR})
+
+add_custom_command(OUTPUT ${DOXYGEN_INDEX_FILE}
+	BYPRODUCTS ${DOXYGEN_OUTPUT_DIR}
+	DEPENDS ${DOXYGEN_OUTPUT_DIR} ${CIFPP_PUBLIC_HEADERS} ${DOXYFILE_OUT}
+	COMMAND ${DOXYGEN_EXECUTABLE} ${DOXYFILE_OUT}
+	MAIN_DEPENDENCY ${DOXYFILE_OUT} ${DOXYFILE_IN}
+	COMMENT "Generating docs")
+
+add_custom_target("Doxygen-${PROJECT_NAME}" ALL DEPENDS ${DOXYGEN_INDEX_FILE})
+
+configure_file(${CMAKE_CURRENT_SOURCE_DIR}/conf.py.in ${CMAKE_CURRENT_SOURCE_DIR}/conf.py @ONLY)
+
+set(SPHINX_SOURCE ${CMAKE_CURRENT_SOURCE_DIR})
+set(SPHINX_BUILD ${CMAKE_CURRENT_BINARY_DIR}/sphinx)
+
+add_custom_target("Sphinx-${PROJECT_NAME}" ALL
+	COMMAND ${SPHINX_EXECUTABLE} -b html
+	-Dbreathe_projects.${PROJECT_NAME}=${DOXYGEN_OUTPUT_DIR}
+	${SPHINX_SOURCE} ${SPHINX_BUILD}
+	DEPENDS ${DOXYGEN_INDEX_FILE}
+	BYPRODUCTS ${CMAKE_CURRENT_SOURCE_DIR}/api
+	WORKING_DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}
+	COMMENT "Generating documentation with Sphinx")
+
+install(DIRECTORY ${CMAKE_CURRENT_BINARY_DIR}/sphinx/
+	DESTINATION ${CMAKE_INSTALL_DOCDIR}
+	PATTERN .doctrees EXCLUDE
+	PATTERN .buildinfo EXCLUDE)

docs/Doxyfile.in

@@ -0,0 +1,10 @@
+EXCLUDE_SYMBOLS        = cif::detail::*, std*
+FILE_PATTERNS          = *.hpp
+STRIP_FROM_PATH        = @DOXYGEN_INPUT_DIR@
+RECURSIVE              = YES
+GENERATE_XML           = YES
+GENERATE_LATEX         = NO
+PREDEFINED             += and=&& or=|| not=! CIFPP_EXPORT= HAVE_LIBCLIPPER=1
+GENERATE_HTML          = NO
+GENERATE_TODOLIST      = NO
+INPUT                  = @DOXYGEN_INPUT_DIR@

docs/_static/.gitignore

@@ -0,0 +1,4 @@
+# Ignore everything in this directory
+*
+# Except this file
+!.gitignore

docs/basics.rst

@@ -0,0 +1,400 @@
+Basic usage
+===========
+
+This library, *libcifpp*, is a generic *CIF* library with some specific additions to work with *mmCIF* files. The main focus of this library is to make sure that files read or written are valid. That is, they are syntactically valid *and* their content is valid with respect to a CIF dictionary, if such a dictionary is available and specified.
+
+Reading a file is as simple as:
+
+.. code-block:: cpp
+
+    #include <cif++.hpp>
+
+    cif::file f("/path/to/file.cif");
+
+The file may also be compressed using *gzip* which is detected automatically.
+
+Writing out the file again is also simple, to write out the terminal you can do:
+
+.. code-block:: cpp
+
+    std::cout << f;
+
+    // or
+    f.save(std::cout);
+
+    // or write a compressed file using gzip compression:
+    f.save("/tmp/f.cif.gz");
+
+CIF files contain one or more datablocks. To print out the names of all datablocks in our file:
+
+.. code-block:: cpp
+
+    for (auto &db : f)
+        std::cout << db.name() << '\n';
+
+Most often *libcifpp* is used to read in structure files in mmCIF format. These files only contain one datablock and so you can safely use code like this:
+
+.. code-block:: cpp
+
+    // get a reference to the first datablock in f
+    auto &db = f.front();
+
+But if you know the name of the datablock, this also works:
+
+.. code-block:: cpp
+
+    // get a reference to the datablock name '1CBS'
+    auto &db = f["1CBS"];
+
+Now, each datablock contains categories. To print out all their names:
+
+.. code-block:: cpp
+
+    for (auto &cat : db)
+        std::cout << cat.name() << '\n';
+
+But you probably know what category you need to use, so lets fetch it by name:
+
+.. _atom_site-label:
+.. code-block:: cpp
+
+    // get a reference to the atom_site category in db
+    auto &atom_site = db["atom_site"];
+
+    // and make sure there's some data in it:
+    assert(not atom_site.empty());
+
+.. note::
+    
+    Note that we omit the leading underscore in the name of the category here.
+
+Categories contain rows of data and each row has fields or items. Referencing a row in a category results in a :cpp:class:`cif::row_handle` object which you can use to request or manipulate item data.
+
+.. code-block:: cpp
+
+    // Get the first row in atom_site
+    auto rh = atom_site.front();
+
+    // Get the label_atom_id value from this row handle as a std::string
+    std::string atom_id = rh["label_atom_id"].as<std::string>();
+
+    // Get the x, y and z coordinates using structered binding
+    const auto &[x, y, z] = rh.get<float,float,float>("Cartn_x", "Cartn_y", "Cartn_z");
+
+    // Assign a new value to the x coordinate or our atom
+    rh["Cartn_x"] = x + 1;
+
+Querying
+--------
+
+Walking over the rows in a category is often not very useful. More often you are interested in specific rows in a category. The function :cpp:func:`cif::category::find` and friends are here to help.
+
+What these functions have in common is that they return data based on a query implemented by :cpp:class:`cif::condition`. These condition objects are built in code using regular C++ syntax. The most basic example of a query is:
+
+.. code-block:: cpp
+
+    cif::condition c = cif::key("id") == 1;
+
+Here the condition is that all rows returned should have a value of 1 in there item named *id*. Likewise you can use other data types and even combine those. Oh, and I said we use regular C++ syntax for conditions, so you may as well use other operators to compare values:
+
+.. code-block:: cpp
+
+    // condition for C-alpha atoms having an occupancy less than 1.0
+    cif::condition c = cif::key("occupancy") < 1.0f and cif::key("label_atom_id") == "CA";
+
+Using the namespace *cif::literals* that code becomes a little less verbose:
+
+.. code-block:: cpp
+
+    using namespace cif::literals;
+    cif::condition c = "occupancy"_key < 1.0f and "label_atom_id"_key == "CA";
+
+Conditions can also be combined:
+
+.. code-block:: cpp
+
+    cif::condition c = "occupancy"_key < 1.0f and "label_atom_id"_key == "CA";
+
+    // extend the condition by requiring the compound ID to be unequal to PRO
+    c = std::move(c) and "label_comp_id"_key != "PRO";
+
+.. note::
+
+    Note the use of std::move here. 
+
+Using queries constructed in this way is simple:
+
+.. code-block:: cpp
+
+    cif::condition c = ...
+    auto result = atom_site.find(std::move(c));
+
+    // or construct a condition inline:
+    auto result = atom_site.find("label_atom_id"_key == "CA");
+
+In the example above the result is a range of :cpp:class:`cif::row_handle` objects. Often, using individual field values is more useful:
+
+.. code-block:: cpp
+
+    // Requesting a single item:
+    for (auto id : atom_site.find<std::string>("label_atom_id"_key == "CA", "id"))
+        std::cout << "ID for CA: " << id << '\n';
+
+    // Requesting multiple items:
+    for (const auto &[id, x, y, z] : atom_site.find<std::string,float,float,float>("label_atom_id"_key == "CA",
+            "id", "Cartn_x", "Cartn_y", "Cartn_z"))
+    {
+        std::cout << "Atom " << id << " is at [" << x << ", " << y << ", " z << "]\n";
+    }
+
+Returning a complete set if often not required, if you only want to have the first you can use :cpp:func:`cif::category::find_first` as shown here:
+
+.. code-block:: cpp
+
+    // return the ID item for the first C-alpha atom
+    std::string v1 = atom_site.find_first<std::string>("label_atom_id"_key == "CA", "id");
+
+    // If you're not sure the row exists, use std::optional
+    auto v2 = atom_site.find_first<std::optional<std::string>>("label_atom_id"_key == "CA", "id");
+    if (v2.has_value())
+        ...
+
+There are cases when you really need exactly one result. The :cpp:func:`cif::category::find1` can be used in that case, it will throw an exception if the query does not result in exactly one row.
+
+NULL and ANY
+------------
+
+Sometimes items may be empty. The trouble is a bit that empty comes in two flavors: unknown and null. Null in *CIF* parlance means the item should not contain a value since it makes no sense in this case, the value stored in the file is a single dot character: ``'.'``. E.g. *atom_site* records may have a NULL value for label_seq_id for atoms that are part of a *non-polymer*.
+
+The other empty value is indicated by a question mark character: ``'?'``. This means the value is simply unknown.
+
+Both these are NULL in *libcifpp* conditions and can be searched for using :cpp:var:`cif::null`.
+
+So you can search for:
+
+.. code-block:: cpp
+
+    cif::condition c = "label_seq_id"_key == cif::null;
+
+You might also want to look for a certain value and don't care in which item it is stored, in that case you can use :cpp:var:`cif::any`.
+
+.. code-block:: cpp
+
+    cif::condition c = cif::any == "foo";
+
+And in linked record you might have the items that have a value in both parent and child or both should be NULL. For that, you can request the value to return by find to be of type std::optional and then use that value to build the query. An example to explain this, let's find the location of the atom that is referenced as the first atom in a struct_conn record:
+
+.. code-block:: cpp
+
+    // Take references to the two categories we need
+    auto struct_conn = db["struct_conn"];
+    auto atom_site = db["atom_site"];
+
+    // Loop over all rows in struct_conn taking only the values we need
+    // Note that the label_seq_id is returned as a std::optional<int>
+    // That means it may contain an integer or may be empty
+    for (const auto &[asym1, seqid1, authseqid1, atomid1] :
+        struct_conn.rows<std::string,std::optional<int>,std::string,std::string,std::string>(
+            "ptnr1_label_asym_id", "ptnr1_label_seq_id", "ptnr1_auth_seq_id", "ptnr1_label_atom_id"
+        ))
+    {
+        // Find the location of the first atom
+        cif::point p1 = atom_site.find1<float,float,float>(
+            "label_asym_id"_key == asym1 and "label_seq_id"_key == seqid1 and "auth_seq_id"_key == authseqid1 and "label_atom_id"_key == atomid1,
+            "cartn_x", "cartn_y", "cartn_z");
+    }
+    
+
+Validation
+----------
+
+CIF files can have a dictionary attached. And based on such a dictionary a :cpp:class:`cif::validator` object can be constructed which in turn can be used to validate the content of the file.
+
+A simple case:
+
+.. code-block:: cpp
+
+    #include <cif++.hpp>
+
+    cif::file f("1cbs.cif.gz");
+    f.load_dictionary("mmcif_pdbx");
+
+    if (not f.is_valid())
+        std::cout << "This file is not valid\n";
+
+If you want to know why it is not valid, you should set the global variable :cpp:var:`cif::VERBOSE` to something higer than zero. Depending on the value more or less diagnostic output is sent to std::cerr.
+
+In the case above we load a dictionary based on its name. You can of course also load dictionaries based on a specific file, that's a bit more work:
+
+.. code-block:: cpp
+
+    std::filesystem::ifstream dictFile("/tmp/my-dictionary.dic");
+    auto &validator = cif::parse_dictionary("my-dictionary", dictFile);
+
+    cif::file f("1cbs.cif.gz");
+
+    // assign the validator
+    f.set_validator(&validator);
+
+    // alternatively, load it by name
+    f.load_dictionary("my-dictionary");
+
+    if (not f.is_valid())
+        std::cout << "This file is not valid\n";
+
+Creating your own dictionary is a lot of work, especially if you are only extending an existing dictionary with a couple of new categories or items. So, what you can do is extend a loaded validator like this (code taken from DSSP):
+
+.. code-block:: cpp
+
+    // db is a cif::datablock reference containing an mmCIF file with DSSP annotations
+    auto &validator = const_cast<cif::validator &>(*db.get_validator());
+    if (validator.get_validator_for_category("dssp_struct_summary") == nullptr)
+    {
+        auto dssp_extension = cif::load_resource("dssp-extension.dic");
+        if (dssp_extension)
+            cif::extend_dictionary(validator, *dssp_extension);
+    }
+
+.. note::
+
+    In the example above we're loading the data using :doc:`/resources`. See the documentation on that for more information.
+
+If a validator has been assigned to a file, assignments to items are checked for valid data. So the following code will throw an exception (see: :ref:`_atom_site-label`):
+
+.. code-block:: cpp
+    
+    auto rh = atom_site.front();
+    rh["Cartn_x"] = "foo";
+
+Linking
+-------
+
+Based on information recorded in dictionary files (see :ref:`Validation`) you can locate linked records in parent or child categories.
+
+To make this example not too complex, lets assume the following example file:
+
+.. code-block:: cif
+
+    data_test
+    loop_
+    _cat_1.id
+    _cat_1.name
+    _cat_1.desc
+    1 aap  Aap
+    2 noot Noot
+    3 mies Mies
+
+    loop_
+    _cat_2.id
+    _cat_2.name
+    _cat_2.num
+    _cat_2.desc
+    1 aap  1 'Een dier'
+    2 aap  2 'Een andere aap'
+    3 noot 1 'walnoot bijvoorbeeld'
+
+And we have a dictionary containing the following link definition:
+
+.. code-block:: cif
+
+    loop_
+    _pdbx_item_linked_group_list.parent_category_id
+    _pdbx_item_linked_group_list.link_group_id
+    _pdbx_item_linked_group_list.parent_name
+    _pdbx_item_linked_group_list.child_name
+    _pdbx_item_linked_group_list.child_category_id
+    cat_1 1 '_cat_1.name' '_cat_2.name' cat_2
+
+So, there are links between *cat_1* and *cat_2* based on the value in items named *name*. Using this information, we can now locate children and parents:
+
+.. code-block:: cpp
+
+    // Assuming the file was loaded in f:
+    auto &cat1 = f.front()["cat_1"];
+    auto &cat2 = f.front()["cat_2"];
+    auto &cat3 = f.front()["cat_3"];
+
+    // Loop over all ape's in cat2
+    for (auto r : cat1.get_children(cat1.find1("name"_key == "aap"), cat2))
+        std::cout << r.get<std::string>("desc") << '\n';
+
+Updating a value in an item in a parent category will update the corresponding value in all related children:
+
+.. code-block:: cpp
+
+    auto r1 = cat1.find1("id"_key == 1);
+    r1["name"] = "aapje";
+
+    auto rs1 = cat2.find("name"_key == "aapje");
+    assert(rs1.size() == 2);
+
+However, changing a value in a child record will not update the parent. This may result in an invalid file since you may then have a child that has no parent:
+
+.. code-block:: cpp
+
+    auto r2 = cat2.find1("id"_key == 3);
+    r2["name"] = "wim";
+
+    assert(f.is_valid() == false);
+
+So you have to fix this yourself by inserting a new item in cat1 with the new value.
+
+.. _splitting-rows:
+Another situation is when you change a value in a parent and updating children might introduce a situation where you need to split a child. To give an example, consider this:
+
+.. code-block:: cif
+
+    data_test
+    loop_
+    _cat_1.id
+    _cat_1.name
+    _cat_1.desc
+    1 aap  Aap
+    2 noot Noot
+    3 mies Mies
+
+    loop_
+    _cat_2.id
+    _cat_2.name
+    _cat_2.num
+    _cat_2.desc
+    1 aap  1 'Een dier'
+    2 aap  2 'Een andere aap'
+    3 noot 1 'walnoot bijvoorbeeld'
+
+    loop_
+    _cat_3.id
+    _cat_3.name
+    _cat_3.num
+    1 aap 1
+    2 aap 2
+
+And we have a dictionary containing the following link definition (reversed compared to the previous example):
+
+.. code-block:: cif
+
+    loop_
+    _pdbx_item_linked_group_list.parent_category_id
+    _pdbx_item_linked_group_list.link_group_id
+    _pdbx_item_linked_group_list.parent_name
+    _pdbx_item_linked_group_list.child_name
+    _pdbx_item_linked_group_list.child_category_id
+    cat_2 1 '_cat_2.name' '_cat_1.name' cat_1
+    cat_3 1 '_cat_3.name' '_cat_2.name' cat_2
+    cat_3 1 '_cat_3.num'  '_cat_2.num'  cat_2
+
+So *cat3* is a parent of *cat2* and *cat2* is a parent of *cat1*. Now, if you change the *name* value of the first row of *cat3* to 'aapje', the corresponding row in *cat2* is updated as well. But when you update *cat2* you have to update *cat1* too. And simply changing the name field in row 1 of *cat1* is wrong. The default behaviour in libcifpp is to split the record in *cat1* and have a new child with the new name whereas the other remains as is.
+
+The new *cat1* will thus be like:
+
+.. code-block:: cif
+
+    loop_
+    _cat_1.id
+    _cat_1.name
+    _cat_1.desc
+    1 aapje Aap
+    2 noot  Noot
+    3 mies  Mies
+    5 aap   Aap
+

docs/bitsandpieces.rst

@@ -0,0 +1,49 @@
+Bits & Pieces
+=============
+
+The *libcifpp* library offers some extra code that makes the life of developers a bit easier.
+
+gzio
+----
+
+To work with compressed data files a *std::streambuf* implemenation was added based on the code in `gxrio <https://github.com/mhekkel/gxrio>`_. This allows you to read and write compressed data streams transparently.
+
+When working with files you can use :cpp:class:`cif::gzio::ifstream` and :cpp:class:`cif::gzio::ofstream`. The selection of whether to use compression or not is based on the file extension. If it is ``.gz`` gzip compression is used:
+
+.. code-block:: cpp
+
+	cif::gzio::ifstream file("my-file.txt.gz");
+
+	std::string line;
+	while (std::getline(file, line))
+		std::cout << line << '\n';
+
+Writing is equally easy:
+
+.. code-block:: cpp
+
+	cif::gzio::ofstream file("/tmp/output.txt.gz");
+	file << "Hello, world!";
+	file.close();
+
+You can also use the :cpp:class:`cif::gzio::istream` and feed it a *std::streambuf* object that may or may not contain compressed data. In that case the first bytes of the input are sniffed and if it is gzip compressed data, decompression will be done.
+
+A progress bar
+--------------
+
+Applications based on *libcifpp* may have a longer run time. To give some feedback to the user running your application in a terminal you can use the :cpp:class:`cif::progress_bar`. This class will display an ASCII progress bar along with optional status messages, but only if output is to a real TTY (terminal).
+
+A progress bar is also shown only if the duration is more than two seconds. To avoid having flashing progress bars for short actions.
+
+The progress bar uses an internal progress counter that starts at zero and ends when the max value has been reached after which it will be removed from the screen. Updating this internal progress counter can be done by adding a number of steps calling :cpp:func:`cif::progress_bar::consumed` or by setting the exact value for the counter by calling :cpp:func:`cif::progress_bar::progress`.
+
+Colouring output
+----------------
+
+It is also nice to emphasise some output in the terminal by using colours. For this you can create output manipulators using :cpp:func:`cif::coloured`. To write a string in white, and bold letters on a red background you can do:
+
+.. code-block:: cpp
+
+	using namespace cif::colour;
+	std::cout << cif::coloured("Hello, world!", white, red, bold) << '\n';
+

docs/compound.rst

@@ -0,0 +1,33 @@
+Chemical Compounds
+==================
+
+The data in *CIF* and *mmCIF* files often describes the structure of some chemical compounds. The structure is recorded in the categories *atom_site* and friends. Records in these categories refer to chemical compounds using a compound ID. This compound ID is the ID field of the *chem_comp* category. For all of the known compounds in the PDB there is an entry in the Chemical Compounds Dictionary or `CCD <https://www.wwpdb.org/data/ccd>`_. If *libcifpp* was properly installed you have a copy of this file somewhere on your disk. And if you have installed the update scripts, a fresh version of this file will be retrieved weekly.
+
+As an alternative to CCD there are the monomer library files from `CCP4 <https://www.ccp4.ac.uk/>`_. These contain somewhat different data but the overlap is good enough for usage in *libcifpp*.
+
+Information about compounds is captured in the :cpp:class:`cif::compound`. An instance of a compound object for a certain compound ID can be obtained by using the singleton :cpp:class:`cif::compound_factory`.
+
+If the compound you want to use is not available in the CCD or in CCP4, you can add that information yourself. For this you can use the method :cpp:func:`cif::compound_factory::push_dictionary`.
+
+So, given that we have CCD, CCP4 monomer library and used defined compound definitions, what will you get when you try to retrieve such a compound by ID? The answer is, the factory has a stack of compound generators. The first thrown on the stack is the one for a CCD file (*components.cif*) if it can be found. Then, if the *CLIBD_MON* environmental variable is defined, a generator for monomer library files is added to the stack. And then all generators for files you added using *push_dictionary* are added in order. The generators are searched in the reverse order in which they were added to see if it creates a compound object for the ID. If no compound was created at all, nullptr is returned.
+
+Updating CCD
+------------
+
+The CCD data is stored in a single file called *components.cif* and can be downloaded from `CCD <https://www.wwpdb.org/data/ccd>`_. 
+
+As can be read in the section on resources (:doc:`/resources`) files in libcifpp are loaded in a specific order. If the CCD datafile was downloaded during installation, a copy can be found in the directory */usr/share/libcifpp/* (if you installed in */usr*). This is a static file and will not be updated until the next installation of libcifpp.
+
+When configuring libcifpp, you can specify the *CIFPP_INSTALL_UPDATE_SCRIPT* option, as in:
+
+.. code-block:: console
+
+	cmake -S . -B build -DCIFPP_INSTALL_UPDATE_SCRIPT=ON # ... more options?
+
+This will install a script named *update-libcifpp-data* in */etc/cron.weekly* or */etc/periodic/weekly*. This file uses a config file named */etc/libcifpp.conf* which you then need to edit. In this config file the following line needs to be uncommented:
+
+.. code-block:: console
+
+	# update=true
+
+After that, the update script will weekly download the latest components.cif file to */var/cache/libcifpp*.

docs/conf.py.in

@@ -0,0 +1,66 @@
+project = '@PROJECT_NAME@'
+copyright = '2023, Maarten L. Hekkelman'
+author = 'Maarten L. Hekkelman'
+release = '@PROJECT_VERSION@'
+
+# -- General configuration ---------------------------------------------------
+
+extensions = [
+    "breathe",
+    "exhale",
+    "myst_parser"
+]
+
+breathe_projects = {
+	"@PROJECT_NAME@": "../build/docs/xml"
+}
+
+myst_enable_extensions = [ "colon_fence" ]
+breathe_default_project = "@PROJECT_NAME@"
+
+# Setup the exhale extension
+exhale_args = {
+    # These arguments are required
+    "containmentFolder":     "./api",
+    "rootFileName":          "library_root.rst",
+    "doxygenStripFromPath":  "../include/",
+    # Heavily encouraged optional argument (see docs)
+    "rootFileTitle":         "API Reference",
+    # Suggested optional arguments
+    # "createTreeView":        True,
+    # TIP: if using the sphinx-bootstrap-theme, you need
+    # "treeViewIsBootstrap": True,
+    "exhaleExecutesDoxygen": False,
+    "contentsDirectives" : False,
+    
+    "verboseBuild": False
+}
+
+# Tell sphinx what the primary language being documented is.
+primary_domain = 'cpp'
+
+# Tell sphinx what the pygments highlight language should be.
+highlight_language = 'cpp'
+
+templates_path = ['_templates']
+exclude_patterns = ['_build', 'Thumbs.db', '.DS_Store']
+
+# -- Options for HTML output -------------------------------------------------
+
+# The theme to use for HTML and HTML Help pages.  See the documentation for
+# a list of builtin themes.
+#
+html_theme = 'sphinx_rtd_theme'
+
+# Add any paths that contain custom static files (such as style sheets) here,
+# relative to this directory. They are copied after the builtin static files,
+# so a file named "default.css" will overwrite the builtin "default.css".
+html_static_path = ['_static']
+
+html_theme_options = {
+}
+
+cpp_index_common_prefix = [
+	'cif::'
+]
+

docs/genindex.rst

@@ -0,0 +1,2 @@
+Index
+=====

docs/index.rst

@@ -0,0 +1,46 @@
+Introduction
+============
+
+Information on 3D structures of proteins originally came formatted in `PDB <http://www.wwpdb.org/documentation/file-format-content/format33/v3.3.html>`_ files. Although the specification for this format had some real restrictions like a mandatory HEADER and CRYST line, many programs implemented this very poorly often writing out only ATOM records. And users became used to this.
+
+The legacy PDB format has some severe limitations rendering it useless for all but very small protein structures. A new format called `mmCIF <https://mmcif.wwpdb.org/>`_ has been around for decades and now is the default format for the Protein Data Bank.
+
+The software developed in the `PDB-REDO <https://pdb-redo.eu/>`_ project aims at improving 3D models based on original experimental data. For this, the tools need to be able to work with both legacy PDB and mmCIF files. A decision was made to make mmCIF leading internally in all programs and convert legacy PDB directly into mmCIF before processing the data. A robust conversion had to be developed to make this possible since, as noted above, files can come with more or less information making it sometimes needed to do a sequence alignment to find out the exact residue numbers.
+
+And so libcif++ came to life, a library to work with mmCIF files. Work on this library started early 2017 and has developed quite a bit since then. To reduce dependency on other libraries, some functionality was added that is not strictly related to reading and writing mmCIF files but may be useful nonetheless. This is mostly code that is used in 3D calculations and symmetry operations.
+
+Design
+------
+
+The main part of the library is a set of classes that work with mmCIF files. They are:
+
+* :cpp:class:`cif::file`
+* :cpp:class:`cif::datablock`
+* :cpp:class:`cif::category`
+
+The :cpp:class:`cif::file` class encapsulates the contents of a mmCIF file. In such a file there are one or more :cpp:class:`cif::datablock` objects and each datablock contains one or more :cpp:class:`cif::category` objects.
+
+Synopsis
+--------
+
+Using *libcifpp* is easy, if you are familiar with modern C++:
+
+.. literalinclude:: ../README.md
+	:language: c++
+	:start-after: ```cpp
+	:end-before: ```
+
+.. toctree::
+   :maxdepth: 2
+   :caption: Contents
+
+   self
+   basics.rst
+   compound.rst
+   model.rst
+   resources.rst
+   symmetry.rst
+   bitsandpieces.rst
+   api/library_root.rst
+   genindex.rst
+

docs/model.rst

@@ -0,0 +1,36 @@
+Molecular Model
+===============
+
+Theoretically it is possible to get along with only the classes *cif::file*, *cif::datablock* and *cif::category*. But to keep your data complete and valid you then have to update lots of categories for all but the simplest manipulations. For this *libcifpp* comes with a higher level API modelling atoms, residues, monomers, polymers and complete structures in their respective classes.
+
+Note that these classes only work properly if you are using *mmCIF* files and have an mmcif_pdbx dictionary available, either compiled in using `mrc <https://github.com/mhekkel/mrc.git>`_ or installed in the proper location.
+
+.. note::
+
+	This part of *libcifpp* is the least developed part. What is available should work but functionality should eventually be extended.
+
+Atom
+----
+
+The :cpp:class:`cif::mm::atom` is a lightweight proxy class giving access to the data stored in *atom_site* and *atom_site_anisotrop*. It only caches the most often used item data and every modification is directly written back into the *mmCIF* categories.
+
+Atoms can be copied by value with low cost. The atom class only contains a pointer to an implementation that is reference counted.
+
+Residue, Monomer and Polymer
+----------------------------
+
+The :cpp:class:`cif::mm::residue`, :cpp:class:`cif::mm::monomer` and :cpp:class:`cif::mm::polymer` implement what you'd expect. A monomer is a residue that is part of a polymer and thus has a sequence number and siblings.
+
+Sugars & Branches
+-----------------
+
+There are also classes for modelling sugars and sugar branches. You can create sugar branches
+
+Structure
+---------
+
+The :cpp:class:`cif::mm::structure` can be used to load one of the models from an *mmCIF* file. By default the first model is loaded. (Multiple models are often only available files containing structures defined using NMR).
+
+A structure holds a reference to a *cif::datablock* and retrieves its data from this datablock and writes any modification back into that datablock.
+
+One of the most useful parts of the structure class is the ability to create and modify residues. This updates related *chem_comp* and *entity* categories as well.

docs/requirements.in

@@ -0,0 +1,5 @@
+sphinx<5
+exhale==0.3.6
+myst-parser
+breathe
+sphinx_rtd_theme==1.3.0

docs/requirements.txt

@@ -0,0 +1,93 @@
+#
+# This file is autogenerated by pip-compile with Python 3.10
+# by the following command:
+#
+#    pip-compile --output-file=requirements.txt requirements.in
+#
+alabaster==0.7.13
+    # via sphinx
+babel==2.12.1
+    # via sphinx
+beautifulsoup4==4.12.2
+    # via exhale
+breathe==4.35.0
+    # via
+    #   -r requirements.in
+    #   exhale
+certifi==2023.7.22
+    # via requests
+charset-normalizer==3.2.0
+    # via requests
+docutils==0.17.1
+    # via
+    #   breathe
+    #   exhale
+    #   myst-parser
+    #   sphinx
+    #   sphinx-rtd-theme
+exhale==0.3.6
+    # via -r requirements.in
+idna==3.4
+    # via requests
+imagesize==1.4.1
+    # via sphinx
+jinja2==3.1.2
+    # via
+    #   myst-parser
+    #   sphinx
+lxml==4.9.3
+    # via exhale
+markdown-it-py==2.2.0
+    # via
+    #   mdit-py-plugins
+    #   myst-parser
+markupsafe==2.1.3
+    # via jinja2
+mdit-py-plugins==0.3.5
+    # via myst-parser
+mdurl==0.1.2
+    # via markdown-it-py
+myst-parser==0.18.1
+    # via -r requirements.in
+packaging==23.1
+    # via sphinx
+pygments==2.16.1
+    # via sphinx
+pyyaml==6.0.1
+    # via myst-parser
+requests==2.31.0
+    # via sphinx
+six==1.16.0
+    # via exhale
+snowballstemmer==2.2.0
+    # via sphinx
+soupsieve==2.4.1
+    # via beautifulsoup4
+sphinx==4.5.0
+    # via
+    #   -r requirements.in
+    #   breathe
+    #   exhale
+    #   myst-parser
+    #   sphinx-rtd-theme
+    #   sphinxcontrib-jquery
+sphinx-rtd-theme==1.3.0
+    # via -r requirements.in
+sphinxcontrib-applehelp==1.0.4
+    # via sphinx
+sphinxcontrib-devhelp==1.0.2
+    # via sphinx
+sphinxcontrib-htmlhelp==2.0.1
+    # via sphinx
+sphinxcontrib-jquery==4.1
+    # via sphinx-rtd-theme
+sphinxcontrib-jsmath==1.0.1
+    # via sphinx
+sphinxcontrib-qthelp==1.0.3
+    # via sphinx
+sphinxcontrib-serializinghtml==1.1.5
+    # via sphinx
+typing-extensions==4.7.1
+    # via myst-parser
+urllib3==2.0.4
+    # via requests

docs/resources.rst

@@ -0,0 +1,47 @@
+Resources
+=========
+
+Programs using libcifpp often need access to common data files. E.g. CIF dictionary files, CCP4 monomer restraints files or the CCD data file. In libcifpp these files are called resources. These files are often also based on external sources that are updated on a regular basis.
+
+Resources can be compiled into the executable so that the resulting
+application can be made portable to other machines. For this you
+need to use `mrc <https://github.com/mhekkel/mrc.git>`_ which only works
+on Un*x like systems using the ELF executable format or on MS Windows
+
+But resources may also be located as files on the filesytem at
+specific locations. And you can specify your own location for
+files (a directory) or even override named resources with your
+own data.
+
+Loading Resources
+-----------------
+
+No matter where the resource is located, you should always use the single libcifpp API call :cpp:func:`cif::load_resource` to load them. This function returns a *std::istream* wrapped inside a *std::unique_ptr*. 
+
+The order in which resources are searched for is:
+
+* Use the resource that was defined by calling :cpp:func:`cif::add_file_resource`
+  for this name.
+
+* Search the paths specified by :cpp:func:`cif::add_data_directory`, last one
+  added is searched first
+
+* Search the so-called *CACHE_DIR*. This location is defined
+  at compile time and based on the installation directory of
+  libcifpp. Usually it is */var/cache/libcifpp*.
+  It is in this directory where the cron job for libcifpp will
+  put the updated files weekly.
+
+* If the *CCP4* environment is available, the
+  *$ENV{CCP4}/share/libcifpp* is searched.
+
+* If the environment variable *LIBCIFPP_DATA_DIR* is set it
+  is searched
+
+* The *DATA_DIR* is searched, this is also a variable defined
+  at compile time, also based on the installation directory
+  of libcifpp. It usually is */usr/share/libcifpp*
+
+* As a last resort an attempt is made to load the data from
+  resources compiled by `mrc <https://github.com/mhekkel/mrc.git>`_.
+

docs/symmetry.rst

@@ -0,0 +1,108 @@
+Symmetry & Geometry
+===================
+
+Although not really a core *CIF* functionality, when working with *mmCIF* files you often need to work with symmetry information. And symmetry works on points in a certain space and thus geometry calculations are also something you need often. Former versions of *libcifpp* used to use `clipper <http://www.ysbl.york.ac.uk/~cowtan/clipper/doc/index.html>`_ to do many of these calculations, but that introduces a dependency and besides, the way clipper numbers symmetry operations is not completely compatible with the way this is done in the PDB.
+
+Points
+------
+
+The most basic type in use is :cpp:type:`cif::point`. It can be thought of as a point in space with three coordinates, but it is also often used as a vector in 3d space. To keep the interface simple there's no separate vector type.
+
+Many functions are available in :ref:`file_cif++_point.hpp` that work on points. There are functions to calculate the :cpp:func:`cif::distance` between two points and also function to calculate dot products, cross products and dihedral angles between sets of points.
+
+Quaternions
+-----------
+
+All operations inside *libcifpp* that perform some kind of rotation use :cpp:type:`cif::quaternion`. The reason to use Quaternions is not only that they are cool, they are faster than multiplying with a matrix and the results also suffer less from numerical instability.
+
+Matrix
+------
+
+Although Quaternions are the preferred way of doing rotations, not every manipulation is a rotation and thus we need a matrix class as well. Matrices and their operations are encoded as matrix_expressions in *libcifpp* allowing the compiler to generate very fast code. See the :ref:`file_cif++_matrix.hpp` for what is on offer.
+
+Crystals
+--------
+
+The *CIF* and *mmCIF* were initially developed to store crystallographic information on structures. Apart from coordinates and the chemical information the crystallographic information is important. This information can be split into two parts, a unit cell and a set of  :ref:`symmetry-ops` making up a spacegroup. The spacegroup number and name are stored in the *symmetry* category. The corresponding symmetry operations can be obtained in *libcifpp* by using the :cpp:class:`cif::spacegroup`. The cell is stored in the category *cell* and likewise can be loaded using the :cpp:class:`cif::cell`. Together these two classes make up a crystal and so we have a :cpp:class:`cif::crystal` which contains both. You can easily create such a crystal object by passing the datablock containing the data to the constructor. As in:
+
+.. code:: cpp
+
+    // Load the file
+    cif::file f("1cbs.cif.gz");
+
+    auto &db = f.front();
+    cif::crystal c(db);
+
+.. _symmetry-ops:
+Symmetry operations
+-------------------
+
+Each basic symmetry operation in the crystallographic world consists of a matrix multiplication followed by a translation. To apply such an operation on a carthesian coordinate you first have to convert the point into a fractional coordinate with respect to the unit cell of the crystal, then apply the matrix and translation operations and then convert the result back into carthesian coordinates. This is all done by the proper routines in *libcifpp*.
+
+Symmetry operations are encoded as a string in *mmCIF* PDBx files. The format is a string with the rotational number followed by an underscore and then the encoded translation in each direction where 5 means no translation. So, the identity operator is ``1_555`` meaning that we have rotational number 1 (which is always the identity rotation, point multiplied with the identity matrix) and a translation of zero in each direction.
+
+To give an idea how this works, here's a piece of code copied from one of the unit tests in *libcifpp*. It takes the *struct_conn* records in a certain PDB file and checks wether the distances in each row correspond to what we can calculate.
+
+.. code:: cpp
+
+    using namespace cif::literals;
+
+    // Load the file
+    cif::file f(gTestDir / "2bi3.cif.gz");
+
+    // Simply assume we can use the first datablock
+    auto &db = f.front();
+
+    // Load the crystal information
+    cif::crystal c(db);
+
+    // Take references to the two categories we need
+    auto struct_conn = db["struct_conn"];
+    auto atom_site = db["atom_site"];
+
+    // Loop over all rows in struct_conn taking only the values we need
+    for (const auto &[
+            asym1, seqid1, authseqid1, atomid1, symm1,
+            asym2, seqid2, authseqid2, atomid2, symm2,
+            dist] : struct_conn.find<
+                std::string,std::optional<int>,std::string,std::string,std::string,
+                std::string,std::optional<int>,std::string,std::string,std::string,
+                float>(
+            cif::key("ptnr1_symmetry") != "1_555" or cif::key("ptnr2_symmetry") != "1_555",
+            "ptnr1_label_asym_id", "ptnr1_label_seq_id", "ptnr1_auth_seq_id", "ptnr1_label_atom_id", "ptnr1_symmetry", 
+            "ptnr2_label_asym_id", "ptnr2_label_seq_id", "ptnr2_auth_seq_id", "ptnr2_label_atom_id", "ptnr2_symmetry", 
+            "pdbx_dist_value"
+        ))
+    {
+        // Find the location of the first atom
+        cif::point p1 = atom_site.find1<float,float,float>(
+            "label_asym_id"_key == asym1 and "label_seq_id"_key == seqid1 and "auth_seq_id"_key == authseqid1 and "label_atom_id"_key == atomid1,
+            "cartn_x", "cartn_y", "cartn_z");
+
+        // Find the location of the second atom
+        cif::point p2 = atom_site.find1<float,float,float>(
+            "label_asym_id"_key == asym2 and "label_seq_id"_key == seqid2 and "auth_seq_id"_key == authseqid2 and "label_atom_id"_key == atomid2,
+            "cartn_x", "cartn_y", "cartn_z");
+
+        // Calculate the position of the first atom using the symmetry operator defined in struct_conn
+        auto sa1 = c.symmetry_copy(p1, cif::sym_op(symm1));
+
+        // Calculate the position of the second atom using the symmetry operator defined in struct_conn
+        auto sa2 = c.symmetry_copy(p2, cif::sym_op(symm2));
+
+        // The distance between these symmetry atoms should be equal to the distance in the struct_conn record
+        assert(cif::distance(sa1, sa2) == dist);
+
+        // And to show how you can obtain the closest symmetry copy of an atom near another one:
+        // here we request the symmetry copy of p2 that lies closest to p1
+        const auto &[d, p, so] = c.closest_symmetry_copy(p1, p2);
+
+        // And that should of course be equal to the location in struct_conn for p2
+        assert(p.m_x == sa2.m_x);
+        assert(p.m_y == sa2.m_y);
+        assert(p.m_z == sa2.m_z);
+
+        // Distance and symmetry operator string should also be the same
+        assert(d == dist);
+        assert(so.string() == symm2);
+    }

examples/CMakeLists.txt

@@ -0,0 +1,7 @@
+cmake_minimum_required(VERSION 3.15)
+project(cifpp_example LANGUAGES CXX)
+
+find_package(cifpp REQUIRED)
+
+add_executable(example example.cpp)
+target_link_libraries(example cifpp::cifpp)

examples/example.cpp

@@ -8,13 +8,16 @@ namespace fs = std::filesystem;
 int main(int argc, char *argv[])
 {
 	if (argc != 2)
+	{
+		std::cerr << "Usage: example <inputfile>\n";
 		exit(1);
+	}
 
-	cif::file file = cif::pdb::read(argv[1]);
+	cif::file file(argv[1]);
 
 	if (file.empty())
 	{
-		std::cerr << "Empty file" << std::endl;
+		std::cerr << "Empty file\n";
 		exit(1);
 	}
 
@@ -22,13 +25,13 @@ int main(int argc, char *argv[])
 	auto &atom_site = db["atom_site"];
 	auto n = atom_site.find(cif::key("label_atom_id") == "OXT").size();
 
-	std::cout << "File contains " << atom_site.size() << " atoms of which " << n << (n == 1 ? " is" : " are") << " OXT" << std::endl
-			  << "residues with an OXT are:" << std::endl;
+	std::cout << "File contains " << atom_site.size() << " atoms of which " << n << (n == 1 ? " is" : " are") << " OXT\n"
+			  << "residues with an OXT are:\n";
 
 	for (const auto &[asym, comp, seqnr] : atom_site.find<std::string, std::string, int>(
 			 cif::key("label_atom_id") == "OXT", "label_asym_id", "label_comp_id", "label_seq_id"))
 	{
-		std::cout << asym << ' ' << comp << ' ' << seqnr << std::endl;
+		std::cout << asym << ' ' << comp << ' ' << seqnr << '\n';
 	}
 
 	return 0;

examples/makefile

@@ -1,8 +0,0 @@
-CXX = c++ -std=c++17
-CXXFLAGS = $(shell pkg-config --cflags libcifpp)
-LIBS = $(shell pkg-config --libs libcifpp)
-
-all: example
-
-example: example.cpp
-	$(CXX) -o $@ $? $(CXXFLAGS) $(LIBS)

include/cif++.hpp

@@ -37,5 +37,5 @@
 
 #include "cif++/model.hpp"
 
-#include "cif++/pdb/io.hpp"
+#include "cif++/pdb.hpp"
 #include "cif++/gzio.hpp"

include/cif++/atom_type.hpp

@@ -24,212 +24,259 @@
  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-// Lib for working with structures as contained in mmCIF and PDB files
+/** \file atom_type.hpp
+ * 
+ * This file contains information about all known elements
+ */
 
 #pragma once
 
+#include "cif++/exports.hpp"
+
 #include <cstdint>
+#include <limits>
 #include <stdexcept>
 #include <string>
 
 namespace cif
 {
 
+/** Atom type as an integer. All known elements are available as a constant. */
+
 enum atom_type : uint8_t
 {
-	Nn = 0, // Unknown
+	Nn = 0, ///< Unknown
 
-	H = 1,  // Hydro­gen
-	He = 2, // He­lium
+	H = 1,  ///< Hydro­gen
+	He = 2, ///< He­lium
 
-	Li = 3,  // Lith­ium
-	Be = 4,  // Beryl­lium
-	B = 5,   // Boron
-	C = 6,   // Carbon
-	N = 7,   // Nitro­gen
-	O = 8,   // Oxy­gen
-	F = 9,   // Fluor­ine
-	Ne = 10, // Neon
+	Li = 3,  ///< Lith­ium
+	Be = 4,  ///< Beryl­lium
+	B = 5,   ///< Boron
+	C = 6,   ///< Carbon
+	N = 7,   ///< Nitro­gen
+	O = 8,   ///< Oxy­gen
+	F = 9,   ///< Fluor­ine
+	Ne = 10, ///< Neon
 
-	Na = 11, // So­dium
-	Mg = 12, // Magne­sium
-	Al = 13, // Alumin­ium
-	Si = 14, // Sili­con
-	P = 15,  // Phos­phorus
-	S = 16,  // Sulfur
-	Cl = 17, // Chlor­ine
-	Ar = 18, // Argon
+	Na = 11, ///< So­dium
+	Mg = 12, ///< Magne­sium
+	Al = 13, ///< Alumin­ium
+	Si = 14, ///< Sili­con
+	P = 15,  ///< Phos­phorus
+	S = 16,  ///< Sulfur
+	Cl = 17, ///< Chlor­ine
+	Ar = 18, ///< Argon
 
-	K = 19,  // Potas­sium
-	Ca = 20, // Cal­cium
-	Sc = 21, // Scan­dium
-	Ti = 22, // Tita­nium
-	V = 23,  // Vana­dium
-	Cr = 24, // Chrom­ium
-	Mn = 25, // Manga­nese
-	Fe = 26, // Iron
-	Co = 27, // Cobalt
-	Ni = 28, // Nickel
-	Cu = 29, // Copper
-	Zn = 30, // Zinc
-	Ga = 31, // Gallium
-	Ge = 32, // Germa­nium
-	As = 33, // Arsenic
-	Se = 34, // Sele­nium
-	Br = 35, // Bromine
-	Kr = 36, // Kryp­ton
+	K = 19,  ///< Potas­sium
+	Ca = 20, ///< Cal­cium
+	Sc = 21, ///< Scan­dium
+	Ti = 22, ///< Tita­nium
+	V = 23,  ///< Vana­dium
+	Cr = 24, ///< Chrom­ium
+	Mn = 25, ///< Manga­nese
+	Fe = 26, ///< Iron
+	Co = 27, ///< Cobalt
+	Ni = 28, ///< Nickel
+	Cu = 29, ///< Copper
+	Zn = 30, ///< Zinc
+	Ga = 31, ///< Gallium
+	Ge = 32, ///< Germa­nium
+	As = 33, ///< Arsenic
+	Se = 34, ///< Sele­nium
+	Br = 35, ///< Bromine
+	Kr = 36, ///< Kryp­ton
 
-	Rb = 37, // Rubid­ium
-	Sr = 38, // Stront­ium
-	Y = 39,  // Yttrium
-	Zr = 40, // Zirco­nium
-	Nb = 41, // Nio­bium
-	Mo = 42, // Molyb­denum
-	Tc = 43, // Tech­netium
-	Ru = 44, // Ruthe­nium
-	Rh = 45, // Rho­dium
-	Pd = 46, // Pallad­ium
-	Ag = 47, // Silver
-	Cd = 48, // Cad­mium
-	In = 49, // Indium
-	Sn = 50, // Tin
-	Sb = 51, // Anti­mony
-	Te = 52, // Tellurium
-	I = 53,  // Iodine
-	Xe = 54, // Xenon
-	Cs = 55, // Cae­sium
-	Ba = 56, // Ba­rium
-	La = 57, // Lan­thanum
+	Rb = 37, ///< Rubid­ium
+	Sr = 38, ///< Stront­ium
+	Y = 39,  ///< Yttrium
+	Zr = 40, ///< Zirco­nium
+	Nb = 41, ///< Nio­bium
+	Mo = 42, ///< Molyb­denum
+	Tc = 43, ///< Tech­netium
+	Ru = 44, ///< Ruthe­nium
+	Rh = 45, ///< Rho­dium
+	Pd = 46, ///< Pallad­ium
+	Ag = 47, ///< Silver
+	Cd = 48, ///< Cad­mium
+	In = 49, ///< Indium
+	Sn = 50, ///< Tin
+	Sb = 51, ///< Anti­mony
+	Te = 52, ///< Tellurium
+	I = 53,  ///< Iodine
+	Xe = 54, ///< Xenon
+	Cs = 55, ///< Cae­sium
+	Ba = 56, ///< Ba­rium
+	La = 57, ///< Lan­thanum
 
-	Hf = 72, // Haf­nium
-	Ta = 73, // Tanta­lum
-	W = 74,  // Tung­sten
-	Re = 75, // Rhe­nium
-	Os = 76, // Os­mium
-	Ir = 77, // Iridium
-	Pt = 78, // Plat­inum
-	Au = 79, // Gold
-	Hg = 80, // Mer­cury
-	Tl = 81, // Thallium
-	Pb = 82, // Lead
-	Bi = 83, // Bis­muth
-	Po = 84, // Polo­nium
-	At = 85, // Asta­tine
-	Rn = 86, // Radon
-	Fr = 87, // Fran­cium
-	Ra = 88, // Ra­dium
-	Ac = 89, // Actin­ium
+	Hf = 72, ///< Haf­nium
+	Ta = 73, ///< Tanta­lum
+	W = 74,  ///< Tung­sten
+	Re = 75, ///< Rhe­nium
+	Os = 76, ///< Os­mium
+	Ir = 77, ///< Iridium
+	Pt = 78, ///< Plat­inum
+	Au = 79, ///< Gold
+	Hg = 80, ///< Mer­cury
+	Tl = 81, ///< Thallium
+	Pb = 82, ///< Lead
+	Bi = 83, ///< Bis­muth
+	Po = 84, ///< Polo­nium
+	At = 85, ///< Asta­tine
+	Rn = 86, ///< Radon
+	Fr = 87, ///< Fran­cium
+	Ra = 88, ///< Ra­dium
+	Ac = 89, ///< Actin­ium
 
-	Rf = 104, // Ruther­fordium
-	Db = 105, // Dub­nium
-	Sg = 106, // Sea­borgium
-	Bh = 107, // Bohr­ium
-	Hs = 108, // Has­sium
-	Mt = 109, // Meit­nerium
-	Ds = 110, // Darm­stadtium
-	Rg = 111, // Roent­genium
-	Cn = 112, // Coper­nicium
-	Nh = 113, // Nihon­ium
-	Fl = 114, // Flerov­ium
-	Mc = 115, // Moscov­ium
-	Lv = 116, // Liver­morium
-	Ts = 117, // Tenness­ine
-	Og = 118, // Oga­nesson
+	Rf = 104, ///< Ruther­fordium
+	Db = 105, ///< Dub­nium
+	Sg = 106, ///< Sea­borgium
+	Bh = 107, ///< Bohr­ium
+	Hs = 108, ///< Has­sium
+	Mt = 109, ///< Meit­nerium
+	Ds = 110, ///< Darm­stadtium
+	Rg = 111, ///< Roent­genium
+	Cn = 112, ///< Coper­nicium
+	Nh = 113, ///< Nihon­ium
+	Fl = 114, ///< Flerov­ium
+	Mc = 115, ///< Moscov­ium
+	Lv = 116, ///< Liver­morium
+	Ts = 117, ///< Tenness­ine
+	Og = 118, ///< Oga­nesson
 
-	Ce = 58, // Cerium
-	Pr = 59, // Praseo­dymium
-	Nd = 60, // Neo­dymium
-	Pm = 61, // Prome­thium
-	Sm = 62, // Sama­rium
-	Eu = 63, // Europ­ium
-	Gd = 64, // Gadolin­ium
-	Tb = 65, // Ter­bium
-	Dy = 66, // Dyspro­sium
-	Ho = 67, // Hol­mium
-	Er = 68, // Erbium
-	Tm = 69, // Thulium
-	Yb = 70, // Ytter­bium
-	Lu = 71, // Lute­tium
+	Ce = 58, ///< Cerium
+	Pr = 59, ///< Praseo­dymium
+	Nd = 60, ///< Neo­dymium
+	Pm = 61, ///< Prome­thium
+	Sm = 62, ///< Sama­rium
+	Eu = 63, ///< Europ­ium
+	Gd = 64, ///< Gadolin­ium
+	Tb = 65, ///< Ter­bium
+	Dy = 66, ///< Dyspro­sium
+	Ho = 67, ///< Hol­mium
+	Er = 68, ///< Erbium
+	Tm = 69, ///< Thulium
+	Yb = 70, ///< Ytter­bium
+	Lu = 71, ///< Lute­tium
 
-	Th = 90,  // Thor­ium
-	Pa = 91,  // Protac­tinium
-	U = 92,   // Ura­nium
-	Np = 93,  // Neptu­nium
-	Pu = 94,  // Pluto­nium
-	Am = 95,  // Ameri­cium
-	Cm = 96,  // Curium
-	Bk = 97,  // Berkel­ium
-	Cf = 98,  // Califor­nium
-	Es = 99,  // Einstei­nium
-	Fm = 100, // Fer­mium
-	Md = 101, // Mende­levium
-	No = 102, // Nobel­ium
-	Lr = 103, // Lawren­cium
+	Th = 90,  ///< Thor­ium
+	Pa = 91,  ///< Protac­tinium
+	U = 92,   ///< Ura­nium
+	Np = 93,  ///< Neptu­nium
+	Pu = 94,  ///< Pluto­nium
+	Am = 95,  ///< Ameri­cium
+	Cm = 96,  ///< Curium
+	Bk = 97,  ///< Berkel­ium
+	Cf = 98,  ///< Califor­nium
+	Es = 99,  ///< Einstei­nium
+	Fm = 100, ///< Fer­mium
+	Md = 101, ///< Mende­levium
+	No = 102, ///< Nobel­ium
+	Lr = 103, ///< Lawren­cium
 
-	D = 119, // Deuterium
+	D = 119, ///< Deuterium
 };
 
 // --------------------------------------------------------------------
-// atom_type_info
+
+/// An enum used to select the desired radius for an atom.
+/// All values are collected from the wikipedia pages on atom radii
 
 enum class radius_type
 {
-	calculated,
-	empirical,
+	calculated, ///< Calculated radius from theoretical models
+	empirical,  ///< Empirically measured covalent radii
+
+	/// @deprecated It is a bit unclear where these values came from. So, better not use them
 	covalent_empirical,
 
-	single_bond,
-	double_bond,
-	triple_bond,
+	single_bond, ///< Bond length for a single covalent bond calculated using statistically analysis
+	double_bond, ///< Bond length for a double covalent bond calculated using statistically analysis
+	triple_bond, ///< Bond length for a triple covalent bond calculated using statistically analysis
 
-	van_der_waals,
+	van_der_waals, ///< Radius of an imaginary hard sphere representing the distance of closest approach for another atom
 
-	type_count
+	type_count ///< Number of radii
 };
 
-constexpr size_t kRadiusTypeCount = static_cast<size_t>(radius_type::type_count);
+/// @brief The number of radii per element which can be requested from atom_type_info
+constexpr std::size_t kRadiusTypeCount = static_cast<std::size_t>(radius_type::type_count);
+
+/// An enum used to select either the effective or the crystal radius of an ion.
+/// See explanation on Wikipedia: https://en.wikipedia.org/wiki/Ionic_radius
 
 enum class ionic_radius_type
 {
-	effective, crystal
+	effective, ///< Based on distance between ions in a crystal structure as determined by X-ray crystallography
+	crystal    ///< Calculated ion radius based on a function of ionic charge and spin
 };
 
+/// Requests for an unknown radius value return kNA
+constexpr float kNA = std::numeric_limits<float>::quiet_NaN();
+
+/// A struct holding the known information for all elements defined in atom_type
+
 struct atom_type_info
 {
+	/// The type as an atom_type
 	atom_type type;
+
+	/// The official name for this element
 	std::string name;
+
+	/// The official symbol for this element
 	std::string symbol;
+
+	/// The weight of this element
 	float weight;
+
+	/// A flag indicating whether the element is a metal
 	bool metal;
+
+	/// Array containing all known radii for this element. A value of kNA is
+	/// stored for unknown values
 	float radii[kRadiusTypeCount];
 };
 
+/// Array of atom_type_info struct for each of the defined elements in atom_type
+
 extern CIFPP_EXPORT const atom_type_info kKnownAtoms[];
 
 // --------------------------------------------------------------------
 // AtomTypeTraits
 
+/// A traits class to access information for known elements
+
 class atom_type_traits
 {
   public:
+	/// Constructor taking an atom_type \a a
 	atom_type_traits(atom_type a);
+
+	/// Constructor based on the element as a string in \a symbol
 	atom_type_traits(const std::string &symbol);
 
-	atom_type type() const { return m_info->type; }
-	std::string name() const { return m_info->name; }
-	std::string symbol() const { return m_info->symbol; }
-	float weight() const { return m_info->weight; }
+	atom_type type() const { return m_info->type; }       ///< Returns the atom_type
+	std::string name() const { return m_info->name; }     ///< Returns the name of the element
+	std::string symbol() const { return m_info->symbol; } ///< Returns the symbol of the element
+	float weight() const { return m_info->weight; }       ///< Returns the average weight of the element
 
-	bool is_metal() const { return m_info->metal; }
+	bool is_metal() const { return m_info->metal; } ///< Returns true if the element is a metal
 
+	/// Return true if the symbol in \a symbol actually exists in the list of known elements in atom_type
 	static bool is_element(const std::string &symbol);
+
+	/// Return true if the symbol in \a symbol exists and is a metal
 	static bool is_metal(const std::string &symbol);
 
+	/// @brief Return the radius for the element, use \a type to select which radius to return
+	/// @param type The selector for which radius to return
+	/// @return The requested radius or kNA if not known (or applicable)
 	float radius(radius_type type = radius_type::single_bond) const
 	{
 		if (type >= radius_type::type_count)
 			throw std::invalid_argument("invalid radius requested");
-		return m_info->radii[static_cast<size_t>(type)] / 100.f;
+		return m_info->radii[static_cast<std::size_t>(type)] / 100.f;
 	}
 
 	/// \brief Return the radius for a charged version of this atom in a solid crystal
@@ -247,35 +294,46 @@ class atom_type_traits
 	/// \brief Return the radius for a charged version of this atom, returns the effective radius by default
 	///
 	/// \param charge  The charge of the ion
+	/// \param type    The requested ion radius type
 	/// \return        The radius of the ion
 	float ionic_radius(int charge, ionic_radius_type type = ionic_radius_type::effective) const
 	{
 		return type == ionic_radius_type::effective ? effective_ionic_radius(charge) : crystal_ionic_radius(charge);
 	}
 
-	// data type encapsulating the Waasmaier & Kirfel scattering factors
-	// in a simplified form (only a and b).
-	// Added the electrion scattering factors as well
+	/**
+	 * @brief data type encapsulating the scattering factors
+	 * in a simplified form (only a and b).
+	 */
 	struct SFData
 	{
+		/** @cond */
 		double a[6], b[6];
+		/** @endcond */
 	};
 
-	// to get the Cval and Siva values, use this constant as charge:
-	enum
-	{
-		kWKSFVal = -99
-	};
+	/// @brief to get the Cval and Siva scattering factor values, use this constant as charge:
+	static constexpr int kWKSFVal = -99;
 
+	/// @brief Return the Waasmaier & Kirfel scattering factor values for the element
+	///
+	/// The coefficients from Waasmaier & Kirfel (1995), Acta Cryst. A51, 416-431.
+	///
+	/// @param charge The charge for which the structure values should be returned, use kWSKFVal to return the *Cval* and *Siva* values
+	/// @return The scattering factors as a SFData struct
 	const SFData &wksf(int charge = 0) const;
+
+	/// @brief Return the electron scattering factor values for the element
+	///
+	/// @return The scattering factors as a SFData struct
 	const SFData &elsf() const;
 
-	// Clipper doesn't like atoms with charges that do not have a scattering factor. And
-	// rightly so, but we need to know in advance if this is the case
+	/// Clipper doesn't like atoms with charges that do not have a scattering factor. And
+	/// rightly so, but we need to know in advance if this is the case
 	bool has_sf(int charge) const;
 
   private:
 	const struct atom_type_info *m_info;
 };
 
-} // namespace pdbx
+} // namespace cif

include/cif++/compound.hpp

@@ -26,17 +26,28 @@
 
 #pragma once
 
-/// \file This file contains the definition for the class compound, encapsulating
-/// the information found for compounds in the CCD.
-
 #include "cif++/atom_type.hpp"
+#include "cif++/datablock.hpp"
+#include "cif++/exports.hpp"
 #include "cif++/point.hpp"
+#include "cif++/utilities.hpp"
 
 #include <map>
 #include <set>
 #include <tuple>
 #include <vector>
 
+/// \file compound.hpp
+/// This file contains the definition for the class compound, encapsulating
+/// the information found for compounds in the CCD.
+///
+/// The data is loaded by default from a file called `components.cif`. This file
+/// is located using load_resource. (See documentation on cif::load_resource for more information)
+///
+/// Note that since version 6 the CCP4 monomer library is no longer used.
+
+/// See also :doc:`/compound` for more information.
+
 namespace cif
 {
 
@@ -46,21 +57,47 @@ class compound;
 struct compound_atom;
 class compound_factory_impl;
 
-/// \brief The bond type as defined in the CCD, possible values taken from the mmcif_pdbx file
+/// \brief The bond type or bond order as defined in the CCD, possible values taken from the mmcif_pdbx file
 enum class bond_type
 {
-	sing, // 'single bond'
-	doub, // 'double bond'
-	trip, // 'triple bond'
-	quad, // 'quadruple bond'
-	arom, // 'aromatic bond'
-	poly, // 'polymeric bond'
-	delo, // 'delocalized double bond'
-	pi,   // 'pi bond'
+	sing, ///< single bond
+	doub, ///< double bond
+	trip, ///< triple bond
+	quad, ///< quadruple bond
+	arom, ///< aromatic bond
+	poly, ///< polymeric bond
+	delo, ///< delocalized double bond
+	pi,   ///< pi bond
 };
 
-std::string to_string(bond_type bondType);
-bond_type from_string(const std::string &bondType);
+/// @brief return the string representation of @a bondType
+std::string bond_type_to_string(bond_type bondType);
+
+/// @brief return the cif::bond_type for the string representation @a bondType
+bond_type parse_bond_type_from_string(const std::string &bondType);
+
+/// \brief The possible stereo config values for a compound_atom.
+///
+/// As the site https://psiberg.com/r-s-nomenclature/ states:
+///
+/// > RS nomenclature is currently the preferred system for assigning absolute
+/// > configuration to chiral molecules. The letters R and S come from the Latin
+/// > words ‘Rectus‘ and ‘Sinister‘ meaning ‘right’ and ‘left’. Molecules that
+/// > rotate the plane of polarized light to right are referred to as ‘R isomers’
+/// > and the molecules that rotate the plane of polarized light to left are
+/// > referred to ‘S isomers’.
+enum class stereo_config_type : uint8_t
+{
+	N = 'N', ///< Not polarizing
+	R = 'R', ///< Rectus
+	S = 'S'  ///< Sinister
+};
+
+/// @brief return the string representation of @a stereo_config
+std::string to_string(stereo_config_type stereo_config);
+
+/// @brief return the cif::stereo_config_type for the string representation @a stereo_config
+stereo_config_type parse_stereo_config_from_string(const std::string &stereo_config);
 
 /// --------------------------------------------------------------------
 /// \brief struct containing information about an atom in a chemical compound.
@@ -68,14 +105,17 @@ bond_type from_string(const std::string &bondType);
 
 struct compound_atom
 {
-	std::string id;
-	atom_type type_symbol;
-	int charge = 0;
-	bool aromatic = false;
-	bool leaving_atom = false;
-	bool stereo_config = false;
-	float x, y, z;
+	std::string id;                                           ///< Identifier for each atom in the chemical component
+	atom_type type_symbol;                                    ///< The element type for each atom in the chemical component.
+	int charge = 0;                                           ///< The formal charge assigned to each atom in the chemical component.
+	bool aromatic = false;                                    ///< Defines atoms in an aromatic moiety
+	bool leaving_atom = false;                                ///< Flags atoms with "leaving" capability
+	stereo_config_type stereo_config = stereo_config_type::N; ///< Defines the stereochemical configuration of the chiral center atom.
+	float x,                                                  ///< The x component of the coordinates for each atom specified as orthogonal angstroms.
+		y,                                                    ///< The y component of the coordinates for each atom specified as orthogonal angstroms.
+		z;                                                    ///< The z component of the coordinates for each atom specified as orthogonal angstroms.
 
+	/// Return the location of the atom as a point
 	point get_location() const
 	{
 		return { x, y, z };
@@ -87,9 +127,10 @@ struct compound_atom
 
 struct compound_bond
 {
-	std::string atom_id[2];
-	bond_type type;
-	bool aromatic = false, stereo_config = false;
+	std::string atom_id[2];    ///< The ID's of the two atoms that define the bond.
+	bond_type type;            ///< The bond order of the chemical bond associated with the specified atoms.
+	bool aromatic = false,     ///< Defines aromatic bonds.
+		stereo_config = false; ///< Defines stereochemical bonds.
 };
 
 /// --------------------------------------------------------------------
@@ -97,7 +138,7 @@ struct compound_bond
 /// This information is derived from the CDD by default.
 ///
 /// To create compounds, you use the factory method. You can add your own
-/// compound definitions by calling the addExtraComponents function and
+/// compound definitions by calling the push_dictionary function and
 /// pass it a valid CCD formatted file.
 
 class compound
@@ -105,40 +146,47 @@ class compound
   public:
 	// accessors
 
-	std::string id() const { return m_id; }
-	std::string name() const { return m_name; }
-	std::string type() const { return m_type; }
-	std::string group() const { return m_group; }
-	std::string formula() const { return m_formula; }
-	float formula_weight() const { return m_formula_weight; }
-	int formal_charge() const { return m_formal_charge; }
+	std::string id() const { return m_id; }                   ///< Return the alphanumeric code for the chemical component.
+	std::string name() const { return m_name; }               ///< Return the name of the chemical component.
+	std::string type() const { return m_type; }               ///< Return the type of monomer.
+	std::string formula() const { return m_formula; }         ///< Return the chemical formula of the chemical component.
+	float formula_weight() const { return m_formula_weight; } ///< Return the formula mass of the chemical component in Daltons.
+	int formal_charge() const { return m_formal_charge; }     ///< Return the formal charge on the chemical component.
 
-	const std::vector<compound_atom> &atoms() const { return m_atoms; }
-	const std::vector<compound_bond> &bonds() const { return m_bonds; }
+	const std::vector<compound_atom> &atoms() const { return m_atoms; } ///< Return the list of atoms for this compound
+	const std::vector<compound_bond> &bonds() const { return m_bonds; } ///< Return the list of bonds for this compound
 
-	compound_atom get_atom_by_atom_id(const std::string &atom_id) const;
+	compound_atom get_atom_by_atom_id(const std::string &atom_id) const; ///< Return the atom with id @a atom_id
 
-	bool atoms_bonded(const std::string &atomId_1, const std::string &atomId_2) const;
-	float bond_length(const std::string &atomId_1, const std::string &atomId_2) const;
+	bool atoms_bonded(const std::string &atomId_1, const std::string &atomId_2) const; ///< Return true if @a atomId_1 is bonded to @a atomId_2
+	float bond_length(const std::string &atomId_1, const std::string &atomId_2) const; ///< Return the bond length between @a atomId_1 and @a atomId_2
 
-	bool is_water() const
+	bool is_water() const ///< Return if the compound is actually a water
 	{
 		return m_id == "HOH" or m_id == "H2O" or m_id == "WAT";
 	}
 
+	/** \brief Return whether this compound has a type of either 'peptide linking' or 'L-peptide linking' */
+	bool is_peptide() const;
+
+	/** \brief Return whether this compound has a type of either 'DNA linking' or 'RNA linking' */
+	bool is_base() const;
+
+	char one_letter_code() const { return m_one_letter_code; }; ///< Return the one letter code to use in a canonical sequence. If unknown the value '\0' is returned
+	std::string parent_id() const { return m_parent_id; };      ///< Return the parent id code in case a parent is specified (e.g. MET for MSE)
+
   private:
 	friend class compound_factory_impl;
-	friend class CCD_compound_factory_impl;
-	friend class CCP4_compound_factory_impl;
+	friend class local_compound_factory_impl;
 
 	compound(cif::datablock &db);
-	compound(cif::datablock &db, const std::string &id, const std::string &name, const std::string &type, const std::string &group);
-
+	
 	std::string m_id;
 	std::string m_name;
 	std::string m_type;
-	std::string m_group;
 	std::string m_formula;
+	char m_one_letter_code = 0;
+	std::string m_parent_id;
 	float m_formula_weight = 0;
 	int m_formal_charge = 0;
 	std::vector<compound_atom> m_atoms;
@@ -148,6 +196,8 @@ class compound
 // --------------------------------------------------------------------
 // Factory class for compound and Link objects
 
+/// Use the compound_factory singleton instance to create compound objects
+
 class compound_factory
 {
   public:
@@ -159,27 +209,86 @@ class compound_factory
 	/// flag to true.
 
 	static void init(bool useThreadLocalInstanceOnly);
+
+	/// Return the singleton instance. If initialized with local threads, this is the
+	/// instance for the current thread.
 	static compound_factory &instance();
+
+	/// Delete and reset the singleton instance. If initialized with local threads, this is the
+	/// instance for the current thread.
 	static void clear();
 
+	/// Set the default dictionary file to @a inDictFile
 	void set_default_dictionary(const std::filesystem::path &inDictFile);
+
+	/// Override any previously loaded dictionary with @a inDictFile
 	void push_dictionary(const std::filesystem::path &inDictFile);
+
+	/** @brief Override any previously loaded dictionary with the data in @a file
+	 *
+	 * @note experimental feature
+	 *
+	 * Load the file @a file as a source for compound information. This may
+	 * be e.g. a regular mmCIF file with extra files containing compound
+	 * information.
+	 *
+	 * Be carefull to remove the block again, best use @ref cif::compound_source
+	 * as a stack based object.
+	 */
+
+	void push_dictionary(const file &file);
+
+	/// Remove the last pushed dictionary
 	void pop_dictionary();
 
+	/// Return whether @a res_name is a valid and known peptide
+	[[deprecated("use is_peptide or is_std_peptide instead)")]]
 	bool is_known_peptide(const std::string &res_name) const;
+
+	/// Return whether @a res_name is a valid and known base
+	[[deprecated("use is_base or is_std_base instead)")]]
 	bool is_known_base(const std::string &res_name) const;
 
+	/// Return whether @a res_name is a peptide
+	bool is_peptide(std::string_view res_name) const;
+
+	/// Return whether @a res_name is a base
+	bool is_base(std::string_view res_name) const;
+
+	/// Return whether @a res_name is one of the standard peptides
+	bool is_std_peptide(std::string_view res_name) const;
+
+	/// Return whether @a res_name is one of the standard bases
+	bool is_std_base(std::string_view res_name) const;
+
+	/// Return whether @a res_name is a monomer (either base or peptide)
+	bool is_monomer(std::string_view res_name) const;
+
+	/// Return whether @a res_name is one of the standard bases or peptides
+	bool is_std_monomer(std::string_view res_name) const
+	{
+		return is_std_base(res_name) or is_std_peptide(res_name);
+	}
+
+	bool is_water(std::string_view res_name) const
+	{
+		return res_name == "HOH" or res_name == "H2O" or res_name == "WAT";
+	}
+
 	/// \brief Create the compound object for \a id
 	///
 	/// This will create the compound instance for \a id if it doesn't exist already.
 	/// The result is owned by this factory and should not be deleted by the user.
 	/// \param id	The compound ID, a three letter code usually
 	/// \result		The compound, or nullptr if it could not be created (missing info)
-	const compound *create(std::string id);
+	const compound *create(std::string_view id);
 
 	~compound_factory();
 
-	static CIFPP_EXPORT const std::map<std::string, char> kAAMap, kBaseMap;
+	CIFPP_EXPORT static const std::map<std::string, char> kAAMap, ///< Globally accessible static list of the default amino acids
+		kBaseMap;                                                 ///< Globally accessible static list of the default bases
+
+	void report_missing_compound(std::string_view compound_id);
 
   private:
 	compound_factory();
@@ -194,4 +303,35 @@ class compound_factory
 	std::shared_ptr<compound_factory_impl> m_impl;
 };
 
+// --------------------------------------------------------------------
+
+/**
+ * @brief Stack based source for compound info.
+ *
+ * Use this class to temporarily add a compound source to the
+ * compound_factory.
+ *
+ * @code{.cpp}
+ * cif::file f("1cbs-with-custom-rea.cif");
+ * cif::compound_source cs(f);
+ *
+ * auto &cf = cif::compound_factory::instance();
+ * auto rea_compound = cf.create("REA");
+ * @endcode
+ */
+
+class compound_source
+{
+  public:
+	compound_source(const cif::file &file)
+	{
+		cif::compound_factory::instance().push_dictionary(file);
+	}
+
+	~compound_source()
+	{
+		cif::compound_factory::instance().pop_dictionary();
+	}
+};
+
 } // namespace cif

include/cif++/condition.hpp

@@ -29,20 +29,116 @@
 #include "cif++/row.hpp"
 
 #include <cassert>
+#include <concepts>
 #include <functional>
 #include <iostream>
 #include <regex>
 #include <utility>
 
+/** \file condition.hpp
+ * This file contains code to create conditions: object encapsulating a
+ * query you can use to find rows in a @ref cif::category
+ *
+ * Conditions are created as standard C++ expressions. That means
+ * you can use the standard comparison operators to compare item
+ * contents with a value and boolean operators to chain everything
+ * together.
+ *
+ * To create a query that simply compares one item with one value:
+ *
+ * @code {.cpp}
+ * cif::condition c = cif::key("id") == 1;
+ * @endcode
+ * 
+ * That will find rows where the ID item contains the number 1. If
+ * using cif::key is a bit too much typing, you can also write:
+ * 
+ * @code{.cpp}
+ * using namespace cif::literals;
+ * 
+ * cif::condition c2 = "id"_key == 1;
+ * @endcode
+ * 
+ * Now if you want both ID = 1 and ID = 2 in the result:
+ * 
+ * @code{.cpp}
+ * auto c3 = "id"_key == 1 or "id"_key == 2;
+ * @endcode
+ * 
+ * There are some special values you can use. To find rows with item that
+ * do not have a value:
+ * 
+ * @code{.cpp}
+ * auto c4 = "type"_key == cif::null;
+ * @endcode 
+ * 
+ * Of if it should not be NULL:
+ * 
+ * @code{.cpp}
+ * auto c5 = "type"_key != cif::null;
+ * @endcode 
+ * 
+ * There's even a way to find all records:
+ * 
+ * @code{.cpp}
+ * auto c6 = cif::all;
+ * @endcode
+ * 
+ * And when you want to search for any item containing the value 'foo':
+ * 
+ * @code{.cpp}
+ * auto c7 = cif::any == "foo";
+ * @endcode 
+ * 
+ * All these conditions can be chained together again:
+ * 
+ * @code{.cpp}
+ * auto c8 = std::move(c3) and std::move(c5);
+ * @endcode
+ */
+
 namespace cif
 {
 
 // --------------------------------------------------------------------
-// let's make life easier
+/// let's make life easier, since @ref cif::category is not known yet,
+/// we declare a function to access its contents
 
+/**
+ * @brief Get the items that can be used as key in conditions for a category
+ * 
+ * @param cat The category whose items to return
+ * @return iset The set of key item names
+ */
+[[deprecated("use get_category_items instead")]]
 iset get_category_fields(const category &cat);
-uint16_t get_column_ix(const category &cat, std::string_view col);
-bool is_column_type_uchar(const category &cat, std::string_view col);
+
+/**
+ * @brief Get the items that can be used as key in conditions for a category
+ * 
+ * @param cat The category whose items to return
+ * @return iset The set of key field names
+ */
+iset get_category_items(const category &cat);
+
+/**
+ * @brief Get the item index for item @a col in category @a cat
+ * 
+ * @param cat The category
+ * @param col The name of the item
+ * @return uint16_t The index
+ */
+uint16_t get_item_ix(const category &cat, std::string_view col);
+
+/**
+ * @brief Return whether the item @a col in category @a cat has a primitive type of *uchar*
+ * 
+ * @param cat The category
+ * @param col The item name
+ * @return true If the primitive type is of type *uchar*
+ * @return false If the primitive type is not of type *uchar*
+ */
+bool is_item_type_uchar(const category &cat, std::string_view col);
 
 // --------------------------------------------------------------------
 // some more templates to be able to do querying
@@ -72,16 +168,32 @@ namespace detail
 	struct not_condition_impl;
 } // namespace detail
 
+/**
+ * @brief The interface class for conditions. This uses the bridge pattern,
+ * which means the implementation is in the member m_impl
+ */
 class condition
 {
   public:
-	using condition_impl = detail::condition_impl;
 
+	/** @cond */
+	using condition_impl = detail::condition_impl;
+	/** @endcond */
+
+	/**
+	 * @brief Construct a new, empty condition object
+	 * 
+	 */
 	condition()
 		: m_impl(nullptr)
 	{
 	}
 
+	/**
+	 * @brief Construct a new condition object with implementation @a impl
+	 * 
+	 * @param impl The implementation to use
+	 */
 	explicit condition(condition_impl *impl)
 		: m_impl(impl)
 	{
@@ -89,6 +201,9 @@ class condition
 
 	condition(const condition &) = delete;
 
+	/**
+	 * @brief Construct a new condition object moving the data from @a rhs
+	 */
 	condition(condition &&rhs) noexcept
 		: m_impl(nullptr)
 	{
@@ -97,6 +212,9 @@ class condition
 
 	condition &operator=(const condition &) = delete;
 
+	/**
+	 * @brief Assignment operator moving the data from @a rhs
+	 */
 	condition &operator=(condition &&rhs) noexcept
 	{
 		std::swap(m_impl, rhs.m_impl);
@@ -109,8 +227,22 @@ class condition
 		m_impl = nullptr;
 	}
 
+	/**
+	 * @brief Prepare the condition to be used on category @a c. This will
+	 * take care of setting the correct indices for items e.g.
+	 * 
+	 * @param c The category this query should act upon
+	 */
 	void prepare(const category &c);
 
+	/**
+	 * @brief This operator returns true if the row referenced by @a r is 
+	 * a match for this condition.
+	 * 
+	 * @param r The reference to a row.
+	 * @return true If there is a match
+	 * @return false If there is no match
+	 */
 	bool operator()(row_handle r) const
 	{
 		assert(this->m_impl != nullptr);
@@ -118,27 +250,53 @@ class condition
 		return m_impl ? m_impl->test(r) : false;
 	}
 
+	/**
+	 * @brief Return true if the condition is not empty
+	 */
 	explicit operator bool() { return not empty(); }
+
+	/**
+	 * @brief Return true if the condition is empty, has no condition
+	 */
 	bool empty() const { return m_impl == nullptr; }
 
+	/**
+	 * @brief If the prepare step found out there is only one hit
+	 * this single hit can be returned by this method.
+	 * 
+	 * @return std::optional<row_handle> The result will contain
+	 * a row reference if there is a single hit, it will be empty otherwise
+	 */
 	std::optional<row_handle> single() const
 	{
 		return m_impl ? m_impl->single() : std::optional<row_handle>();
 	}
 
-	friend condition operator||(condition &&a, condition &&b);
-	friend condition operator&&(condition &&a, condition &&b);
+	friend condition operator||(condition &&a, condition &&b); /**< Return a condition which is the logical OR or condition @a and @b */
+	friend condition operator&&(condition &&a, condition &&b); /**< Return a condition which is the logical AND or condition @a and @b */
 
+	/// @cond
 	friend struct detail::or_condition_impl;
 	friend struct detail::and_condition_impl;
 	friend struct detail::not_condition_impl;
+	/// @endcond
 
+	/**
+	 * @brief Swap two conditions
+	 */
 	void swap(condition &rhs)
 	{
 		std::swap(m_impl, rhs.m_impl);
 		std::swap(m_prepared, rhs.m_prepared);
 	}
 
+	/**
+	 * @brief Operator to use to write out a condition to @a os, for debugging purposes
+	 * 
+	 * @param os The std::ostream to write to
+	 * @param cond The condition to write
+	 * @return std::ostream& The same as @a os
+	 */
 	friend std::ostream &operator<<(std::ostream &os, const condition &cond)
 	{
 		if (cond.m_impl)
@@ -157,14 +315,14 @@ namespace detail
 {
 	struct key_is_empty_condition_impl : public condition_impl
 	{
-		key_is_empty_condition_impl(const std::string &item_tag)
-			: m_item_tag(item_tag)
+		key_is_empty_condition_impl(const std::string &item_name)
+			: m_item_name(item_name)
 		{
 		}
 
 		condition_impl *prepare(const category &c) override
 		{
-			m_item_ix = get_column_ix(c, m_item_tag);
+			m_item_ix = get_item_ix(c, m_item_name);
 			return this;
 		}
 
@@ -175,23 +333,23 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << m_item_tag << " IS NULL";
+			os << m_item_name << " IS NULL";
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix = 0;
 	};
 
 	struct key_is_not_empty_condition_impl : public condition_impl
 	{
-		key_is_not_empty_condition_impl(const std::string &item_tag)
-			: m_item_tag(item_tag)
+		key_is_not_empty_condition_impl(const std::string &item_name)
+			: m_item_name(item_name)
 		{
 		}
 
 		condition_impl *prepare(const category &c) override
 		{
-			m_item_ix = get_column_ix(c, m_item_tag);
+			m_item_ix = get_item_ix(c, m_item_name);
 			return this;
 		}
 
@@ -202,18 +360,18 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << m_item_tag << " IS NOT NULL";
+			os << m_item_name << " IS NOT NULL";
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix = 0;
 	};
 
 	struct key_equals_condition_impl : public condition_impl
 	{
 		key_equals_condition_impl(item &&i)
-			: m_item_tag(i.name())
-			, m_value(i.value())
+			: m_item_name(i.name())
+			, m_value(std::forward<item>(i).value())
 		{
 		}
 
@@ -226,7 +384,7 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << m_item_tag << (m_icase ? "^ " : " ") << " == " << m_value;
+			os << m_item_name << (m_icase ? "^ " : " ") << " == " << m_value;
 		}
 
 		virtual std::optional<row_handle> single() const override
@@ -242,13 +400,13 @@ namespace detail
 				if (m_single_hit.has_value() or ri->m_single_hit.has_value())
 					return m_single_hit == ri->m_single_hit;
 				else
-					// watch out, both m_item_ix might be the same while tags might be diffent (in case they both do not exist in the category)
-					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_tag == ri->m_item_tag;
+					// watch out, both m_item_ix might be the same while item_names might be diffent (in case they both do not exist in the category)
+					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_name == ri->m_item_name;
 			}
 			return this == rhs;
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix = 0;
 		bool m_icase = false;
 		std::string m_value;
@@ -258,7 +416,7 @@ namespace detail
 	struct key_equals_or_empty_condition_impl : public condition_impl
 	{
 		key_equals_or_empty_condition_impl(key_equals_condition_impl *equals)
-			: m_item_tag(equals->m_item_tag)
+			: m_item_name(equals->m_item_name)
 			, m_value(equals->m_value)
 			, m_icase(equals->m_icase)
 			, m_single_hit(equals->m_single_hit)
@@ -267,8 +425,8 @@ namespace detail
 
 		condition_impl *prepare(const category &c) override
 		{
-			m_item_ix = get_column_ix(c, m_item_tag);
-			m_icase = is_column_type_uchar(c, m_item_tag);
+			m_item_ix = get_item_ix(c, m_item_name);
+			m_icase = is_item_type_uchar(c, m_item_name);
 			return this;
 		}
 
@@ -284,7 +442,7 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << '(' << m_item_tag << (m_icase ? "^ " : " ") << " == " << m_value << " OR " << m_item_tag << " IS NULL)";
+			os << '(' << m_item_name << (m_icase ? "^ " : " ") << " == " << m_value << " OR " << m_item_name << " IS NULL)";
 		}
 
 		virtual std::optional<row_handle> single() const override
@@ -300,24 +458,124 @@ namespace detail
 				if (m_single_hit.has_value() or ri->m_single_hit.has_value())
 					return m_single_hit == ri->m_single_hit;
 				else
-					// watch out, both m_item_ix might be the same while tags might be diffent (in case they both do not exist in the category)
-					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_tag == ri->m_item_tag;
+					// watch out, both m_item_ix might be the same while item_names might be diffent (in case they both do not exist in the category)
+					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_name == ri->m_item_name;
 			}
 			return this == rhs;
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix = 0;
 		std::string m_value;
 		bool m_icase = false;
 		std::optional<row_handle> m_single_hit;
 	};
 
+	struct key_equals_number_condition_impl : public condition_impl
+	{
+		key_equals_number_condition_impl(const std::string &name, double v)
+			: m_item_name(name)
+			, m_value(v)
+		{
+		}
+
+		condition_impl *prepare(const category &c) override;
+
+		bool test(row_handle r) const override
+		{
+			return m_single_hit.has_value() ? *m_single_hit == r : r[m_item_ix].compare(m_value) == 0;
+		}
+
+		void str(std::ostream &os) const override
+		{
+			os << m_item_name << " == " << m_value;
+		}
+
+		virtual std::optional<row_handle> single() const override
+		{
+			return m_single_hit;
+		}
+
+		virtual bool equals(const condition_impl *rhs) const override
+		{
+			if (typeid(*rhs) == typeid(key_equals_number_condition_impl))
+			{
+				auto ri = static_cast<const key_equals_number_condition_impl *>(rhs);
+				if (m_single_hit.has_value() or ri->m_single_hit.has_value())
+					return m_single_hit == ri->m_single_hit;
+				else
+					// watch out, both m_item_ix might be the same while item_names might be diffent (in case they both do not exist in the category)
+					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_name == ri->m_item_name;
+			}
+			return this == rhs;
+		}
+
+		std::string m_item_name;
+		uint16_t m_item_ix = 0;
+		double m_value;
+		std::optional<row_handle> m_single_hit;
+	};
+
+	struct key_equals_number_or_empty_condition_impl : public condition_impl
+	{
+		key_equals_number_or_empty_condition_impl(key_equals_number_condition_impl *equals)
+			: m_item_name(equals->m_item_name)
+			, m_value(equals->m_value)
+			, m_single_hit(equals->m_single_hit)
+		{
+		}
+
+		condition_impl *prepare(const category &c) override
+		{
+			m_item_ix = get_item_ix(c, m_item_name);
+			return this;
+		}
+
+		bool test(row_handle r) const override
+		{
+			bool result = false;
+			if (m_single_hit.has_value())
+				result = *m_single_hit == r;
+			else
+				result = r[m_item_ix].empty() or r[m_item_ix].compare(m_value) == 0;
+			return result;
+		}
+
+		void str(std::ostream &os) const override
+		{
+			os << '(' << m_item_name << " == " << m_value << " OR " << m_item_name << " IS NULL)";
+		}
+
+		virtual std::optional<row_handle> single() const override
+		{
+			return m_single_hit;
+		}
+
+		virtual bool equals(const condition_impl *rhs) const override
+		{
+			if (typeid(*rhs) == typeid(key_equals_number_or_empty_condition_impl))
+			{
+				auto ri = static_cast<const key_equals_number_or_empty_condition_impl *>(rhs);
+				if (m_single_hit.has_value() or ri->m_single_hit.has_value())
+					return m_single_hit == ri->m_single_hit;
+				else
+					// watch out, both m_item_ix might be the same while item_names might be diffent (in case they both do not exist in the category)
+					return m_item_ix == ri->m_item_ix and m_value == ri->m_value and m_item_name == ri->m_item_name;
+			}
+			return this == rhs;
+		}
+
+		std::string m_item_name;
+		uint16_t m_item_ix = 0;
+		double m_value;
+		std::optional<row_handle> m_single_hit;
+	};
+
 	struct key_compare_condition_impl : public condition_impl
 	{
 		template <typename COMP>
-		key_compare_condition_impl(const std::string &item_tag, COMP &&comp, const std::string &s)
-			: m_item_tag(item_tag)
+		key_compare_condition_impl(const std::string &item_name, COMP &&comp, const std::string &s)
+			: m_item_name(item_name)
 			, m_compare(std::move(comp))
 			, m_str(s)
 		{
@@ -325,8 +583,8 @@ namespace detail
 
 		condition_impl *prepare(const category &c) override
 		{
-			m_item_ix = get_column_ix(c, m_item_tag);
-			m_icase = is_column_type_uchar(c, m_item_tag);
+			m_item_ix = get_item_ix(c, m_item_name);
+			m_icase = is_item_type_uchar(c, m_item_name);
 			return this;
 		}
 
@@ -337,10 +595,10 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << m_item_tag << (m_icase ? "^ " : " ") << m_str;
+			os << m_item_name << (m_icase ? "^ " : " ") << m_str;
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix = 0;
 		bool m_icase = false;
 		std::function<bool(row_handle, bool)> m_compare;
@@ -349,8 +607,8 @@ namespace detail
 
 	struct key_matches_condition_impl : public condition_impl
 	{
-		key_matches_condition_impl(const std::string &item_tag, const std::regex &rx)
-			: m_item_tag(item_tag)
+		key_matches_condition_impl(const std::string &item_name, const std::regex &rx)
+			: m_item_name(item_name)
 			, m_item_ix(0)
 			, mRx(rx)
 		{
@@ -358,7 +616,7 @@ namespace detail
 
 		condition_impl *prepare(const category &c) override
 		{
-			m_item_ix = get_column_ix(c, m_item_tag);
+			m_item_ix = get_item_ix(c, m_item_name);
 			return this;
 		}
 
@@ -370,10 +628,10 @@ namespace detail
 
 		void str(std::ostream &os) const override
 		{
-			os << m_item_tag << " =~ expression";
+			os << m_item_name << " =~ expression";
 		}
 
-		std::string m_item_tag;
+		std::string m_item_name;
 		uint16_t m_item_ix;
 		std::regex mRx;
 	};
@@ -393,7 +651,7 @@ namespace detail
 			auto &c = r.get_category();
 
 			bool result = false;
-			for (auto &f : get_category_fields(c))
+			for (auto &f : get_category_items(c))
 			{
 				try
 				{
@@ -431,7 +689,7 @@ namespace detail
 			auto &c = r.get_category();
 
 			bool result = false;
-			for (auto &f : get_category_fields(c))
+			for (auto &f : get_category_items(c))
 			{
 				try
 				{
@@ -691,6 +949,9 @@ namespace detail
 
 } // namespace detail
 
+/**
+ * @brief Create a condition containing the logical AND of conditions @a a and @a b
+ */
 inline condition operator and(condition &&a, condition &&b)
 {
 	if (a.m_impl and b.m_impl)
@@ -700,6 +961,9 @@ inline condition operator and(condition &&a, condition &&b)
 	return condition(std::move(b));
 }
 
+/**
+ * @brief Create a condition containing the logical OR of conditions @a a and @a b
+ */
 inline condition operator or(condition &&a, condition &&b)
 {
 	if (a.m_impl and b.m_impl)
@@ -710,19 +974,40 @@ inline condition operator or(condition &&a, condition &&b)
 			auto ci = static_cast<detail::key_equals_condition_impl *>(a.m_impl);
 			auto ce = static_cast<detail::key_is_empty_condition_impl *>(b.m_impl);
 
-			if (ci->m_item_tag == ce->m_item_tag)
+			if (ci->m_item_name == ce->m_item_name)
 				return condition(new detail::key_equals_or_empty_condition_impl(ci));
 		}
-		else if (typeid(*b.m_impl) == typeid(detail::key_equals_condition_impl) and
+		
+		if (typeid(*b.m_impl) == typeid(detail::key_equals_condition_impl) and
 				 typeid(*a.m_impl) == typeid(detail::key_is_empty_condition_impl))
 		{
 			auto ci = static_cast<detail::key_equals_condition_impl *>(b.m_impl);
 			auto ce = static_cast<detail::key_is_empty_condition_impl *>(a.m_impl);
 
-			if (ci->m_item_tag == ce->m_item_tag)
+			if (ci->m_item_name == ce->m_item_name)
 				return condition(new detail::key_equals_or_empty_condition_impl(ci));
 		}
 
+		if (typeid(*a.m_impl) == typeid(detail::key_equals_number_condition_impl) and
+			typeid(*b.m_impl) == typeid(detail::key_is_empty_condition_impl))
+		{
+			auto ci = static_cast<detail::key_equals_number_condition_impl *>(a.m_impl);
+			auto ce = static_cast<detail::key_is_empty_condition_impl *>(b.m_impl);
+
+			if (ci->m_item_name == ce->m_item_name)
+				return condition(new detail::key_equals_number_or_empty_condition_impl(ci));
+		}
+		
+		if (typeid(*b.m_impl) == typeid(detail::key_equals_number_condition_impl) and
+				 typeid(*a.m_impl) == typeid(detail::key_is_empty_condition_impl))
+		{
+			auto ci = static_cast<detail::key_equals_number_condition_impl *>(b.m_impl);
+			auto ce = static_cast<detail::key_is_empty_condition_impl *>(a.m_impl);
+
+			if (ci->m_item_name == ce->m_item_name)
+				return condition(new detail::key_equals_number_or_empty_condition_impl(ci));
+		}
+
 		return condition(new detail::or_condition_impl(std::move(a), std::move(b)));
 	}
 
@@ -732,152 +1017,325 @@ inline condition operator or(condition &&a, condition &&b)
 	return condition(std::move(b));
 }
 
+/**
+ * @brief A helper class to make it possible to search for empty items (NULL)
+ * 
+ * @code{.cpp}
+ * "id"_key == cif::empty_type();
+ * @endcode
+ */
+
 struct empty_type
 {
 };
 
-/// \brief A helper to make it possible to have conditions like ("id"_key == cif::null)
+/**
+ * @brief A helper to make it possible to have conditions like
+ * 
+ * @code{.cpp}
+ * "id"_key == cif::null;
+ * @endcode
+ */
 
 inline constexpr empty_type null = empty_type();
 
+/**
+ * @brief Class to use in creating conditions, creates a reference to a item or item
+ * 
+ */
 struct key
 {
-	explicit key(const std::string &itemTag)
-		: m_item_tag(itemTag)
+	/**
+	 * @brief Construct a new key object using @a item_name as name
+	 * 
+	 * @param item_name 
+	 */
+	explicit key(const std::string &item_name)
+		: m_item_name(item_name)
 	{
 	}
 
-	explicit key(const char *itemTag)
-		: m_item_tag(itemTag)
+	/**
+	 * @brief Construct a new key object using @a item_name as name
+	 * 
+	 * @param item_name 
+	 */
+	explicit key(const char *item_name)
+		: m_item_name(item_name)
+	{
+	}
+
+	/**
+	 * @brief Construct a new key object using @a item_name as name
+	 * 
+	 * @param item_name 
+	 */
+	explicit key(std::string_view item_name)
+		: m_item_name(item_name)
 	{
 	}
 
 	key(const key &) = delete;
 	key &operator=(const key &) = delete;
 
-	std::string m_item_tag;
+	std::string m_item_name; ///< The item name
 };
 
 template <typename T>
+concept Numeric = ((std::is_floating_point_v<T> or std::is_integral_v<T>) and not std::is_same_v<T, bool>);
+
+/**
+ * @brief Operator to create an equals condition based on a key @a key and a numeric value @a v
+ */
+template <Numeric T>
 condition operator==(const key &key, const T &v)
 {
-	return condition(new detail::key_equals_condition_impl({ key.m_item_tag, v }));
+	return condition(new detail::key_equals_number_condition_impl(key.m_item_name, v));
 }
 
-inline condition operator==(const key &key, const char *value)
+/**
+ * @brief Operator to create an equals condition based on a key @a key and a value @a value
+ */
+inline condition operator==(const key &key, std::string_view value)
 {
-	if (value != nullptr and *value != 0)
-		return condition(new detail::key_equals_condition_impl({ key.m_item_tag, value }));
+	if (not value.empty())
+		return condition(new detail::key_equals_condition_impl({ key.m_item_name, value }));
 	else
-		return condition(new detail::key_is_empty_condition_impl(key.m_item_tag));
+		return condition(new detail::key_is_empty_condition_impl(key.m_item_name));
 }
 
-// inline condition_t operator==(const key& key, const detail::ItemReference& v)
-// {
-// 	if (v.empty())
-// 		return condition_t(new detail::key_is_empty_condition_impl(key.m_item_tag));
-// 	else
-// 		return condition_t(new detail::key_compare_condition_impl(key.m_item_tag, [tag = key.m_item_tag, v](const category& c, const row& r, bool icase)
-// 			{ return r[tag].template compare<(v, icase) == 0; }));
-// }
+/**
+ * @brief Operator to create an equals condition based on a key @a key and a value @a value
+ */
+template <typename T>
+	requires std::is_same_v<T, bool>
+inline condition operator==(const key &key, T value)
+{
+	return condition(new detail::key_equals_condition_impl({ key.m_item_name, value ? "y" : "n" }));
+}
 
+/**
+ * @brief Operator to create a not equals condition based on a key @a key and a value @a v
+ */
 template <typename T>
 condition operator!=(const key &key, const T &v)
 {
 	return condition(new detail::not_condition_impl(operator==(key, v)));
 }
 
-inline condition operator!=(const key &key, const char *v)
+/**
+ * @brief Operator to create a not equals condition based on a key @a key and a value @a value
+ */
+inline condition operator!=(const key &key, std::string_view value)
 {
-	std::string value(v ? v : "");
 	return condition(new detail::not_condition_impl(operator==(key, value)));
 }
 
-template <typename T>
+/**
+ * @brief Operator to create a greater than condition based on a key @a key and a value @a v
+ */
+template <Numeric T>
 condition operator>(const key &key, const T &v)
 {
 	std::ostringstream s;
 	s << " > " << v;
 
 	return condition(new detail::key_compare_condition_impl(
-		key.m_item_tag, [tag = key.m_item_tag, v](row_handle r, bool icase)
-		{ return r[tag].template compare<T>(v, icase) > 0; },
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v) > 0; },
 		s.str()));
 }
 
-template <typename T>
+/**
+ * @brief Operator to create a greater than or equals condition based on a key @a key and a value @a v
+ */
+template <Numeric T>
 condition operator>=(const key &key, const T &v)
 {
 	std::ostringstream s;
 	s << " >= " << v;
 
 	return condition(new detail::key_compare_condition_impl(
-		key.m_item_tag, [tag = key.m_item_tag, v](row_handle r, bool icase)
-		{ return r[tag].template compare<T>(v, icase) >= 0; },
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v) >= 0; },
 		s.str()));
 }
 
-template <typename T>
+/**
+ * @brief Operator to create a less than condition based on a key @a key and a value @a v
+ */
+template <Numeric T>
 condition operator<(const key &key, const T &v)
 {
 	std::ostringstream s;
 	s << " < " << v;
 
 	return condition(new detail::key_compare_condition_impl(
-		key.m_item_tag, [tag = key.m_item_tag, v](row_handle r, bool icase)
-		{ return r[tag].template compare<T>(v, icase) < 0; },
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v) < 0; },
 		s.str()));
 }
 
-template <typename T>
+/**
+ * @brief Operator to create a less than or equals condition based on a key @a key and a value @a v
+ */
+template <Numeric T>
 condition operator<=(const key &key, const T &v)
 {
 	std::ostringstream s;
 	s << " <= " << v;
 
 	return condition(new detail::key_compare_condition_impl(
-		key.m_item_tag, [tag = key.m_item_tag, v](row_handle r, bool icase)
-		{ return r[tag].template compare<T>(v, icase) <= 0; },
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v) <= 0; },
 		s.str()));
 }
 
+/**
+ * @brief Operator to create a greater than condition based on a key @a key and a value @a v
+ */
+inline condition operator>(const key &key, std::string_view v)
+{
+	std::ostringstream s;
+	s << " > " << v;
+
+	return condition(new detail::key_compare_condition_impl(
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v, icase) > 0; },
+		s.str()));
+}
+
+/**
+ * @brief Operator to create a greater than or equals condition based on a key @a key and a value @a v
+ */
+inline condition operator>=(const key &key, std::string_view v)
+{
+	std::ostringstream s;
+	s << " >= " << v;
+
+	return condition(new detail::key_compare_condition_impl(
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v, icase) >= 0; },
+		s.str()));
+}
+
+/**
+ * @brief Operator to create a less than condition based on a key @a key and a value @a v
+ */
+inline condition operator<(const key &key, std::string_view v)
+{
+	std::ostringstream s;
+	s << " < " << v;
+
+	return condition(new detail::key_compare_condition_impl(
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v, icase) < 0; },
+		s.str()));
+}
+
+/**
+ * @brief Operator to create a less than or equals condition based on a key @a key and a value @a v
+ */
+inline condition operator<=(const key &key, std::string_view v)
+{
+	std::ostringstream s;
+	s << " <= " << v;
+
+	return condition(new detail::key_compare_condition_impl(
+		key.m_item_name, [item_name = key.m_item_name, v](row_handle r, bool icase)
+		{ return r[item_name].compare(v, icase) <= 0; },
+		s.str()));
+}
+
+/**
+ * @brief Operator to create a condition based on a key @a key and a regular expression @a rx
+ */
 inline condition operator==(const key &key, const std::regex &rx)
 {
-	return condition(new detail::key_matches_condition_impl(key.m_item_tag, rx));
+	return condition(new detail::key_matches_condition_impl(key.m_item_name, rx));
 }
 
+/**
+ * @brief Operator to create a condition based on a key @a key which should be empty/null
+ */
 inline condition operator==(const key &key, const empty_type &)
 {
-	return condition(new detail::key_is_empty_condition_impl(key.m_item_tag));
+	return condition(new detail::key_is_empty_condition_impl(key.m_item_name));
 }
 
+/**
+ * @brief Operator to create a condition based on a key @a key which should be not empty/null
+ */
 inline condition operator!=(const key &key, const empty_type &)
 {
-	return condition(new detail::key_is_not_empty_condition_impl(key.m_item_tag));
+	return condition(new detail::key_is_not_empty_condition_impl(key.m_item_name));
 }
 
+/**
+ * @brief Create a condition to search any item for a value @a v if @a v contains a value
+ * compare to null if not.
+ */
+template <typename T>
+condition operator==(const key &key, const std::optional<T> &v)
+{
+	if (v.has_value())
+		return condition(new detail::key_equals_condition_impl({ key.m_item_name, *v }));
+	else
+		return condition(new detail::key_is_empty_condition_impl(key.m_item_name));
+}
+
+/**
+ * @brief Create a condition to search any item for a value @a v if @a v contains a value
+ * compare to null if not.
+ */
+template <typename T>
+condition operator!=(const key &key, const std::optional<T> &v)
+{
+	if (v.has_value())
+		return condition(new detail::not_condition_impl(condition(new detail::key_equals_condition_impl({ key.m_item_name, *v }))));
+	else
+		return condition(new detail::not_condition_impl(condition(new detail::key_is_empty_condition_impl(key.m_item_name))));
+}
+
+/**
+ * @brief Operator to create a boolean opposite of the condition in @a rhs
+ */
 inline condition operator not(condition &&rhs)
 {
 	return condition(new detail::not_condition_impl(std::move(rhs)));
 }
 
+/** @cond */
 struct any_type
 {
 };
+/** @endcond */
 
+/**
+ * @brief A helper for any item constructs
+ */
 inline constexpr any_type any = any_type{};
 
+/**
+ * @brief Create a condition to search any item for a value @a v
+ */
 template <typename T>
 condition operator==(const any_type &, const T &v)
 {
 	return condition(new detail::any_is_condition_impl<T>(v));
 }
 
+/**
+ * @brief Create a condition to search any item for a regular expression @a rx
+ */
 inline condition operator==(const any_type &, const std::regex &rx)
 {
 	return condition(new detail::any_matches_condition_impl(rx));
 }
 
+/**
+ * @brief Create a condition to return all rows
+ */
 inline condition all()
 {
 	return condition(new detail::all_condition_impl());
@@ -885,7 +1343,14 @@ inline condition all()
 
 namespace literals
 {
-	inline key operator""_key(const char *text, size_t length)
+	/**
+	 * @brief Return a cif::key for the item name @a text
+	 * 
+	 * @param text The name of the item
+	 * @param length The length of @a text
+	 * @return key The cif::key created
+	 */
+	inline key operator""_key(const char *text, std::size_t length)
 	{
 		return key(std::string(text, length));
 	}

include/cif++/datablock.hpp

@@ -29,57 +29,206 @@
 #include "cif++/category.hpp"
 #include "cif++/forward_decl.hpp"
 
+/** \file datablock.hpp
+ * Each valid mmCIF file contains at least one @ref cif::datablock.
+ * A datablock has a name and can contain one or more @ref cif::category "categories"
+ */
+
 namespace cif
 {
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief A datablock is a list of category objects with some additional features
+ * 
+ */
+
 class datablock : public std::list<category>
 {
   public:
 	datablock() = default;
 
+	/**
+	 * @brief Construct a new datablock object with name @a name
+	 * 
+	 * @param name The name for the new datablock
+	 */
 	datablock(std::string_view name)
 		: m_name(name)
 	{
 	}
 
+	/** @cond */
 	datablock(const datablock &);
-	datablock(datablock &&) = default;
 
-	datablock &operator=(const datablock &);
-	datablock &operator=(datablock &&) = default;
+	datablock(datablock &&db) noexcept
+	{
+		swap_(*this, db);
+	}
+
+	datablock &operator=(datablock db)
+	{
+		swap_(*this, db);
+		return *this;
+	}
+	/** @endcond */
+
+	friend void swap_(datablock &a, datablock &b) noexcept
+	{
+		std::swap(a.m_name, b.m_name);
+		std::swap(a.m_validator, b.m_validator);
+		std::swap(static_cast<std::list<category>&>(a), static_cast<std::list<category>&>(b));
+	}
 
 	// --------------------------------------------------------------------
 
+	/**
+	 * @brief Return the name of this datablock
+	 */
 	const std::string &name() const { return m_name; }
 
+	/**
+	 * @brief Set the name of this datablock to @a name
+	 * 
+	 * @param name The new name
+	 */
 	void set_name(std::string_view name)
 	{
 		m_name = name;
 	}
 
+	/**
+	 * @brief Attempt to load the dictionary specified in audit_conform category
+	 * 
+	 */
+	void load_dictionary();
+
+	/**
+	 * @brief Load the dictionary named @a dict_name
+	 * 
+	 * @param dict_name 
+	 */
+	void load_dictionary(std::string_view dict_name);
+
+	/**
+	 * @brief Set the validator object to @a v
+	 * 
+	 * @param v The new validator object, may be null
+	 */
 	void set_validator(const validator *v);
 
+	/**
+	 * @brief Get the validator object
+	 * 
+	 * @return const validator* The validator or nullptr if there is none
+	 */
 	const validator *get_validator() const;
 
+	/**
+	 * @brief Validates the content of this datablock and all its content
+	 * 
+	 * @return true If the content is valid
+	 * @return false If the content is not valid
+	 */
 	bool is_valid() const;
+
+	/**
+	 * @brief Validates the content of this datablock and all its content
+	 * and updates or removes the audit_conform category to match the result.
+	 * 
+	 * @return true If the content is valid
+	 * @return false If the content is not valid
+	 */
+	bool is_valid();
+
+	/**
+	 * @brief Validates all contained data for valid links between parents and children
+	 * as defined in the validator
+	 * 
+	 * @return true If all links are valid
+	 * @return false If all links are not valid
+	 */
 	bool validate_links() const;
 
 	// --------------------------------------------------------------------
 
+	/**
+	 * @brief Return the category named @a name, will create a new and empty
+	 * category named @a name if it does not exist.
+	 * 
+	 * @param name The name of the category to return
+	 * @return category& Reference to the named category
+	 */
 	category &operator[](std::string_view name);
+
+	/**
+	 * @brief Return the const category named @a name, will return a reference
+	 * to a static empty category if it was not found.
+	 * 
+	 * @param name The name of the category to return
+	 * @return category& Reference to the named category
+	 */
 	const category &operator[](std::string_view name) const;
 
+	/**
+	 * @brief Return a pointer to the category named @a name or nullptr if
+	 * it does not exist.
+	 * 
+	 * @param name The name of the category
+	 * @return category* Pointer to the category found or nullptr
+	 */
 	category *get(std::string_view name);
+
+	/**
+	 * @brief Return a pointer to the category named @a name or nullptr if
+	 * it does not exist.
+	 * 
+	 * @param name The name of the category
+	 * @return category* Pointer to the category found or nullptr
+	 */
 	const category *get(std::string_view name) const;
 
+	/**
+	 * @brief Tries to find a category with name @a name and will create a
+	 * new one if it is not found. The result is a tuple of an iterator
+	 * pointing to the category and a boolean indicating whether the category
+	 * was created or not.
+	 * 
+	 * @param name The name for the category
+	 * @return std::tuple<iterator, bool> A tuple containing an iterator pointing
+	 * at the category and a boolean indicating whether the category was newly
+	 * created.
+	 */
 	std::tuple<iterator, bool> emplace(std::string_view name);
 
-	std::vector<std::string> get_tag_order() const;
-	void write(std::ostream &os) const;
-	void write(std::ostream &os, const std::vector<std::string> &tag_order);
+	/**
+	 * @brief Get the preferred order of the categories when writing them
+	 */
+	[[deprecated("use get_item_order instead")]]
+	std::vector<std::string> get_tag_order() const
+	{
+		return get_item_order();
+	}
 
+	/**
+	 * @brief Get the preferred order of the categories when writing them
+	 */
+	std::vector<std::string> get_item_order() const;
+
+	/**
+	 * @brief Write out the contents to @a os
+	 */
+	void write(std::ostream &os) const;
+
+	/**
+	 * @brief Write out the contents to @a os using the order defined in @a item_name_order
+	 */
+	void write(std::ostream &os, const std::vector<std::string> &item_name_order);
+
+	/**
+	 * @brief Friend operator<< to write datablock @a db to std::ostream @a os
+	 */
 	friend std::ostream &operator<<(std::ostream &os, const datablock &db)
 	{
 		db.write(os);
@@ -88,6 +237,9 @@ class datablock : public std::list<category>
 
 	// --------------------------------------------------------------------
 
+	/**
+	 * @brief Comparison operator to compare two datablock for equal content
+	 */
 	bool operator==(const datablock &rhs) const;
 
   private:

include/cif++/dictionary_parser.hpp

@@ -28,10 +28,23 @@
 
 #include "cif++/validate.hpp"
 
+/**
+ * @file validate.hpp
+ * 
+ * Functions to create and manipulate validator objects 
+ */
+
 namespace cif
 {
 
+/**
+ * @brief Parse the contents of @a is and create a new validator object with name @a name
+ */
 validator parse_dictionary(std::string_view name, std::istream &is);
+
+/**
+ * @brief Extend the definitions in validator @a v with the contents of stream @a is
+ */
 void extend_dictionary(validator &v, std::istream &is);
 
 } // namespace cif

include/cif++/file.hpp

@@ -28,35 +28,68 @@
 
 #include <list>
 
-#include "cif++/exports.hpp"
 #include "cif++/datablock.hpp"
 #include "cif++/parser.hpp"
 
+/** \file file.hpp
+ * 
+ * The file class defined here encapsulates the contents of an mmCIF file
+ * It is mainly a list of @ref cif::datablock objects
+ * 
+ * The class file has methods to load dictionaries. These dictionaries are
+ * loaded from resources (if available) or from disk from several locations.
+ * 
+ * See the documentation on load_resource() in file: utilities.hpp for more
+ * information on how data is loaded. 
+ */
+
 namespace cif
 {
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief The class file is actually a list of datablock objects
+ * 
+ */
+
 class file : public std::list<datablock>
 {
   public:
 	file() = default;
 
+	/**
+	 * @brief Construct a new file object using the data in the file @a p as content
+	 * 
+	 * @param p Path to a file containing the data to load
+	 */
 	explicit file(const std::filesystem::path &p)
 	{
 		load(p);
 	}
 
+	/**
+	 * @brief Construct a new file object using the data in the std::istream @a is
+	 * 
+	 * @param is The istream containing the data to load
+	 */
 	explicit file(std::istream &is)
 	{
 		load(is);
 	}
 
-	explicit file(const char *data, size_t length)
+	/**
+	 * @brief Construct a new file object with data in the constant string defined
+	 * by @a data and @a length
+	 * 
+	 * @param data The pointer to the character string with data to load
+	 * @param length The length of the data
+	 */
+	explicit file(const char *data, std::size_t length)
 	{
 		struct membuf : public std::streambuf
 		{
-			membuf(char *text, size_t length)
+			membuf(char *text, std::size_t length)
 			{
 				this->setg(text, text, text + length);
 			}
@@ -66,58 +99,137 @@ class file : public std::list<datablock>
 		load(is);
 	}
 
-	file(const file &) = default;
-	file(file &&) = default;
-	file &operator=(const file &) = default;
-	file &operator=(file &&) = default;
-
-	void set_validator(const validator *v);
-
-	const validator *get_validator() const
+	/** @cond */
+	file(const file &rhs)
+		: std::list<datablock>(rhs)
 	{
-		return m_validator;
 	}
 
+	file(file &&rhs)
+	{
+		this->swap(rhs);
+	}
+
+	file &operator=(file f)
+	{
+		this->swap(f);
+		return *this;
+	}
+
+	/** @endcond */
+
+	/**
+	 * @brief Validate the content and return true if everything was valid.
+	 * 
+	 * Will throw an exception if there is no validator defined.
+	 * 
+	 * If each category was valid, validate_links will also be called.
+	 * 
+	 * @return true If the content is valid
+	 * @return false If the content is not valid
+	 */
 	bool is_valid() const;
+
+	/**
+	 * @brief Validate the content and return true if everything was valid.
+	 * 
+	 * Will attempt to load the referenced dictionary if none was specified.
+	 * 
+	 * If each category was valid, validate_links will also be called.
+	 * 
+	 * @return true If the content is valid
+	 * @return false If the content is not valid
+	 */
 	bool is_valid();
+
+	/**
+	 * @brief Validate the links for all datablocks contained.
+	 * 
+	 * Will throw an exception if no validator was specified.
+	 * 
+	 * @return true If all links were valid
+	 * @return false If all links were not valid
+	 */
 	bool validate_links() const;
 
-	void load_dictionary();
-	void load_dictionary(std::string_view name);
-
+	/**
+	 * @brief Return true if a datablock with the name @a name is part of this file
+	 */
 	bool contains(std::string_view name) const;
 
+	/**
+	 * @brief return a reference to the first datablock in the file
+	 */
 	datablock &front()
 	{
 		assert(not empty());
 		return std::list<datablock>::front();
 	}
 
+	/**
+	 * @brief return a const reference to the first datablock in the file
+	 */
 	const datablock &front() const
 	{
 		assert(not empty());
 		return std::list<datablock>::front();
 	}
 
+	/**
+	 * @brief return a reference to the datablock named @a name
+	 */
 	datablock &operator[](std::string_view name);
+
+	/**
+	 * @brief return a const reference to the datablock named @a name
+	 */
 	const datablock &operator[](std::string_view name) const;
 
+	/**
+	 * @brief Tries to find a datablock with name @a name and will create a
+	 * new one if it is not found. The result is a tuple of an iterator
+	 * pointing to the datablock and a boolean indicating whether the datablock
+	 * was created or not.
+	 * 
+	 * @param name The name for the datablock
+	 * @return std::tuple<iterator, bool> A tuple containing an iterator pointing
+	 * at the datablock and a boolean indicating whether the datablock was newly
+	 * created.
+	 */
 	std::tuple<iterator, bool> emplace(std::string_view name);
 
+	/** Load the data from the file specified by @a p */
 	void load(const std::filesystem::path &p);
+
+	/** Load the data from @a is */
 	void load(std::istream &is);
 
+	/** Load the data from the file specified by @a p using validator @a v */
+	void load(const std::filesystem::path &p, const validator &v);
+
+	/** Load the data from @a is using validator @a v */
+	void load(std::istream &is, const validator &v);
+
+	/** Load the data from the file specified by @a p using a validator constructed from dictionary @a dict */
+	void load(const std::filesystem::path &p, std::string_view dict);
+
+	/** Load the data from @a is using a validator constructed from dictionary @a dict */
+	void load(std::istream &is, std::string_view dict);
+
+	/** Save the data to the file specified by @a p */
 	void save(const std::filesystem::path &p) const;
+
+	/** Save the data to @a is */
 	void save(std::ostream &os) const;
 
+	/**
+	 * @brief Friend operator<< to write file @a f to std::ostream @a os
+	 */
 	friend std::ostream &operator<<(std::ostream &os, const file &f)
 	{
 		f.save(os);
 		return os;
 	}
-
-  private:
-	const validator *m_validator = nullptr;
 };
 
 } // namespace cif

include/cif++/format.hpp

@@ -28,10 +28,12 @@
 
 #include <string>
 
-/// \file format.hpp
-/// File containing a basic reimplementation of boost::format
-/// but then a bit more simplistic. Still this allowed me to move my code
-/// from using boost::format to something without external dependency easily.
+/**  \file format.hpp
+ * 
+ * File containing a basic reimplementation of boost::format
+ * but then a bit more simplistic. Still this allowed me to move my code
+ * from using boost::format to something without external dependency easily.
+ */
 
 namespace cif
 {
@@ -85,6 +87,8 @@ namespace detail
 
 } // namespace
 
+/** @cond */
+
 template <typename... Args>
 class format_plus_arg
 {
@@ -121,7 +125,7 @@ class format_plus_arg
 
   private:
 
-	template <size_t... I>
+	template <std::size_t... I>
 	void copy_vargs(std::index_sequence<I...>)
 	{
 		((std::get<I>(m_vargs) = *std::get<I>(m_args)), ...);
@@ -132,6 +136,23 @@ class format_plus_arg
 	vargs_vector_type m_vargs;
 };
 
+/** @endcond */
+
+/**
+ * @brief A simplistic reimplementation of boost::format, in fact it is
+ * actually a way to call the C function snprintf to format the arguments
+ * in @a args into the format string @a fmt
+ * 
+ * The string in @a fmt should thus be a C style format string.
+ * 
+ * TODO: Move to C++23 style of printing.
+ * 
+ * @tparam Args The types of the arguments
+ * @param fmt The format string
+ * @param args The arguments
+ * @return An object that can be written out to a std::ostream using operator<<
+ */
+
 template <typename... Args>
 constexpr auto format(std::string_view fmt, Args... args)
 {
@@ -144,11 +165,20 @@ constexpr auto format(std::string_view fmt, Args... args)
 class fill_out_streambuf : public std::streambuf
 {
   public:
+
+	/** @cond */
+
 	using base_type = std::streambuf;
 	using int_type = base_type::int_type;
 	using char_type = base_type::char_type;
 	using traits_type = base_type::traits_type;
 
+	/** @endcond */
+
+	/**
+	 * @brief Construct a new fill out streambuf object based on ostream @a os and a
+	 * width to fill out to of @a width
+	 */
 	fill_out_streambuf(std::ostream &os, int width = 80)
 		: m_os(os)
 		, m_upstream(os.rdbuf())
@@ -156,11 +186,21 @@ class fill_out_streambuf : public std::streambuf
 	{
 	}
 
+	/** @cond */
+
 	~fill_out_streambuf()
 	{
 		m_os.rdbuf(m_upstream);
 	}
 
+	/** @endcond */
+
+	/**
+	 * @brief The magic happens here. Write out a couple of spaces when
+	 * the last character to write is a newline to make the line as
+	 * wide as the requested width.
+	 */
+	
 	virtual int_type
 	overflow(int_type ic = traits_type::eof())
 	{
@@ -191,8 +231,10 @@ class fill_out_streambuf : public std::streambuf
 		return result;
 	}
 
+	/** Return the upstream streambuf */
 	std::streambuf *get_upstream() const { return m_upstream; }
 
+	/** Return how many lines have been written */
 	int get_line_count() const { return m_line_count; }
 
   private:

include/cif++/forward_decl.hpp

@@ -26,11 +26,16 @@
 
 #pragma once
 
-#include "cif++/exports.hpp"
-
 #include <string>
 #include <vector>
 
+/**
+ * @file forward_decl.hpp
+ * 
+ * File containing only forward declarations
+ * 
+ */
+
 namespace cif
 {
 

include/cif++/gzio.hpp

@@ -13,28 +13,30 @@
 
 #include <zlib.h>
 
-/// \file gzio.hpp
-///
-/// Single header file for the implementation of stream classes
-/// that can transparently read and write compressed files.
-///
-/// The gzio::istream_buf class sniffs the input and decides whether to use
-/// a decompressor if a signature was recognized.
-///
-/// There's also an ifstream and ofstream class here that can
-/// read and write compressed files. In this case the decission
-/// whether to use a compressions/decompression algorithm is
-/// based on the extension of the \a filename argument.
-
-// This is a stripped down version of the gxrio library from
-// https://github.com/mhekkel/gxrio.git
-// Most notably, the lzma support has been removed since getting
-// that to work in Windows proved to be too much work.
+/** \file gzio.hpp
+ * 
+ * Single header file for the implementation of stream classes
+ * that can transparently read and write compressed files.
+ * 
+ * The gzio::istream_buf class sniffs the input and decides whether to use
+ * a decompressor if a signature was recognized.
+ * 
+ * There's also an ifstream and ofstream class here that can
+ * read and write compressed files. In this case the decission
+ * whether to use a compressions/decompression algorithm is
+ * based on the extension of the \a filename argument.
+ *
+ * This is a stripped down version of the gxrio library from
+ * https://github.com/mhekkel/gxrio.git
+ * Most notably, the lzma support has been removed since getting
+ * that to work in Windows proved to be too much work.
+ */
 
 namespace cif::gzio
 {
 
-const size_t kDefaultBufferSize = 256;
+/** The default buffer size to use */
+const std::size_t kDefaultBufferSize = 256;
 
 // --------------------------------------------------------------------
 
@@ -50,6 +52,8 @@ template <typename CharT, typename Traits>
 class basic_streambuf : public std::basic_streambuf<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	using char_type = CharT;
 	using traits_type = Traits;
 
@@ -77,14 +81,21 @@ class basic_streambuf : public std::basic_streambuf<CharT, Traits>
 		return *this;
 	}
 
+	/** @endcond */
+
+	/** Set the upstream streambuf to @a upstream */
 	void set_upstream(streambuf_type *upstream)
 	{
 		m_upstream = upstream;
 	}
 
+	/** @cond */
+
 	virtual basic_streambuf *init(streambuf_type *sb) = 0;
 	virtual basic_streambuf *close() = 0;
 
+	/** @endcond */
+
   protected:
 	/// \brief The upstream streambuf object, usually this is a basic_filebuf
 	streambuf_type *m_upstream = nullptr;
@@ -101,10 +112,12 @@ class basic_streambuf : public std::basic_streambuf<CharT, Traits>
 /// This implementation of streambuf can decompress (inflate) data compressed
 /// using zlib.
 
-template <typename CharT, typename Traits, size_t BufferSize = kDefaultBufferSize>
+template <typename CharT, typename Traits, std::size_t BufferSize = kDefaultBufferSize>
 class basic_igzip_streambuf : public basic_streambuf<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	static_assert(sizeof(CharT) == 1, "Unfortunately, support for wide characters is not implemented yet.");
 
 	using char_type = CharT;
@@ -121,6 +134,8 @@ class basic_igzip_streambuf : public basic_streambuf<CharT, Traits>
 
 	basic_igzip_streambuf(const basic_igzip_streambuf &) = delete;
 
+	/** @endcond */
+
 	/// \brief Move constructor
 	basic_igzip_streambuf(basic_igzip_streambuf &&rhs)
 		: base_type(std::move(rhs))
@@ -141,6 +156,8 @@ class basic_igzip_streambuf : public basic_streambuf<CharT, Traits>
 		}
 	}
 
+	/** @cond */
+
 	basic_igzip_streambuf &operator=(const basic_igzip_streambuf &) = delete;
 
 	/// \brief Move operator= implementation
@@ -171,6 +188,8 @@ class basic_igzip_streambuf : public basic_streambuf<CharT, Traits>
 		close();
 	}
 
+	/** @endcond */
+
 	/// \brief This closes the zlib stream and sets the get pointers to null.
 	base_type *close() override
 	{
@@ -298,10 +317,12 @@ class basic_igzip_streambuf : public basic_streambuf<CharT, Traits>
 ///
 /// This implementation of streambuf can compress (deflate) data using zlib.
 
-template <typename CharT, typename Traits, size_t BufferSize = kDefaultBufferSize>
+template <typename CharT, typename Traits, std::size_t BufferSize = kDefaultBufferSize>
 class basic_ogzip_streambuf : public basic_streambuf<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	static_assert(sizeof(CharT) == 1, "Unfortunately, support for wide characters is not implemented yet.");
 
 	using char_type = CharT;
@@ -332,6 +353,8 @@ class basic_ogzip_streambuf : public basic_streambuf<CharT, Traits>
 
 	basic_ogzip_streambuf &operator=(const basic_ogzip_streambuf &) = delete;
 
+	/** @endcond */
+
 	/// \brief Move operator=
 	basic_ogzip_streambuf &operator=(basic_ogzip_streambuf &&rhs)
 	{
@@ -487,6 +510,8 @@ template <typename CharT, typename Traits>
 class basic_istream : public std::basic_istream<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	using base_type = std::basic_istream<CharT, Traits>;
 
 	using traits_type = Traits;
@@ -498,6 +523,8 @@ class basic_istream : public std::basic_istream<CharT, Traits>
 
 	using gzip_streambuf_type = basic_igzip_streambuf<char_type, traits_type>;
 
+	/** @endcond */
+
 	/// \brief Regular move constructor
 	basic_istream(basic_istream &&rhs)
 		: base_type(std::move(rhs))
@@ -591,6 +618,8 @@ template <typename CharT, typename Traits>
 class basic_ifstream : public basic_istream<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	using base_type = basic_istream<CharT, Traits>;
 
 	using char_type = CharT;
@@ -608,6 +637,8 @@ class basic_ifstream : public basic_istream<CharT, Traits>
 		close();
 	}
 
+	/** @endcond */
+
 	/// \brief Construct an ifstream
 	/// \param filename Null terminated string specifying the file to open
 	/// \param mode The mode in which to open the file
@@ -647,10 +678,13 @@ class basic_ifstream : public basic_istream<CharT, Traits>
 			this->rdbuf(&m_filebuf);
 	}
 
+	/** @cond */
 	basic_ifstream(const basic_ifstream &) = delete;
 
 	basic_ifstream &operator=(const basic_ifstream &) = delete;
 
+	/** @endcond */
+
 	/// \brief Move version of operator=
 	basic_ifstream &operator=(basic_ifstream &&rhs)
 	{
@@ -777,6 +811,8 @@ template <typename CharT, typename Traits>
 class basic_ostream : public std::basic_ostream<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	using base_type = std::basic_ostream<CharT, Traits>;
 
 	using char_type = CharT;
@@ -785,6 +821,8 @@ class basic_ostream : public std::basic_ostream<CharT, Traits>
 	using z_streambuf_type = basic_streambuf<char_type, traits_type>;
 	using upstreambuf_type = std::basic_streambuf<char_type, traits_type>;
 
+	/** @endcond */
+
 	/// \brief Regular move constructor
 	basic_ostream(basic_ostream &&rhs)
 		: base_type(std::move(rhs))
@@ -846,6 +884,8 @@ template <typename CharT, typename Traits>
 class basic_ofstream : public basic_ostream<CharT, Traits>
 {
   public:
+	/** @cond */
+
 	using base_type = basic_ostream<CharT, Traits>;
 
 	using char_type = CharT;
@@ -861,6 +901,8 @@ class basic_ofstream : public basic_ostream<CharT, Traits>
 		close();
 	}
 
+	/** @endcond */
+
 	/// \brief Construct an ofstream
 	/// \param filename Null terminated string specifying the file to open
 	/// \param mode The mode in which to open the file
@@ -899,10 +941,14 @@ class basic_ofstream : public basic_ostream<CharT, Traits>
 			this->rdbuf(&m_filebuf);
 	}
 
+	/** @cond */
+
 	basic_ofstream(const basic_ofstream &) = delete;
 
 	basic_ofstream &operator=(const basic_ofstream &) = delete;
 
+	/** @endcond */
+
 	/// \brief Move operator=
 	basic_ofstream &operator=(basic_ofstream &&rhs)
 	{
@@ -1032,11 +1078,13 @@ class basic_ofstream : public basic_ostream<CharT, Traits>
 
 // --------------------------------------------------------------------
 
-/// \brief Convenience typedefs
+/// \brief Convenience typedef for a regular istream
 using istream = basic_istream<char, std::char_traits<char>>;
+
+/// \brief Convenience typedef for a file ifstream
 using ifstream = basic_ifstream<char, std::char_traits<char>>;
 
-// using ostream = basic_ostream<char, std::char_traits<char>>;
+/// \brief Convenience typedef for a file ofstream
 using ofstream = basic_ofstream<char, std::char_traits<char>>;
 
 } // namespace gzio

include/cif++/item.hpp

@@ -41,22 +41,61 @@
 #include <optional>
 #include <utility>
 
-/// \file item.hpp
-/// This file contains the declaration of item but also the item_value and item_handle
-/// These handle the storage of and access to the data for a single data field. 
+/** \file item.hpp
+ *
+ * This file contains the declaration of item but also the item_value and item_handle
+ * These handle the storage of and access to the data for a single data item.
+ */
 
 namespace cif
 {
 
 // --------------------------------------------------------------------
-/// \brief item is a transient class that is used to pass data into rows
-///        but it also takes care of formatting data. 
+/** @brief item is a transient class that is used to pass data into rows
+ * but it also takes care of formatting data.
+ * 
+ * 
+ * 
+ * The class cif::item is often used implicitly when creating a row in a category
+ * using the emplace function.
+ * 
+ * @code{.cpp}
+ * cif::category cat("my-cat");
+ * cat.emplace({
+ *   { "item-1", 1 },                             // <- stores an item with value 1
+ *   { "item-2", 1.0, 2 },                        // <- stores an item with value 1.00
+ *   { "item-3", std::optional<int>() },          // <- stores an item with value ?
+ *   { "item-4", std::make_optional<int>(42) },   // <- stores an item with value 42
+ *   { "item-5" }                                 // <- stores an item with value .
+ * });
+ * 
+ * std::cout << cat << '\n';
+ * @endcode
+ * 
+ * Will result in:
+ * 
+ * @code{.txt}
+ * _my-cat.item-1 1
+ * _my-cat.item-2 1.00
+ * _my-cat.item-3 ?
+ * _my-cat.item-4 42
+ * _my-cat.item-5 .
+ * @endcode
+ */
 class item
 {
   public:
 	/// \brief Default constructor, empty item
 	item() = default;
 
+	/// \brief constructor for an item with name \a name and as
+	/// content the character '.', i.e. an inapplicable value.
+	item(std::string_view name)
+		: m_name(name)
+		, m_value({ '.' })
+	{
+	}
+
 	/// \brief constructor for an item with name \a name and as
 	/// content a single character string with content \a value
 	item(std::string_view name, char value)
@@ -66,7 +105,7 @@ class item
 	}
 
 	/// \brief constructor for an item with name \a name and as
-	/// content a the formatted floating point value \a value with
+	/// content the formatted floating point value \a value with
 	/// precision \a precision
 	template <typename T, std::enable_if_t<std::is_floating_point_v<T>, int> = 0>
 	item(std::string_view name, const T &value, int precision)
@@ -78,11 +117,9 @@ class item
 		char buffer[32];
 
 		auto r = to_chars(buffer, buffer + sizeof(buffer) - 1, value, chars_format::fixed, precision);
-		if (r.ec != std::errc())
+		if ((bool)r.ec)
 			throw std::runtime_error("Could not format number");
 
-		assert(r.ptr >= buffer and r.ptr < buffer + sizeof(buffer));
-		*r.ptr = 0;
 		m_value.assign(buffer, r.ptr - buffer);
 	}
 
@@ -99,34 +136,30 @@ class item
 		char buffer[32];
 
 		auto r = to_chars(buffer, buffer + sizeof(buffer) - 1, value, chars_format::general);
-		if (r.ec != std::errc())
+		if ((bool)r.ec)
 			throw std::runtime_error("Could not format number");
 
-		assert(r.ptr >= buffer and r.ptr < buffer + sizeof(buffer));
-		*r.ptr = 0;
 		m_value.assign(buffer, r.ptr - buffer);
 	}
 
 	/// \brief constructor for an item with name \a name and as
-	/// content a the formatted integral value \a value
-	template <typename T, std::enable_if_t<std::is_integral_v<T> and not std::is_same_v<T,bool>, int> = 0>
+	/// content the formatted integral value \a value
+	template <typename T, std::enable_if_t<std::is_integral_v<T> and not std::is_same_v<T, bool>, int> = 0>
 	item(const std::string_view name, const T &value)
 		: m_name(name)
 	{
 		char buffer[32];
 
 		auto r = std::to_chars(buffer, buffer + sizeof(buffer) - 1, value);
-		if (r.ec != std::errc())
+		if ((bool)r.ec)
 			throw std::runtime_error("Could not format number");
 
-		assert(r.ptr >= buffer and r.ptr < buffer + sizeof(buffer));
-		*r.ptr = 0;
 		m_value.assign(buffer, r.ptr - buffer);
 	}
 
 	/// \brief constructor for an item with name \a name and as
-	/// content a the formatted boolean value \a value
-	template <typename T, std::enable_if_t<std::is_same_v<T,bool>, int> = 0>
+	/// content the formatted boolean value \a value
+	template <typename T, std::enable_if_t<std::is_same_v<T, bool>, int> = 0>
 	item(const std::string_view name, const T &value)
 		: m_name(name)
 	{
@@ -135,22 +168,62 @@ class item
 
 	/// \brief constructor for an item with name \a name and as
 	/// content value \a value
-	item(const std::string_view name, const std::string_view value)
+	item(const std::string_view name, std::string_view value)
 		: m_name(name)
 		, m_value(value)
 	{
 	}
 
+	/// \brief constructor for an item with name \a name and as
+	/// content value \a value
+	template<typename T, std::enable_if_t<std::is_same_v<T, std::string>, int> = 0>
+	item(const std::string_view name, T &&value)
+		: m_name(name)
+		, m_value(std::move(value))
+	{
+	}
+
+	/// \brief constructor for an item with name \a name and as
+	/// content the optional value \a value
+	template <typename T>
+	item(const std::string_view name, const std::optional<T> &value)
+		: m_name(name)
+	{
+		if (value.has_value())
+		{
+			item tmp(name, *value);
+			std::swap(tmp.m_value, m_value);
+		}
+		else
+			m_value.assign("?");
+	}
+
+	/// \brief constructor for an item with name \a name and as
+	/// content the formatted floating point value \a value with
+	/// precision \a precision
+	template <typename T, std::enable_if_t<std::is_floating_point_v<T>, int> = 0>
+	item(std::string_view name, const std::optional<T> &value, int precision)
+		: m_name(name)
+	{
+		if (value.has_value())
+		{
+			item tmp(name, *value, precision);
+			std::swap(tmp.m_value, m_value);
+		}
+		else
+			m_value.assign("?");
+	}
+
+	/** @cond */
 	item(const item &rhs) = default;
-
 	item(item &&rhs) noexcept = default;
-
 	item &operator=(const item &rhs) = default;
-
 	item &operator=(item &&rhs) noexcept = default;
+	/** @endcond */
 
-	std::string_view name() const { return m_name; }
-	std::string_view value() const { return m_value; }
+	std::string_view name() const { return m_name; }   ///< Return the name of the item
+	std::string_view value() const & { return m_value; } ///< Return the value of the item
+	std::string value() const && { return std::move(m_value); } ///< Return the value of the item
 
 	/// \brief replace the content of the stored value with \a v
 	void value(std::string_view v) { m_value = v; }
@@ -158,21 +231,23 @@ class item
 	/// \brief empty means either null or unknown
 	bool empty() const { return m_value.empty(); }
 
-	/// \brief returns true if the field contains '.'
+	/// \brief returns true if the item contains '.'
 	bool is_null() const { return m_value == "."; }
 
-	/// \brief returns true if the field contains '?'
+	/// \brief returns true if the item contains '?'
 	bool is_unknown() const { return m_value == "?"; }
 
 	/// \brief the length of the value string
-	size_t length() const { return m_value.length(); }
+	std::size_t length() const { return m_value.length(); }
 
 	/// \brief support for structured binding
-	template<size_t N>
+	template <std::size_t N>
 	decltype(auto) get() const
 	{
-		     if constexpr (N == 0) return name();
-		else if constexpr (N == 1) return value();
+		if constexpr (N == 0)
+			return name();
+		else if constexpr (N == 1)
+			return value();
 	}
 
   private:
@@ -190,7 +265,9 @@ class item
 
 struct item_value
 {
+	/** @cond */
 	item_value() = default;
+	/** @endcond */
 
 	/// \brief constructor
 	item_value(std::string_view text)
@@ -210,19 +287,17 @@ struct item_value
 		}
 	}
 
-	item_value(item_value &&rhs)
+	/** @cond */
+	item_value(item_value &&rhs) noexcept
 		: m_length(std::exchange(rhs.m_length, 0))
 		, m_storage(std::exchange(rhs.m_storage, 0))
 	{
 	}
 
-	item_value &operator=(item_value &&rhs)
+	item_value &operator=(item_value &&rhs) noexcept
 	{
-		if (this != &rhs)
-		{
-			m_length = std::exchange(rhs.m_length, m_length);
-			m_storage = std::exchange(rhs.m_storage, m_storage);
-		}
+		std::swap(m_length, rhs.m_length);
+		std::swap(m_storage, rhs.m_storage);
 		return *this;
 	}
 
@@ -236,24 +311,29 @@ struct item_value
 
 	item_value(const item_value &) = delete;
 	item_value &operator=(const item_value &) = delete;
+	/** @endcond */
 
+	/** operator bool, allows easy checking for empty items */
 	explicit operator bool() const
 	{
 		return m_length != 0;
 	}
 
-	size_t m_length = 0;
+	std::size_t m_length = 0; ///< Length of the data
 	union
 	{
-		char m_local_data[8];
-		char *m_data;
-		uint64_t m_storage;
+		char m_local_data[8]; ///< Storage area for small strings (strings smaller than kBufferSize)
+		char *m_data;         ///< Pointer to a string stored in the heap
+		uint64_t m_storage;   ///< Alternative storage of the data, used in move operations
 	};
 
-	static constexpr size_t kBufferSize = sizeof(m_local_data);
+	/** The maximum length of locally stored strings */
+	static constexpr std::size_t kBufferSize = sizeof(m_local_data);
 
 	// By using std::string_view instead of c_str we obain a
 	// nice performance gain since we avoid many calls to strlen.
+
+	/** Return the content of the item as a std::string_view */
 	constexpr inline std::string_view text() const
 	{
 		return { m_length >= kBufferSize ? m_data : m_local_data, m_length };
@@ -268,18 +348,68 @@ struct item_value
 struct item_handle
 {
   public:
+	/** @cond */
 	// conversion helper class
 	template <typename T, typename = void>
 	struct item_value_as;
+	/** @endcond */
 
+	/**
+	 * @brief Assign value @a value to the item referenced
+	 *
+	 * @tparam T Type of the value
+	 * @param value The value
+	 * @return reference to this item_handle
+	 */
 	template <typename T>
 	item_handle &operator=(const T &value)
 	{
-		item v{ "", value };
-		assign_value(v);
+		assign_value(item{ "", value }.value());
 		return *this;
 	}
 
+	/**
+	 * @brief Assign value @a value to the item referenced
+	 *
+	 * @tparam T Type of the value
+	 * @param value The value
+	 * @return reference to this item_handle
+	 */
+	template <typename T>
+	item_handle &operator=(T &&value)
+	{
+		assign_value(item{ "", std::forward<T>(value) }.value());
+		return *this;
+	}
+
+	/**
+	 * @brief Assign value @a value to the item referenced
+	 *
+	 * @tparam T Type of the value
+	 * @param value The value
+	 * @return reference to this item_handle
+	 */
+	template <std::size_t N>
+	item_handle &operator=(const char (&value)[N])
+	{
+		assign_value(item{ "", std::move(value) }.value());
+		return *this;
+	}
+
+	/**
+	 * @brief A method with a variable number of arguments that will be concatenated and
+	 * assigned as a string. Use it like this:
+	 *
+	 * @code{.cpp}
+	 * cif::item_handle ih;
+	 * is.os("The result of ", 1, " * ", 42, " is of course ", 42);
+	 * @endcode
+	 *
+	 * And the content will then be `The result of 1 * 42 is of course 42`.
+	 *
+	 * @tparam Ts Types of the parameters
+	 * @param v The parameters to concatenate
+	 */
 	template <typename... Ts>
 	void os(const Ts &...v)
 	{
@@ -288,8 +418,10 @@ struct item_handle
 		this->operator=(ss.str());
 	}
 
+	/** Swap contents of this and @a b */
 	void swap(item_handle &b);
 
+	/** Return the contents of this item as type @tparam T */
 	template <typename T = std::string>
 	auto as() const -> T
 	{
@@ -297,18 +429,36 @@ struct item_handle
 		return item_value_as<value_type>::convert(*this);
 	}
 
+	/** Return the contents of this item as type @tparam T or, if not
+	 * set, use @a dv as the default value.
+	 */
 	template <typename T>
 	auto value_or(const T &dv) const
 	{
 		return empty() ? dv : this->as<T>();
 	}
 
+	/**
+	 * @brief Compare the contents of this item with value @a value
+	 * optionally ignoring character case, if @a icase is true.
+	 * Returns 0 if both are equal, -1 if this sorts before @a value
+	 * and 1 if this sorts after @a value
+	 *
+	 * @tparam T Type of the value @a value
+	 * @param value The value to compare with
+	 * @param icase Flag indicating if we should compare character case sensitive
+	 * @return -1, 0 or 1
+	 */
 	template <typename T>
 	int compare(const T &value, bool icase = true) const
 	{
 		return item_value_as<T>::compare(*this, value, icase);
 	}
 
+	/**
+	 * @brief Compare the value contained with the value @a value and
+	 * return true if both are equal.
+	 */
 	template <typename T>
 	bool operator==(const T &value) const
 	{
@@ -317,45 +467,64 @@ struct item_handle
 	}
 
 	// We may not have C++20 yet...
+
+	/**
+	 * @brief Compare the value contained with the value @a value and
+	 * return true if both are not equal.
+	 */
 	template <typename T>
 	bool operator!=(const T &value) const
 	{
 		return not operator==(value);
 	}
 
-	// empty means either null or unknown
+	/**
+	 * @brief Returns true if the content string is empty or
+	 * only contains '.' meaning null or '?' meaning unknown
+	 * in a mmCIF context
+	 */
 	bool empty() const
 	{
 		auto txt = text();
 		return txt.empty() or (txt.length() == 1 and (txt.front() == '.' or txt.front() == '?'));
 	}
 
+	/** Easy way to test for an empty item */
 	explicit operator bool() const { return not empty(); }
 
-	// is_null means the field contains '.'
+	/// is_null return true if the item contains '.'
 	bool is_null() const
 	{
 		auto txt = text();
 		return txt.length() == 1 and txt.front() == '.';
 	}
 
-	// is_unknown means the field contains '?'
+	/// is_unknown returns true if the item contains '?'
 	bool is_unknown() const
 	{
 		auto txt = text();
 		return txt.length() == 1 and txt.front() == '?';
 	}
 
+	/** Return a std::string_view for the contents */
 	std::string_view text() const;
 
-	item_handle(uint16_t column, row_handle &row)
-		: m_column(column)
+	/**
+	 * @brief Construct a new item handle object
+	 *
+	 * @param item Item index
+	 * @param row Reference to the row
+	 */
+	item_handle(uint16_t item, row_handle &row)
+		: m_item_ix(item)
 		, m_row_handle(row)
 	{
 	}
 
-	static CIFPP_EXPORT const item_handle s_null_item;
+	/** A variable holding an empty item */
+	CIFPP_EXPORT static const item_handle s_null_item;
 
+	/** friend to swap two item handles */
 	friend void swap(item_handle a, item_handle b)
 	{
 		a.swap(b);
@@ -364,14 +533,15 @@ struct item_handle
   private:
 	item_handle();
 
-	uint16_t m_column;
+	uint16_t m_item_ix;
 	row_handle &m_row_handle;
 
-	void assign_value(const item &value);
+	void assign_value(std::string_view value);
 };
 
 // So sad that older gcc implementations of from_chars did not support floats yet...
 
+/** @cond */
 template <typename T>
 struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> and not std::is_same_v<T, bool>>>
 {
@@ -385,17 +555,22 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> an
 		{
 			auto txt = ref.text();
 
-			std::from_chars_result r = selected_charconv<value_type>::from_chars(txt.data(), txt.data() + txt.size(), result);
+			auto b = txt.data();
+			auto e = txt.data() + txt.size();
 
-			if (r.ec != std::errc())
+			std::from_chars_result r = (b + 1 < e and *b == '+' and std::isdigit(b[1])) ? selected_charconv<value_type>::from_chars(b + 1, e, result) : selected_charconv<value_type>::from_chars(b, e, result);
+
+			if ((bool)r.ec or r.ptr != e)
 			{
 				result = {};
 				if (cif::VERBOSE)
 				{
 					if (r.ec == std::errc::invalid_argument)
-						std::cerr << "Attempt to convert " << std::quoted(txt) << " into a number" << std::endl;
+						std::cerr << "Attempt to convert " << std::quoted(txt) << " into a number\n";
 					else if (r.ec == std::errc::result_out_of_range)
-						std::cerr << "Conversion of " << std::quoted(txt) << " into a type that is too small" << std::endl;
+						std::cerr << "Conversion of " << std::quoted(txt) << " into a type that is too small\n";
+					else
+						std::cerr << "Not a valid number " << std::quoted(txt) << '\n';
 				}
 			}
 		}
@@ -409,25 +584,32 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> an
 
 		auto txt = ref.text();
 
-		if (txt.empty())
+		if (ref.empty())
 			result = 1;
 		else
 		{
 			value_type v = {};
 
-			std::from_chars_result r = selected_charconv<value_type>::from_chars(txt.data(), txt.data() + txt.size(), v);
+			auto b = txt.data();
+			auto e = txt.data() + txt.size();
 
-			if (r.ec != std::errc())
+			std::from_chars_result r = (b + 1 < e and *b == '+' and std::isdigit(b[1])) ? selected_charconv<value_type>::from_chars(b + 1, e, v) : selected_charconv<value_type>::from_chars(b, e, v);
+
+			if ((bool)r.ec or r.ptr != e)
 			{
 				if (cif::VERBOSE)
 				{
 					if (r.ec == std::errc::invalid_argument)
-						std::cerr << "Attempt to convert " << std::quoted(txt) << " into a number" << std::endl;
+						std::cerr << "Attempt to convert " << std::quoted(txt) << " into a number\n";
 					else if (r.ec == std::errc::result_out_of_range)
-						std::cerr << "Conversion of " << std::quoted(txt) << " into a type that is too small" << std::endl;
+						std::cerr << "Conversion of " << std::quoted(txt) << " into a type that is too small\n";
+					else
+						std::cerr << "Not a valid number " << std::quoted(txt) << '\n';
 				}
 				result = 1;
 			}
+			else if (std::abs(v - value) <= std::numeric_limits<value_type>::epsilon())
+				result = 0;
 			else if (v < value)
 				result = -1;
 			else if (v > value)
@@ -482,7 +664,7 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_same_v<T, bool>>>
 	}
 };
 
-template <size_t N>
+template <std::size_t N>
 struct item_handle::item_value_as<char[N]>
 {
 	static std::string convert(const item_handle &ref)
@@ -546,22 +728,33 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_same_v<T, std::str
 	}
 };
 
+/** @endcond */
+
 } // namespace cif
 
 namespace std
 {
 
-template<> struct tuple_size<::cif::item>
-            : public std::integral_constant<std::size_t, 2> {};
+/** @cond */
 
-template<> struct tuple_element<0, ::cif::item>
+template <>
+struct tuple_size<::cif::item>
+	: public std::integral_constant<std::size_t, 2>
+{
+};
+
+template <>
+struct tuple_element<0, ::cif::item>
 {
 	using type = decltype(std::declval<::cif::item>().name());
 };
 
-template<> struct tuple_element<1, ::cif::item>
+template <>
+struct tuple_element<1, ::cif::item>
 {
 	using type = decltype(std::declval<::cif::item>().value());
 };
 
-}
+/** @endcond */
+
+} // namespace std

include/cif++/iterator.hpp

@@ -30,22 +30,46 @@
 
 #include <array>
 
+/**
+ * @file iterator.hpp
+ *
+ * This file contains several implementations of generic iterators.
+ *
+ * Using partial specialization we can have implementation for
+ * iterators that return row_handles, a single value or tuples of
+ * multiple values.
+ *
+ */
+
 namespace cif
 {
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief Implementation of an iterator that can return
+ * multiple values in a tuple. Of course, that tuple can
+ * then be used in structured binding to receive the values
+ * in a for loop e.g.
+ *
+ * @tparam Category The category for this iterator
+ * @tparam Ts The types this iterator can be dereferenced to
+ */
 template <typename Category, typename... Ts>
 class iterator_impl
 {
   public:
+	/** @cond */
 	template <typename, typename...>
 	friend class iterator_impl;
 
 	friend class category;
+	/** @endcond */
 
-	static constexpr size_t N = sizeof...(Ts);
+	/** variable that contains the number of elements in the tuple */
+	static constexpr std::size_t N = sizeof...(Ts);
 
+	/** @cond */
 	using category_type = std::remove_cv_t<Category>;
 	using row_type = std::conditional_t<std::is_const_v<Category>, const row, row>;
 
@@ -60,41 +84,38 @@ class iterator_impl
 	iterator_impl() = default;
 
 	iterator_impl(const iterator_impl &rhs) = default;
+	iterator_impl(iterator_impl &&rhs) = default;
 
 	template <typename C2, typename... T2s>
 	iterator_impl(const iterator_impl<C2, T2s...> &rhs)
-		: m_category(rhs.m_category)
-		, m_current(rhs.m_current)
+		: m_current(const_cast<row_handle&>(rhs.m_current))
 		, m_value(rhs.m_value)
-		, m_column_ix(rhs.m_column_ix)
+		, m_item_ix(rhs.m_item_ix)
 	{
 	}
 
 	template <typename IRowType>
 	iterator_impl(iterator_impl<IRowType, Ts...> &rhs)
-		: m_category(rhs.m_category)
-		, m_current(const_cast<row_type *>(rhs.m_current))
+		: m_current(const_cast<row_handle&>(rhs.m_current))
 		, m_value(rhs.m_value)
-		, m_column_ix(rhs.m_column_ix)
+		, m_item_ix(rhs.m_item_ix)
 	{
 		m_value = get(std::make_index_sequence<N>());
 	}
 
 	template <typename IRowType>
 	iterator_impl(const iterator_impl<IRowType> &rhs, const std::array<uint16_t, N> &cix)
-		: m_category(rhs.m_category)
-		, m_current(rhs.m_current)
-		, m_column_ix(cix)
+		: m_current(const_cast<row_handle&>(rhs.m_current))
+		, m_item_ix(cix)
 	{
 		m_value = get(std::make_index_sequence<N>());
 	}
 
-	iterator_impl &operator=(const iterator_impl &i)
+	iterator_impl &operator=(iterator_impl i)
 	{
-		m_category = i.m_category;
-		m_current = i.m_current;
-		m_column_ix = i.m_column_ix;
-		m_value = i.m_value;
+		std::swap(m_current, i.m_current);
+		std::swap(m_item_ix, i.m_item_ix);
+		std::swap(m_value, i.m_value);
 		return *this;
 	}
 
@@ -112,18 +133,18 @@ class iterator_impl
 
 	operator const row_handle() const
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	operator row_handle()
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	iterator_impl &operator++()
 	{
-		if (m_current != nullptr)
-			m_current = m_current->m_next;
+		if (m_current)
+			m_current.m_row = m_current.m_row->m_next;
 
 		m_value = get(std::make_index_sequence<N>());
 
@@ -152,29 +173,32 @@ class iterator_impl
 		return m_current != rhs.m_current;
 	}
 
+	/** @endcond */
+
   private:
-	template <size_t... Is>
+	template <std::size_t... Is>
 	tuple_type get(std::index_sequence<Is...>) const
 	{
-		if (m_current != nullptr)
-		{
-			row_handle rh{*m_category, *m_current};
-			return tuple_type{rh[m_column_ix[Is]].template as<Ts>()...};
-		}
-
-		return {};
+		return m_current ? tuple_type{ m_current[m_item_ix[Is]].template as<Ts>()... } : tuple_type{};
 	}
 
-	category_type *m_category = nullptr;
-	row_type *m_current = nullptr;
+	row_handle m_current;
 	value_type m_value;
-	std::array<uint16_t, N> m_column_ix;
+	std::array<uint16_t, N> m_item_ix;
 };
 
-template<typename Category>
+/**
+ * @brief Implementation of an iterator that returns
+ * only row_handles
+ *
+ * @tparam Category The category for this iterator
+ */
+template <typename Category>
 class iterator_impl<Category>
 {
   public:
+	/** @cond */
+
 	template <typename, typename...>
 	friend class iterator_impl;
 
@@ -185,37 +209,34 @@ class iterator_impl<Category>
 	using iterator_category = std::forward_iterator_tag;
 	using value_type = row_handle;
 	using difference_type = std::ptrdiff_t;
-	using pointer = row_handle;
-	using reference = row_handle;
+	using pointer = value_type *;
+	using reference = value_type &;
 
 	iterator_impl() = default;
 
 	iterator_impl(const iterator_impl &rhs) = default;
+	iterator_impl(iterator_impl &&rhs) = default;
 
 	template <typename C2>
 	iterator_impl(const iterator_impl<C2> &rhs)
-		: m_category(rhs.m_category)
-		, m_current(const_cast<row_type*>(rhs.m_current))
+		: m_current(const_cast<row_handle &>(rhs.m_current))
 	{
 	}
 
 	iterator_impl(Category &cat, row *current)
-		: m_category(const_cast<category_type *>(&cat))
-		, m_current(current)
+		: m_current(cat, *current)
 	{
 	}
 
 	template <typename IRowType>
 	iterator_impl(const iterator_impl<IRowType> &rhs, const std::array<uint16_t, 0> &)
-		: m_category(rhs.m_category)
-		, m_current(rhs.m_current)
+		: m_current(const_cast<row_handle &>(rhs.m_current))
 	{
 	}
 
-	iterator_impl &operator=(const iterator_impl &i)
+	iterator_impl &operator=(iterator_impl i)
 	{
-		m_category = i.m_category;
-		m_current = i.m_current;
+		std::swap(m_current, i.m_current);
 		return *this;
 	}
 
@@ -223,7 +244,7 @@ class iterator_impl<Category>
 
 	reference operator*()
 	{
-		return {*m_category, *m_current};
+		return m_current;
 	}
 
 	pointer operator->()
@@ -233,18 +254,18 @@ class iterator_impl<Category>
 
 	operator const row_handle() const
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	operator row_handle()
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	iterator_impl &operator++()
 	{
-		if (m_current != nullptr)
-			m_current = m_current->m_next;
+		if (m_current)
+			m_current.m_row = m_current.m_row->m_next;
 
 		return *this;
 	}
@@ -271,16 +292,25 @@ class iterator_impl<Category>
 		return m_current != rhs.m_current;
 	}
 
+	/** @endcond */
+
   private:
-	category_type *m_category = nullptr;
-	row_type *m_current = nullptr;
+	row_handle m_current;
 };
 
+/**
+ * @brief Implementation of an iterator that can return
+ * a single value.
+ *
+ * @tparam Category The category for this iterator
+ * @tparam T The type this iterator can be dereferenced to
+ */
 
-template<typename Category, typename T>
+template <typename Category, typename T>
 class iterator_impl<Category, T>
 {
   public:
+	/** @cond */
 	template <typename, typename...>
 	friend class iterator_impl;
 
@@ -298,41 +328,38 @@ class iterator_impl<Category, T>
 	iterator_impl() = default;
 
 	iterator_impl(const iterator_impl &rhs) = default;
+	iterator_impl(iterator_impl &&rhs) = default;
 
 	template <typename C2, typename T2>
 	iterator_impl(const iterator_impl<C2, T2> &rhs)
-		: m_category(rhs.m_category)
-		, m_current(rhs.m_current)
+		: m_current(rhs.m_current)
 		, m_value(rhs.m_value)
-		, m_column_ix(rhs.m_column_ix)
+		, m_item_ix(rhs.m_item_ix)
 	{
 	}
 
 	template <typename IRowType>
 	iterator_impl(iterator_impl<IRowType, T> &rhs)
-		: m_category(rhs.m_category)
-		, m_current(const_cast<row_type *>(rhs.m_current))
+		: m_current(const_cast<row_handle&>(rhs.m_current))
 		, m_value(rhs.m_value)
-		, m_column_ix(rhs.m_column_ix)
-	{
-		m_value = get(m_current);
-	}
-
-	template <typename IRowType>
-	iterator_impl(const iterator_impl<IRowType> &rhs, const std::array<uint16_t, 1> &cix)
-		: m_category(rhs.m_category)
-		, m_current(rhs.m_current)
-		, m_column_ix(cix[0])
+		, m_item_ix(rhs.m_item_ix)
 	{
 		m_value = get();
 	}
 
-	iterator_impl &operator=(const iterator_impl &i)
+	template <typename IRowType>
+	iterator_impl(const iterator_impl<IRowType> &rhs, const std::array<uint16_t, 1> &cix)
+		: m_current(const_cast<row_handle&>(rhs.m_current))
+		, m_item_ix(cix[0])
 	{
-		m_category = i.m_category;
-		m_current = i.m_current;
-		m_column_ix = i.m_column_ix;
-		m_value = i.m_value;
+		m_value = get();
+	}
+
+	iterator_impl &operator=(iterator_impl i)
+	{
+		std::swap(m_current, i.m_current);
+		std::swap(m_item_ix, i.m_item_ix);
+		std::swap(m_value, i.m_value);
 		return *this;
 	}
 
@@ -350,18 +377,18 @@ class iterator_impl<Category, T>
 
 	operator const row_handle() const
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	operator row_handle()
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	iterator_impl &operator++()
 	{
-		if (m_current != nullptr)
-			m_current = m_current->m_next;
+		if (m_current)
+			m_current.m_row = m_current.m_row->m_next;
 
 		m_value = get();
 
@@ -390,32 +417,40 @@ class iterator_impl<Category, T>
 		return m_current != rhs.m_current;
 	}
 
+	/** @endcond */
+
   private:
 	value_type get() const
 	{
-		if (m_current != nullptr)
-		{
-			row_handle rh{*m_category, *m_current};
-			return rh[m_column_ix].template as<T>();
-		}
-
-		return {};
+		return m_current ?  m_current[m_item_ix].template as<value_type>() : value_type{};
 	}
 
-	category_type *m_category = nullptr;
-	row_type *m_current = nullptr;
+	row_handle m_current;
 	value_type m_value;
-	uint16_t m_column_ix;
+	uint16_t m_item_ix;
 };
 
 // --------------------------------------------------------------------
 // iterator proxy
 
+/**
+ * @brief An iterator_proxy is used as a result type for methods that
+ * return a range of values you want to iterate over.
+ *
+ * E.g. the class cif::category contains the method cif::category::rows()
+ * that returns an iterator_proxy that allows you to iterate over
+ * all the rows in the category.
+ *
+ * @tparam Category The category for the iterators
+ * @tparam Ts The types the iterators return. See class: iterator
+ */
+
 template <typename Category, typename... Ts>
 class iterator_proxy
 {
   public:
-	static constexpr const size_t N = sizeof...(Ts);
+	/** @cond */
+	static constexpr const std::size_t N = sizeof...(Ts);
 
 	using category_type = Category;
 	using row_type = std::conditional_t<std::is_const_v<category_type>, const row, row>;
@@ -423,51 +458,61 @@ class iterator_proxy
 	using iterator = iterator_impl<category_type, Ts...>;
 	using row_iterator = iterator_impl<category_type>;
 
-	iterator_proxy(category_type &cat, row_iterator pos, char const *const columns[N]);
-	iterator_proxy(category_type &cat, row_iterator pos, std::initializer_list<char const *> columns);
+	iterator_proxy(category_type &cat, row_iterator pos, char const *const items[N]);
+	iterator_proxy(category_type &cat, row_iterator pos, std::initializer_list<char const *> items);
 
 	iterator_proxy(iterator_proxy &&p);
 	iterator_proxy &operator=(iterator_proxy &&p);
 
 	iterator_proxy(const iterator_proxy &) = delete;
 	iterator_proxy &operator=(const iterator_proxy &) = delete;
+	/** @endcond */
 
-	iterator begin() const { return iterator(m_begin, m_column_ix); }
-	iterator end() const { return iterator(m_end, m_column_ix); }
+	iterator begin() const { return iterator(m_begin, m_item_ix); } ///< Return the iterator pointing to the first row
+	iterator end() const { return iterator(m_end, m_item_ix); }     ///< Return the iterator pointing past the last row
 
-	bool empty() const { return m_begin == m_end; }
-
-	explicit operator bool() const { return not empty(); }
-
-	size_t size() const { return std::distance(begin(), end()); }
+	bool empty() const { return m_begin == m_end; }               ///< Return true if the range is empty
+	explicit operator bool() const { return not empty(); }        ///< Easy way to detect if the range is empty
+	std::size_t size() const { return std::distance(begin(), end()); } ///< Return size of the range
 
 	// row front() { return *begin(); }
 	// row back() { return *(std::prev(end())); }
 
-	category_type &category() const { return *m_category; }
+	category_type &category() const { return *m_category; } ///< Return the category the iterator belong to
 
+	/** swap */
 	void swap(iterator_proxy &rhs)
 	{
 		std::swap(m_category, rhs.m_category);
 		std::swap(m_begin, rhs.m_begin);
 		std::swap(m_end, rhs.m_end);
-		std::swap(m_column_ix, rhs.m_column_ix);
+		std::swap(m_item_ix, rhs.m_item_ix);
 	}
 
   private:
 	category_type *m_category;
 	row_iterator m_begin, m_end;
-	std::array<uint16_t, N> m_column_ix;
+	std::array<uint16_t, N> m_item_ix;
 };
 
 // --------------------------------------------------------------------
 // conditional iterator proxy
 
+/**
+ * @brief A conditional iterator proxy is similar to an iterator_proxy
+ * in that it can be used to return a range of rows you can iterate over.
+ * In the case of an conditional_iterator_proxy a cif::condition is used
+ * to filter out only those rows that match the condition.
+ *
+ * @tparam CategoryType The category the iterators belong to
+ * @tparam Ts The types to which the iterators can be dereferenced
+ */
 template <typename CategoryType, typename... Ts>
 class conditional_iterator_proxy
 {
   public:
-	static constexpr const size_t N = sizeof...(Ts);
+	/** @cond */
+	static constexpr const std::size_t N = sizeof...(Ts);
 
 	using category_type = std::remove_cv_t<CategoryType>;
 
@@ -493,22 +538,23 @@ class conditional_iterator_proxy
 
 		reference operator*()
 		{
-			return *mBegin;
+			return *m_begin;
 		}
 
 		pointer operator->()
 		{
-			return &*mBegin;
+			m_current = *m_begin;
+			return &m_current;
 		}
 
 		conditional_iterator_impl &operator++()
 		{
-			while (mBegin != mEnd)
+			while (m_begin != m_end)
 			{
-				if (++mBegin == mEnd)
+				if (++m_begin == m_end)
 					break;
-
-				if (m_condition->operator()(mBegin))
+				
+				if (m_condition->operator()(m_begin))
 					break;
 			}
 
@@ -522,18 +568,22 @@ class conditional_iterator_proxy
 			return result;
 		}
 
-		bool operator==(const conditional_iterator_impl &rhs) const { return mBegin == rhs.mBegin; }
-		bool operator!=(const conditional_iterator_impl &rhs) const { return mBegin != rhs.mBegin; }
+		bool operator==(const conditional_iterator_impl &rhs) const { return m_begin == rhs.m_begin; }
+		bool operator!=(const conditional_iterator_impl &rhs) const { return m_begin != rhs.m_begin; }
+
+		bool operator==(const row_iterator &rhs) const { return m_begin == rhs; }
+		bool operator!=(const row_iterator &rhs) const { return m_begin != rhs; }
 
 		template <typename IRowType, typename... ITs>
-		bool operator==(const iterator_impl<IRowType, ITs...> &rhs) const { return mBegin == rhs; }
+		bool operator==(const iterator_impl<IRowType, ITs...> &rhs) const { return m_begin == rhs; }
 
 		template <typename IRowType, typename... ITs>
-		bool operator!=(const iterator_impl<IRowType, ITs...> &rhs) const { return mBegin != rhs; }
+		bool operator!=(const iterator_impl<IRowType, ITs...> &rhs) const { return m_begin != rhs; }
 
 	  private:
-		CategoryType *mCat;
-		base_iterator mBegin, mEnd;
+		CategoryType *m_cat;
+		base_iterator m_begin, m_end;
+		value_type m_current;
 		const condition *m_condition;
 	};
 
@@ -549,20 +599,21 @@ class conditional_iterator_proxy
 	conditional_iterator_proxy(const conditional_iterator_proxy &) = delete;
 	conditional_iterator_proxy &operator=(const conditional_iterator_proxy &) = delete;
 
-	iterator begin() const;
-	iterator end() const;
+	/** @endcond */
 
-	bool empty() const;
+	iterator begin() const; ///< Return the iterator pointing to the first row
+	iterator end() const;   ///< Return the iterator pointing past the last row
 
-	explicit operator bool() const { return not empty(); }
+	bool empty() const;                                           ///< Return true if the range is empty
+	explicit operator bool() const { return not empty(); }        ///< Easy way to detect if the range is empty
+	std::size_t size() const { return std::distance(begin(), end()); } ///< Return size of the range
 
-	size_t size() const { return std::distance(begin(), end()); }
-
-	row_handle front() { return *begin(); }
+	row_handle front() { return *begin(); } ///< Return reference to the first row
 	// row_handle back() { return *begin(); }
 
-	CategoryType &category() const { return *m_cat; }
+	CategoryType &category() const { return *m_cat; } ///< Category the iterators belong to
 
+	/** swap */
 	void swap(conditional_iterator_proxy &rhs);
 
   private:
@@ -574,27 +625,28 @@ class conditional_iterator_proxy
 
 // --------------------------------------------------------------------
 
+/** @cond */
 template <typename Category, typename... Ts>
-iterator_proxy<Category, Ts...>::iterator_proxy(Category &cat, row_iterator pos, char const *const columns[N])
+iterator_proxy<Category, Ts...>::iterator_proxy(Category &cat, row_iterator pos, char const *const items[N])
 	: m_category(&cat)
 	, m_begin(pos)
 	, m_end(cat.end())
 {
 	for (uint16_t i = 0; i < N; ++i)
-		m_column_ix[i] = m_category->get_column_ix(columns[i]);
+		m_item_ix[i] = m_category->get_item_ix(items[i]);
 }
 
 template <typename Category, typename... Ts>
-iterator_proxy<Category, Ts...>::iterator_proxy(Category &cat, row_iterator pos, std::initializer_list<char const *> columns)
+iterator_proxy<Category, Ts...>::iterator_proxy(Category &cat, row_iterator pos, std::initializer_list<char const *> items)
 	: m_category(&cat)
 	, m_begin(pos)
 	, m_end(cat.end())
 {
-	// static_assert(columns.size() == N, "The list of column names should be exactly the same as the list of requested columns");
+	// static_assert(items.size() == N, "The list of item names should be exactly the same as the list of requested items");
 
 	std::uint16_t i = 0;
-	for (auto column : columns)
-		m_column_ix[i++] = m_category->get_column_ix(column);
+	for (auto item : items)
+		m_item_ix[i++] = m_category->get_item_ix(item);
 }
 
 // --------------------------------------------------------------------
@@ -602,11 +654,13 @@ iterator_proxy<Category, Ts...>::iterator_proxy(Category &cat, row_iterator pos,
 template <typename Category, typename... Ts>
 conditional_iterator_proxy<Category, Ts...>::conditional_iterator_impl::conditional_iterator_impl(
 	Category &cat, row_iterator pos, const condition &cond, const std::array<uint16_t, N> &cix)
-	: mCat(&cat)
-	, mBegin(pos, cix)
-	, mEnd(cat.end(), cix)
+	: m_cat(&cat)
+	, m_begin(pos, cix)
+	, m_end(cat.end(), cix)
 	, m_condition(&cond)
 {
+	if (m_condition == nullptr or m_condition->empty())
+		m_begin = m_end;
 }
 
 template <typename Category, typename... Ts>
@@ -629,15 +683,20 @@ conditional_iterator_proxy<Category, Ts...>::conditional_iterator_proxy(Category
 	, mCBegin(pos)
 	, mCEnd(cat.end())
 {
-	static_assert(sizeof...(Ts) == sizeof...(Ns), "Number of column names should be equal to number of requested value types");
+	static_assert(sizeof...(Ts) == sizeof...(Ns), "Number of item names should be equal to number of requested value types");
 
-	m_condition.prepare(cat);
+	if (m_condition)
+	{
+		m_condition.prepare(cat);
 
-	while (mCBegin != mCEnd and not m_condition(*mCBegin))
-		++mCBegin;
+		while (mCBegin != mCEnd and not m_condition(*mCBegin))
+			++mCBegin;
+	}
+	else
+		mCBegin = mCEnd;
 
 	uint16_t i = 0;
-	((mCix[i++] = m_cat->get_column_ix(names)), ...);
+	((mCix[i++] = m_cat->get_item_ix(names)), ...);
 }
 
 template <typename Category, typename... Ts>
@@ -675,4 +734,6 @@ void conditional_iterator_proxy<Category, Ts...>::swap(conditional_iterator_prox
 	std::swap(mCix, rhs.mCix);
 }
 
+/** @endcond */
+
 } // namespace cif

include/cif++/matrix.hpp

@@ -35,33 +35,51 @@
 #include <type_traits>
 #include <vector>
 
+/**
+ * @file matrix.hpp
+ * 
+ * Some basic matrix operations and classes to hold matrices.
+ * 
+ * We're using expression templates for optimal performance.
+ * 
+ */
+
 namespace cif
 {
 // --------------------------------------------------------------------
 // We're using expression templates here
 
+/**
+ * @brief Base for the matrix expression templates
+ * This all uses the Curiously recurring template pattern
+ * 
+ * @tparam M The type of the derived class
+ */
 template <typename M>
 class matrix_expression
 {
   public:
-	constexpr size_t dim_m() const { return static_cast<const M &>(*this).dim_m(); }
-	constexpr size_t dim_n() const { return static_cast<const M &>(*this).dim_n(); }
+	constexpr std::size_t dim_m() const { return static_cast<const M &>(*this).dim_m(); } ///< Return the size (dimension) in direction m
+	constexpr std::size_t dim_n() const { return static_cast<const M &>(*this).dim_n(); } ///< Return the size (dimension) in direction n
 
-	constexpr bool empty() const { return dim_m() == 0 or dim_n() == 0; }
+	constexpr bool empty() const { return dim_m() == 0 or dim_n() == 0; } ///< Convenient way to test for empty matrices
 
-	constexpr auto &operator()(size_t i, size_t j)
+	/** Return a reference to element [ @a i, @a j ] */
+	constexpr auto &operator()(std::size_t i, std::size_t j)
 	{
 		return static_cast<M &>(*this).operator()(i, j);
 	}
 
-	constexpr auto operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return static_cast<const M &>(*this).operator()(i, j);
 	}
 
-	void swap_row(size_t r1, size_t r2)
+	/** Swap the contents of rows @a r1 and @a r2 */
+	void swap_row(std::size_t r1, std::size_t r2)
 	{
-		for (size_t c = 0; c < dim_m(); ++c)
+		for (std::size_t c = 0; c < dim_m(); ++c)
 		{
 			auto v = operator()(r1, c);
 			operator()(r1, c) = operator()(r2, c);
@@ -69,9 +87,10 @@ class matrix_expression
 		}
 	}
 
-	void swap_col(size_t c1, size_t c2)
+	/** Swap the contents of columns @a c1 and @a c2 */
+	void swap_col(std::size_t c1, std::size_t c2)
 	{
-		for (size_t r = 0; r < dim_n(); ++r)
+		for (std::size_t r = 0; r < dim_n(); ++r)
 		{
 			auto &a = operator()(r, c1);
 			auto &b = operator()(r, c2);
@@ -79,15 +98,16 @@ class matrix_expression
 		}
 	}
 
+	/** write the matrix @a m to std::ostream @a os */
 	friend std::ostream &operator<<(std::ostream &os, const matrix_expression &m)
 	{
 		os << '[';
 
-		for (size_t i = 0; i < m.dim_m(); ++i)
+		for (std::size_t i = 0; i < m.dim_m(); ++i)
 		{
 			os << '[';
 
-			for (size_t j = 0; j < m.dim_n(); ++j)
+			for (std::size_t j = 0; j < m.dim_n(); ++j)
 			{
 				os << m(i, j);
 				if (j + 1 < m.dim_n())
@@ -107,29 +127,51 @@ class matrix_expression
 };
 
 // --------------------------------------------------------------------
-// matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
-// element m i,j is mapped to [i * n + j] and thus storage is row major
+
+/**
+ * @brief Storage class implementation of matrix_expression.
+ * 
+ * @tparam F The type of the stored values
+ *  
+ * matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
+ * element m i,j is mapped to [i * n + j] and thus storage is row major
+ */
 
 template <typename F = float>
 class matrix : public matrix_expression<matrix<F>>
 {
   public:
+	/** The value type */
 	using value_type = F;
 
+	/**
+	 * @brief Copy construct a new matrix object using @a m
+	 * 
+	 * @tparam M2 Type of @a m
+	 * @param m The matrix expression to copy values from
+	 */
 	template <typename M2>
 	matrix(const matrix_expression<M2> &m)
 		: m_m(m.dim_m())
 		, m_n(m.dim_n())
 		, m_data(m_m * m_n)
 	{
-		for (size_t i = 0; i < m_m; ++i)
+		for (std::size_t i = 0; i < m_m; ++i)
 		{
-			for (size_t j = 0; j < m_n; ++j)
+			for (std::size_t j = 0; j < m_n; ++j)
 				operator()(i, j) = m(i, j);
 		}
 	}
 
-	matrix(size_t m, size_t n, value_type v = 0)
+	/**
+	 * @brief Construct a new matrix object with dimension @a m and @a n
+	 * setting the values to @a v
+	 * 
+	 * @param m Requested dimension M
+	 * @param n Requested dimension N
+	 * @param v Value to store in each element
+	 */
+	matrix(std::size_t m, std::size_t n, value_type v = 0)
 		: m_m(m)
 		, m_n(n)
 		, m_data(m_m * m_n)
@@ -137,23 +179,27 @@ class matrix : public matrix_expression<matrix<F>>
 		std::fill(m_data.begin(), m_data.end(), v);
 	}
 
+	/** @cond */
 	matrix() = default;
 	matrix(matrix &&m) = default;
 	matrix(const matrix &m) = default;
 	matrix &operator=(matrix &&m) = default;
 	matrix &operator=(const matrix &m) = default;
+	/** @endcond */
 
-	constexpr size_t dim_m() const { return m_m; }
-	constexpr size_t dim_n() const { return m_n; }
+	constexpr std::size_t dim_m() const { return m_m; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_n; } ///< Return dimension n
 
-	constexpr value_type operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		assert(i < m_m);
 		assert(j < m_n);
 		return m_data[i * m_n + j];
 	}
 
-	constexpr value_type &operator()(size_t i, size_t j)
+	/** Return a reference to element [ @a i, @a j ] */
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		assert(i < m_m);
 		assert(j < m_n);
@@ -161,65 +207,85 @@ class matrix : public matrix_expression<matrix<F>>
 	}
 
   private:
-	size_t m_m = 0, m_n = 0;
+	std::size_t m_m = 0, m_n = 0;
 	std::vector<value_type> m_data;
 };
 
 // --------------------------------------------------------------------
 // special case, 3x3 matrix
 
-template <typename F, size_t M, size_t N>
+/**
+ * @brief Storage class implementation of matrix_expression
+ * with compile time fixed size.
+ * 
+ * @tparam F The type of the stored values
+ *  
+ * matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
+ * element m i,j is mapped to [i * n + j] and thus storage is row major
+ */
+
+template <typename F, std::size_t M, std::size_t N>
 class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 {
   public:
+	/** The value type */
 	using value_type = F;
 
-	static constexpr size_t kSize = M * N;
+	/** The storage size */
+	static constexpr std::size_t kSize = M * N;
 
+	/** Copy constructor */
 	template <typename M2>
 	matrix_fixed(const M2 &m)
 	{
 		assert(M == m.dim_m() and N == m.dim_n());
-		for (size_t i = 0; i < M; ++i)
+		for (std::size_t i = 0; i < M; ++i)
 		{
-			for (size_t j = 0; j < N; ++j)
+			for (std::size_t j = 0; j < N; ++j)
 				operator()(i, j) = m(i, j);
 		}
 	}
 
+	/** default constructor */
 	matrix_fixed(value_type v = 0)
 	{
 		m_data.fill(v);
 	}
 
+	/** Alternate constructor taking an array of values to store */
 	matrix_fixed(const F (&v)[kSize])
 	{
 		fill(v, std::make_index_sequence<kSize>{});
 	}
 
+	/** @cond */
 	matrix_fixed(matrix_fixed &&m) = default;
 	matrix_fixed(const matrix_fixed &m) = default;
 	matrix_fixed &operator=(matrix_fixed &&m) = default;
 	matrix_fixed &operator=(const matrix_fixed &m) = default;
+	/** @endcond */
 
-	template<size_t... Ixs>
+	/** Store the values in @a a in the matrix */
+	template<std::size_t... Ixs>
 	matrix_fixed& fill(const F (&a)[kSize], std::index_sequence<Ixs...>)
 	{
 		m_data = { a[Ixs]... };
 		return *this;
 	}
 
-	constexpr size_t dim_m() const { return M; }
-	constexpr size_t dim_n() const { return N; }
+	constexpr std::size_t dim_m() const { return M; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return N; } ///< Return dimension n
 
-	constexpr value_type operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		assert(i < M);
 		assert(j < N);
 		return m_data[i * N + j];
 	}
 
-	constexpr value_type &operator()(size_t i, size_t j)
+	/** Return a reference to element [ @a i, @a j ] */
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		assert(i < M);
 		assert(j < N);
@@ -230,44 +296,60 @@ class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 	std::array<value_type, M * N> m_data;
 };
 
+/** typedef of a fixed matrix of size 3x3 */
 template <typename F>
 using matrix3x3 = matrix_fixed<F, 3, 3>;
 
+/** typedef of a fixed matrix of size 4x4 */
 template <typename F>
 using matrix4x4 = matrix_fixed<F, 4, 4>;
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief Storage class implementation of symmetric matrix_expression
+ * 
+ * @tparam F The type of the stored values
+ *  
+ * matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
+ * element m i,j is mapped to [i * n + j] and thus storage is row major
+ */
 template <typename F = float>
 class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
 {
   public:
+	/** The value type */
 	using value_type = F;
 
-	symmetric_matrix(size_t n, value_type v = 0)
+	/** constructor for a matrix of size @a n x @a n elements with value @a v */
+	symmetric_matrix(std::size_t n, value_type v = 0)
 		: m_n(n)
 		, m_data((m_n * (m_n + 1)) / 2)
 	{
 		std::fill(m_data.begin(), m_data.end(), v);
 	}
 
+	/** @cond */
 	symmetric_matrix() = default;
 	symmetric_matrix(symmetric_matrix &&m) = default;
 	symmetric_matrix(const symmetric_matrix &m) = default;
 	symmetric_matrix &operator=(symmetric_matrix &&m) = default;
 	symmetric_matrix &operator=(const symmetric_matrix &m) = default;
+	/** @endcond */
 
-	constexpr size_t dim_m() const { return m_n; }
-	constexpr size_t dim_n() const { return m_n; }
+	constexpr std::size_t dim_m() const { return m_n; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_n; } ///< Return dimension n
 
-	constexpr value_type operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		return i < j
 		           ? m_data[(j * (j + 1)) / 2 + i]
 		           : m_data[(i * (i + 1)) / 2 + j];
 	}
 
-	constexpr value_type &operator()(size_t i, size_t j)
+	/** Return a reference to element [ @a i, @a j ] */
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		if (i > j)
 			std::swap(i, j);
@@ -276,39 +358,54 @@ class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
 	}
 
   private:
-	size_t m_n;
+	std::size_t m_n;
 	std::vector<value_type> m_data;
 };
 
 // --------------------------------------------------------------------
 
-template <typename F, size_t M>
+/**
+ * @brief Storage class implementation of symmetric matrix_expression
+ * with compile time fixed size.
+ * 
+ * @tparam F The type of the stored values
+ *  
+ * matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
+ * element m i,j is mapped to [i * n + j] and thus storage is row major
+ */
+template <typename F, std::size_t M>
 class symmetric_matrix_fixed : public matrix_expression<symmetric_matrix_fixed<F, M>>
 {
   public:
+	/** The value type */
 	using value_type = F;
 
+	/** constructor with all elements set to value @a v */
 	symmetric_matrix_fixed(value_type v = 0)
 	{
 		std::fill(m_data.begin(), m_data.end(), v);
 	}
 
+	/** @cond */
 	symmetric_matrix_fixed(symmetric_matrix_fixed &&m) = default;
 	symmetric_matrix_fixed(const symmetric_matrix_fixed &m) = default;
 	symmetric_matrix_fixed &operator=(symmetric_matrix_fixed &&m) = default;
 	symmetric_matrix_fixed &operator=(const symmetric_matrix_fixed &m) = default;
+	/** @endcond */
 
-	constexpr size_t dim_m() const { return M; }
-	constexpr size_t dim_n() const { return M; }
+	constexpr std::size_t dim_m() const { return M; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return M; } ///< Return dimension n
 
-	constexpr value_type operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		return i < j
 		           ? m_data[(j * (j + 1)) / 2 + i]
 		           : m_data[(i * (i + 1)) / 2 + j];
 	}
 
-	constexpr value_type &operator()(size_t i, size_t j)
+	/** Return a reference to element [ @a i, @a j ] */
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		if (i > j)
 			std::swap(i, j);
@@ -320,44 +417,65 @@ class symmetric_matrix_fixed : public matrix_expression<symmetric_matrix_fixed<F
 	std::array<value_type, (M * (M + 1)) / 2> m_data;
 };
 
+/** typedef of a fixed symmetric matrix of size 3x3 */
 template <typename F>
 using symmetric_matrix3x3 = symmetric_matrix_fixed<F, 3>;
 
+/** typedef of a fixed symmetric matrix of size 4x4 */
 template <typename F>
 using symmetric_matrix4x4 = symmetric_matrix_fixed<F, 4>;
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief implementation of symmetric matrix_expression with a value
+ * of 1 for the diagonal values and 0 for all the others.
+ *  
+ * @tparam F The type of the stored values
+ *  
+ * matrix is m x n, addressing i,j is 0 <= i < m and 0 <= j < n
+ * element m i,j is mapped to [i * n + j] and thus storage is row major
+ */
 template <typename F = float>
 class identity_matrix : public matrix_expression<identity_matrix<F>>
 {
   public:
+	/** the value type */
 	using value_type = F;
 
-	identity_matrix(size_t n)
+	/** constructor taking a dimension @a n */
+	identity_matrix(std::size_t n)
 		: m_n(n)
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_n; }
-	constexpr size_t dim_n() const { return m_n; }
+	constexpr std::size_t dim_m() const { return m_n; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_n; } ///< Return dimension n
 
-	constexpr value_type operator()(size_t i, size_t j) const
+	/** Return the value of element [ @a i, @a j ] */
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		return static_cast<value_type>(i == j ? 1 : 0);
 	}
 
   private:
-	size_t m_n;
+	std::size_t m_n;
 };
 
 // --------------------------------------------------------------------
 // matrix functions, implemented as expression templates
 
+/**
+ * @brief Implementation of a substraction operation as a matrix expression
+ * 
+ * @tparam M1 Type of matrix 1
+ * @tparam M2 Type of matrix 2
+ */
 template <typename M1, typename M2>
 class matrix_subtraction : public matrix_expression<matrix_subtraction<M1, M2>>
 {
   public:
+	/** constructor */
 	matrix_subtraction(const M1 &m1, const M2 &m2)
 		: m_m1(m1)
 		, m_m2(m2)
@@ -366,10 +484,11 @@ class matrix_subtraction : public matrix_expression<matrix_subtraction<M1, M2>>
 		assert(m_m1.dim_n() == m_m2.dim_n());
 	}
 
-	constexpr size_t dim_m() const { return m_m1.dim_m(); }
-	constexpr size_t dim_n() const { return m_m1.dim_n(); }
+	constexpr std::size_t dim_m() const { return m_m1.dim_m(); } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_m1.dim_n(); } ///< Return dimension n
 
-	constexpr auto operator()(size_t i, size_t j) const
+	/** Access to the value of element [ @a i, @a j ] */
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return m_m1(i, j) - m_m2(i, j);
 	}
@@ -379,16 +498,24 @@ class matrix_subtraction : public matrix_expression<matrix_subtraction<M1, M2>>
 	const M2 &m_m2;
 };
 
+/** operator to subtract two matrices and return a matrix expression */
 template <typename M1, typename M2>
 auto operator-(const matrix_expression<M1> &m1, const matrix_expression<M2> &m2)
 {
 	return matrix_subtraction(m1, m2);
 }
 
+/**
+ * @brief Implementation of a multiplication operation as a matrix expression
+ * 
+ * @tparam M1 Type of matrix 1
+ * @tparam M2 Type of matrix 2
+ */
 template <typename M1, typename M2>
 class matrix_matrix_multiplication : public matrix_expression<matrix_matrix_multiplication<M1, M2>>
 {
   public:
+	/** constructor */
 	matrix_matrix_multiplication(const M1 &m1, const M2 &m2)
 		: m_m1(m1)
 		, m_m2(m2)
@@ -396,16 +523,17 @@ class matrix_matrix_multiplication : public matrix_expression<matrix_matrix_mult
 		assert(m1.dim_m() == m2.dim_n());
 	}
 
-	constexpr size_t dim_m() const { return m_m1.dim_m(); }
-	constexpr size_t dim_n() const { return m_m1.dim_n(); }
+	constexpr std::size_t dim_m() const { return m_m1.dim_m(); } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_m1.dim_n(); } ///< Return dimension n
 
-	constexpr auto operator()(size_t i, size_t j) const
+	/** Access to the value of element [ @a i, @a j ] */
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		using value_type = decltype(m_m1(0, 0));
 
 		value_type result = {};
 
-		for (size_t k = 0; k < m_m1.dim_m(); ++k)
+		for (std::size_t k = 0; k < m_m1.dim_m(); ++k)
 			result += m_m1(i, k) * m_m2(k, j);
 
 		return result;
@@ -416,22 +544,31 @@ class matrix_matrix_multiplication : public matrix_expression<matrix_matrix_mult
 	const M2 &m_m2;
 };
 
+/**
+ * @brief Implementation of a multiplication operation of a matrix and a scalar value as a matrix expression
+ * 
+ * @tparam M1 Type of matrix
+ * @tparam M2 Type of scalar value
+ */
 template <typename M, typename T>
 class matrix_scalar_multiplication : public matrix_expression<matrix_scalar_multiplication<M, T>>
 {
   public:
+	/** value type */
 	using value_type = T;
 
+	/** constructor */
 	matrix_scalar_multiplication(const M &m, value_type v)
 		: m_m(m)
 		, m_v(v)
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_m.dim_m(); }
-	constexpr size_t dim_n() const { return m_m.dim_n(); }
+	constexpr std::size_t dim_m() const { return m_m.dim_m(); } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_m.dim_n(); } ///< Return dimension n
 
-	constexpr auto operator()(size_t i, size_t j) const
+	/** Access to the value of element [ @a i, @a j ] */
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return m_m(i, j) * m_v;
 	}
@@ -441,12 +578,14 @@ class matrix_scalar_multiplication : public matrix_expression<matrix_scalar_mult
 	value_type m_v;
 };
 
+/** First implementation of operator*, enabled if the second parameter is a scalar */
 template <typename M1, typename T, std::enable_if_t<std::is_floating_point_v<T>, int> = 0>
 auto operator*(const matrix_expression<M1> &m, T v)
 {
 	return matrix_scalar_multiplication(m, v);
 }
 
+/** First implementation of operator*, enabled if the second parameter is not a scalar and thus must be a matrix, right? */
 template <typename M1, typename M2, std::enable_if_t<not std::is_floating_point_v<M2>, int> = 0>
 auto operator*(const matrix_expression<M1> &m1, const matrix_expression<M2> &m2)
 {
@@ -455,9 +594,14 @@ auto operator*(const matrix_expression<M1> &m1, const matrix_expression<M2> &m2)
 
 // --------------------------------------------------------------------
 
+/** Generic routine to calculate the determinant of a matrix
+ * 
+ * @note This is currently only implemented for fixed matrices of size 3x3
+ */
 template <typename M>
 auto determinant(const M &m);
 
+/** Implementation of the determinant function for fixed size matrices of size 3x3 */
 template <typename F = float>
 auto determinant(const matrix3x3<F> &m)
 {
@@ -466,9 +610,14 @@ auto determinant(const matrix3x3<F> &m)
 			m(0, 2) * (m(1, 0) * m(2, 1) - m(1, 1) * m(2, 0)));
 }
 
+/** Generic routine to calculate the inverse of a matrix
+ * 
+ * @note This is currently only implemented for fixed matrices of size 3x3
+ */
 template <typename M>
 M inverse(const M &m);
 
+/** Implementation of the inverse function for fixed size matrices of size 3x3 */
 template <typename F = float>
 matrix3x3<F> inverse(const matrix3x3<F> &m)
 {
@@ -491,29 +640,36 @@ matrix3x3<F> inverse(const matrix3x3<F> &m)
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief Implementation of a cofactor calculation as a matrix expression
+ * 
+ * @tparam M Type of matrix
+ */
 template <typename M>
 class matrix_cofactors : public matrix_expression<matrix_cofactors<M>>
 {
   public:
+	/** constructor */
 	matrix_cofactors(const M &m)
 		: m_m(m)
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_m.dim_m(); }
-	constexpr size_t dim_n() const { return m_m.dim_n(); }
+	constexpr std::size_t dim_m() const { return m_m.dim_m(); } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return m_m.dim_n(); } ///< Return dimension n
 
-	constexpr auto operator()(size_t i, size_t j) const
+	/** Access to the value of element [ @a i, @a j ] */
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
-		const size_t ixs[4][3] = {
+		const std::size_t ixs[4][3] = {
 			{ 1, 2, 3 },
 			{ 0, 2, 3 },
 			{ 0, 1, 3 },
 			{ 0, 1, 2 }
 		};
 
-		const size_t *ix = ixs[i];
-		const size_t *iy = ixs[j];
+		const std::size_t *ix = ixs[i];
+		const std::size_t *iy = ixs[j];
 
 		auto result =
 			m_m(ix[0], iy[0]) * m_m(ix[1], iy[1]) * m_m(ix[2], iy[2]) +

include/cif++/parser.hpp

@@ -30,14 +30,24 @@
 
 #include <map>
 
+/**
+ * @file parser.hpp
+ * 
+ * This file contains the declaration of an mmCIF parser
+ */
+
 namespace cif
 {
 
+class validator;
+
 // --------------------------------------------------------------------
 
+/** Exception that is thrown when the mmCIF file contains a parsing error */
 class parse_error : public std::runtime_error
 {
   public:
+	/// \brief constructor
 	parse_error(uint32_t line_nr, const std::string &message)
 		: std::runtime_error("parse error at line " + std::to_string(line_nr) + ": " + message)
 	{
@@ -46,57 +56,91 @@ class parse_error : public std::runtime_error
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief The sac_parser is a similar to SAX parsers (Simple API for XML, 
+ * in our case it is Simple API for CIF)
+ * 
+ * This is a hand crafted, optimised parser for reading cif files,
+ * both cif 1.0 and cif 1.1 is supported. But version 2.0 is not.
+ * That means that the content of files strictly contains only
+ * ASCII characters. Anything else will generate an error.
+ * 
+ * This class is an abstract base class. Derived classes should
+ * implement the produce_ methods.
+ */
+
 // TODO: Need to implement support for transformed long lines
 
 class sac_parser
 {
   public:
-	using datablock_index = std::map<std::string, std::size_t>;
-
-	virtual ~sac_parser() = default;
-
-	enum CharTraitsMask : uint8_t
+	/** @cond */
+	struct iless_op
 	{
-		kOrdinaryMask = 1 << 0,
-		kNonBlankMask = 1 << 1,
-		kTextLeadMask = 1 << 2,
-		kAnyPrintMask = 1 << 3
+		bool operator()(std::string_view a, std::string_view b) const
+		{
+			return icompare(a, b) < 0;
+		}
 	};
 
+	using datablock_index = std::map<std::string, std::size_t, iless_op>;
+
+	virtual ~sac_parser() = default;
+	/** @endcond */
+
+	/// \brief The parser only supports ASCII so we can
+	/// create a table with character properties.
+	enum CharTraitsMask : uint8_t
+	{
+		kOrdinaryMask = 1 << 0,	///< The character is in the Ordinary class
+		kNonBlankMask = 1 << 1,	///< The character is in the NonBlank class
+		kTextLeadMask = 1 << 2,	///< The character is in the TextLead class
+		kAnyPrintMask = 1 << 3	///< The character is in the AnyPrint class
+	};
+
+	/// \brief Return true if the character @a ch is a *space* character
 	static constexpr bool is_space(int ch)
 	{
 		return ch == ' ' or ch == '\t' or ch == '\r' or ch == '\n';
 	}
 
+	/// \brief Return true if the character @a ch is a *white* character
 	static constexpr bool is_white(int ch)
 	{
 		return is_space(ch) or ch == '#';
 	}
 
+	/// \brief Return true if the character @a ch is a *ordinary* character
 	static constexpr bool is_ordinary(int ch)
 	{
 		return ch >= 0x20 and ch <= 0x7f and (kCharTraitsTable[ch - 0x20] & kOrdinaryMask) != 0;
 	}
 
+	/// \brief Return true if the character @a ch is a *non_blank* character
 	static constexpr bool is_non_blank(int ch)
 	{
 		return ch > 0x20 and ch <= 0x7f and (kCharTraitsTable[ch - 0x20] & kNonBlankMask) != 0;
 	}
 
+	/// \brief Return true if the character @a ch is a *text_lead* character
 	static constexpr bool is_text_lead(int ch)
 	{
 		return ch >= 0x20 and ch <= 0x7f and (kCharTraitsTable[ch - 0x20] & kTextLeadMask) != 0;
 	}
 
+	/// \brief Return true if the character @a ch is a *any_print* character
 	static constexpr bool is_any_print(int ch)
 	{
 		return ch == '\t' or
 		       (ch >= 0x20 and ch <= 0x7f and (kCharTraitsTable[ch - 0x20] & kAnyPrintMask) != 0);
 	}
 
+	/// \brief Return true if the string in @a text can safely be written without quotation
 	static bool is_unquoted_string(std::string_view text);
 
   protected:
+	/** @cond */
+
 	static constexpr uint8_t kCharTraitsTable[128] = {
 		//	0	1	2	3	4	5	6	7	8	9	a	b	c	d	e	f
 		14, 15, 14, 14, 14, 15, 15, 14, 15, 15, 15, 15, 15, 15, 15, 15, //	2
@@ -109,9 +153,9 @@ class sac_parser
 
 	enum class CIFToken
 	{
-		Unknown,
+		UNKNOWN,
 
-		Eof,
+		END_OF_FILE,
 
 		DATA,
 		LOOP,
@@ -119,24 +163,24 @@ class sac_parser
 		SAVE_,
 		SAVE_NAME,
 		STOP,
-		Tag,
-		Value
+		ITEM_NAME,
+		VALUE
 	};
 
 	static constexpr const char *get_token_name(CIFToken token)
 	{
 		switch (token)
 		{
-			case CIFToken::Unknown: return "Unknown";
-			case CIFToken::Eof: return "Eof";
+			case CIFToken::UNKNOWN: return "Unknown";
+			case CIFToken::END_OF_FILE: return "Eof";
 			case CIFToken::DATA: return "DATA";
 			case CIFToken::LOOP: return "LOOP";
 			case CIFToken::GLOBAL: return "GLOBAL";
 			case CIFToken::SAVE_: return "SAVE";
 			case CIFToken::SAVE_NAME: return "SAVE+name";
 			case CIFToken::STOP: return "STOP";
-			case CIFToken::Tag: return "Tag";
-			case CIFToken::Value: return "Value";
+			case CIFToken::ITEM_NAME: return "Tag";
+			case CIFToken::VALUE: return "Value";
 			default: return "Invalid token parameter";
 		}
 	}
@@ -152,17 +196,45 @@ class sac_parser
 
 	void match(CIFToken token);
 
+	/** @endcond */
+
   public:
+
+	/** \brief Parse only a single datablock in the string @a datablock
+	 * The start of the datablock is first located and then data
+	 * is parsed up until the next start of a datablock or the end of
+	 * the data.
+	 * */
 	bool parse_single_datablock(const std::string &datablock);
 
+	/** \brief Return an index for all the datablocks found, that is
+	 * the index will contain the names and offsets for each.
+	 */
 	datablock_index index_datablocks();
 
+	/**
+	 * @brief Parse the datablock named @a datablock
+	 * 
+	 * This will first lookup the datablock's offset in the index @a index
+	 * and then start parsing from that location until the next datablock.
+	 * 
+	 * @param datablock Name of the datablock to parse
+	 * @param index The index created using index_datablocks
+	 * @return true If the datablock was found
+	 * @return false If the datablock was not found
+	 */
 	bool parse_single_datablock(const std::string &datablock, const datablock_index &index);
 
+	/**
+	 * @brief Parse the file
+	 * 
+	 */
 	void parse_file();
 
   protected:
 
+	/** @cond */
+
 	sac_parser(std::istream &is, bool init = true);
 
 	void parse_global();
@@ -174,7 +246,7 @@ class sac_parser
 	void error(const std::string &msg)
 	{
 		if (cif::VERBOSE > 0)
-			std::cerr << "Error parsing mmCIF: " << msg << std::endl;
+			std::cerr << "Error parsing mmCIF: " << msg << '\n';
 
 		throw parse_error(m_line_nr, msg);
 	}
@@ -182,7 +254,7 @@ class sac_parser
 	void warning(const std::string &msg)
 	{
 		if (cif::VERBOSE > 0)
-			std::cerr << "parser warning at line " << m_line_nr << ": " << msg << std::endl;
+			std::cerr << "parser warning at line " << m_line_nr << ": " << msg << '\n';
 	}
 
 	// production methods, these are pure virtual here
@@ -205,9 +277,9 @@ class sac_parser
 		QuotedString,
 		QuotedStringQuote,
 		UnquotedString,
-		Tag,
-		TextField,
-		TextFieldNL,
+		ItemName,
+		TextItem,
+		TextItemNL,
 		Reserved,
 		Value
 	};
@@ -222,19 +294,37 @@ class sac_parser
 	// token buffer
 	std::vector<char> m_token_buffer;
 	std::string_view m_token_value;
+
+	/** @endcond */
 };
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief An actual implementation of a sac_parser generating data in a file
+ * 
+ * This parser will create the cif::file, cif::datablock and cif::category
+ * objects required to contain all data
+ */
 class parser : public sac_parser
 {
   public:
+	/// \brief constructor, generates data into @a file from @a is using validator @a v
+	parser(std::istream &is, file &file, const validator *v)
+		: sac_parser(is)
+		, m_file(file)
+		, m_validator(v)
+	{
+	}
+
+	/// \brief constructor, generates data into @a file from @a is
 	parser(std::istream &is, file &file)
 		: sac_parser(is)
 		, m_file(file)
 	{
 	}
 
+	/** @cond */
 	void produce_datablock(std::string_view name) override;
 
 	void produce_category(std::string_view name) override;
@@ -247,7 +337,10 @@ class parser : public sac_parser
 	file &m_file;
 	datablock *m_datablock = nullptr;
 	category *m_category = nullptr;
+	const validator *m_validator = nullptr;
 	row_handle m_row;
+
+	/** @endcond */
 };
 
 } // namespace cif

include/cif++/pdb.hpp

@@ -0,0 +1,216 @@
+/*-
+ * SPDX-License-Identifier: BSD-2-Clause
+ *
+ * Copyright (c) 2023 NKI/AVL, Netherlands Cancer Institute
+ *
+ * Redistribution and use in source and binary forms, with or without
+ * modification, are permitted provided that the following conditions are met:
+ *
+ * 1. Redistributions of source code must retain the above copyright notice, this
+ *    list of conditions and the following disclaimer
+ * 2. Redistributions in binary form must reproduce the above copyright notice,
+ *    this list of conditions and the following disclaimer in the documentation
+ *    and/or other materials provided with the distribution.
+ *
+ * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
+ * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
+ * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
+ * DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT OWNER OR CONTRIBUTORS BE LIABLE FOR
+ * ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL, EXEMPLARY, OR CONSEQUENTIAL DAMAGES
+ * (INCLUDING, BUT NOT LIMITED TO, PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES;
+ * LOSS OF USE, DATA, OR PROFITS; OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND
+ * ON ANY THEORY OF LIABILITY, WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT
+ * (INCLUDING NEGLIGENCE OR OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS
+ * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
+ */
+
+#pragma once
+
+#include "cif++/file.hpp"
+
+#include <system_error>
+
+/**
+ * @file pdb.hpp
+ *
+ * This file presents the API to read and write files in the
+ * legacy and ancient PDB format.
+ *
+ * The code works on the basis of best effort since it is
+ * impossible to have correct round trip fidelity.
+ *
+ */
+
+namespace cif::pdb
+{
+
+/// --------------------------------------------------------------------
+// PDB to mmCIF
+
+/** @brief Read a file in either mmCIF or PDB format from file @a file,
+ * compressed or not, depending on the content.
+ */
+
+file read(const std::filesystem::path &file);
+
+/** @brief Read a file in either mmCIF or PDB format from std::istream @a is,
+ * compressed or not, depending on the content.
+ */
+
+file read(std::istream &is);
+
+/**
+ * @brief Read a file in legacy PDB format from std::istream @a is and
+ * put the data into @a cifFile
+ */
+file read_pdb_file(std::istream &pdbFile);
+
+// mmCIF to PDB
+
+/** @brief Write out the data in @a db in legacy PDB format
+ * to std::ostream @a os
+ */
+void write(std::ostream &os, const datablock &db);
+
+/** @brief Write out the data in @a f in legacy PDB format
+ * to std::ostream @a os
+ */
+inline void write(std::ostream &os, const file &f)
+{
+	write(os, f.front());
+}
+
+/** @brief Write out the data in @a db to file @a file
+ * in legacy PDB format or mmCIF format, depending on the
+ * filename extension.
+ *
+ * If extension of @a file is *.gz* the resulting file will
+ * be written in gzip compressed format.
+ */
+void write(const std::filesystem::path &file, const datablock &db);
+
+/** @brief Write out the data in @a f to file @a file
+ * in legacy PDB format or mmCIF format, depending on the
+ * filename extension.
+ *
+ * If extension of @a file is *.gz* the resulting file will
+ * be written in gzip compressed format.
+ */
+inline void write(const std::filesystem::path &p, const file &f)
+{
+	write(p, f.front());
+}
+
+// --------------------------------------------------------------------
+
+/** \brief Reconstruct all missing categories for an assumed PDBx file.
+ *
+ * Some people believe that simply dumping some atom records is enough.
+ *
+ * \param file The cif::file that hopefully contains some valid data
+ * \param dictionary The mmcif dictionary to use
+ * \result Returns true if the resulting file is valid
+ */
+
+bool reconstruct_pdbx(file &pdbx_file, std::string_view dictionary = "mmcif_pdbx");
+
+/** \brief This is an extension to cif::validator, use the logic in common
+ * PDBx files to see if the file is internally consistent.
+ *
+ * This function for now checks if the following categories are consistent:
+ *
+ * atom_site -> pdbx_poly_seq_scheme -> entity_poly_seq -> entity_poly -> entity
+ *
+ * Use the common \ref cif::VERBOSE flag to turn on diagnostic messages.
+ * 
+ * This function throws a std::system_error in case of an error
+ *
+ * \param file The input file
+ * \param dictionary The mmcif dictionary to use
+ * \result Returns true if the file was valid and consistent
+ */
+
+bool is_valid_pdbx_file(const file &pdbx_file, std::string_view dictionary = "mmcif_pdbx");
+
+/** \brief This is an extension to cif::validator, use the logic in common
+ * PDBx files to see if the file is internally consistent.
+ *
+ * This function for now checks if the following categories are consistent:
+ *
+ * atom_site -> pdbx_poly_seq_scheme -> entity_poly_seq -> entity_poly -> entity
+ *
+ * Use the common \ref cif::VERBOSE flag to turn on diagnostic messages.
+ * 
+ * The dictionary is assumed to be specified in the file or to be the
+ * default mmcif_pdbx.dic dictionary.
+ *
+ * \param file The input file
+ * \param ec The error_code in case something was wrong
+ * \result Returns true if the file was valid and consistent
+ */
+
+bool is_valid_pdbx_file(const file &pdbx_file, std::error_code &ec);
+
+/** \brief This is an extension to cif::validator, use the logic in common
+ * PDBx files to see if the file is internally consistent.
+ *
+ * This function for now checks if the following categories are consistent:
+ *
+ * atom_site -> pdbx_poly_seq_scheme -> entity_poly_seq -> entity_poly -> entity
+ *
+ * Use the common \ref cif::VERBOSE flag to turn on diagnostic messages.
+ * 
+ * \param file The input file
+ * \param dictionary The dictionary to use
+ * \param ec The error_code in case something was wrong
+ * \result Returns true if the file was valid and consistent
+ */
+
+bool is_valid_pdbx_file(const file &pdbx_file, std::string_view dictionary,
+	std::error_code &ec);
+
+// --------------------------------------------------------------------
+// Other I/O related routines
+
+/** @brief Return the HEADER line for the data in @a data
+ *
+ * The line returned should be compatible with the legacy PDB
+ * format and is e.g. used in the DSSP program.
+ *
+ * @param data The datablock to use as source for the requested data
+ * @param truncate_at The maximum length of the line returned
+ */
+
+std::string get_HEADER_line(const datablock &data, std::string::size_type truncate_at = 127);
+/** @brief Return the COMPND line for the data in @a data
+ *
+ * The line returned should be compatible with the legacy PDB
+ * format and is e.g. used in the DSSP program.
+ *
+ * @param data The datablock to use as source for the requested data
+ * @param truncate_at The maximum length of the line returned
+ */
+
+std::string get_COMPND_line(const datablock &data, std::string::size_type truncate_at = 127);
+/** @brief Return the SOURCE line for the data in @a data
+ *
+ * The line returned should be compatible with the legacy PDB
+ * format and is e.g. used in the DSSP program.
+ *
+ * @param data The datablock to use as source for the requested data
+ * @param truncate_at The maximum length of the line returned
+ */
+
+std::string get_SOURCE_line(const datablock &data, std::string::size_type truncate_at = 127);
+/** @brief Return the AUTHOR line for the data in @a data
+ *
+ * The line returned should be compatible with the legacy PDB
+ * format and is e.g. used in the DSSP program.
+ *
+ * @param data The datablock to use as source for the requested data
+ * @param truncate_at The maximum length of the line returned
+ */
+
+std::string get_AUTHOR_line(const datablock &data, std::string::size_type truncate_at = 127);
+
+} // namespace cif::pdb

include/cif++/pdb/cif2pdb.hpp

@@ -26,17 +26,8 @@
 
 #pragma once
 
-#include "cif++/datablock.hpp"
+/// \file cif2pdb.hpp
+/// \deprecated This file is no longer used. Please use "cif++/pdb.hpp" instead
 
-namespace cif::pdb
-{
+#warning "Use of this file is deprecated, please use "cif++/pdb.hpp"
 
-/// \brief Just the HEADER, COMPND, SOURCE and AUTHOR lines
-void write_header_lines(std::ostream &os, const datablock &data);
-
-std::string get_HEADER_line(const datablock &data, std::string::size_type truncate_at = 127);
-std::string get_COMPND_line(const datablock &data, std::string::size_type truncate_at = 127);
-std::string get_SOURCE_line(const datablock &data, std::string::size_type truncate_at = 127);
-std::string get_AUTHOR_line(const datablock &data, std::string::size_type truncate_at = 127);
-
-} // namespace pdbx

include/cif++/pdb/io.hpp

@@ -26,35 +26,7 @@
 
 #pragma once
 
-#include "cif++/datablock.hpp"
+/// \file io.hpp
+/// \deprecated This file is no longer used. Please use "cif++/pdb.hpp" instead
 
-namespace cif::pdb
-{
-
-/// \brief Read a file in either mmCIF or PDB format, compressed or not,
-/// depending on the content.
-file read(const std::filesystem::path &file);
-
-/// \brief Read a file in either mmCIF or PDB format, compressed or not,
-/// depending on the content.
-file read(std::istream &is);
-
-/// \brief Write out a file in PDB format
-void write(std::ostream &os, const datablock &db);
-
-/// \brief Write out a file in PDB format
-inline void write(std::ostream &os, const file &f)
-{
-	write(os, f.front());
-}
-
-/// \brief Write out a file in PDB format or mmCIF format, depending on the filename extension
-void write(const std::filesystem::path &file, const datablock &db);
-
-/// \brief Write out a file in PDB format or mmCIF format, depending on the filename extension
-inline void write(const std::filesystem::path &p, const file &f)
-{
-	write(p, f.front());
-}
-
-}
+#warning "Use of this file is deprecated, please use "cif++/pdb.hpp"

include/cif++/pdb/pdb2cif.hpp

@@ -26,39 +26,7 @@
 
 #pragma once
 
-#include "cif++/file.hpp"
-namespace cif::pdb
-{
+/// \file pdb2cif.hpp
+/// \deprecated This file is no longer used. Please use "cif++/pdb.hpp" instead
 
-// --------------------------------------------------------------------
-
-struct PDBRecord
-{
-	PDBRecord *mNext;
-	uint32_t mLineNr;
-	char mName[11];
-	size_t mVlen;
-	char mValue[1];
-
-	PDBRecord(uint32_t lineNr, const std::string &name, const std::string &value);
-	~PDBRecord();
-
-	void *operator new(size_t);
-	void *operator new(size_t size, size_t vLen);
-
-	void operator delete(void *p);
-	void operator delete(void *p, size_t vLen);
-
-	bool is(const char *name) const;
-
-	char vC(size_t column);
-	std::string vS(size_t columnFirst, size_t columnLast = std::numeric_limits<size_t>::max());
-	int vI(int columnFirst, int columnLast);
-	std::string vF(size_t columnFirst, size_t columnLast);
-};
-
-// --------------------------------------------------------------------
-
-void ReadPDBFile(std::istream &pdbFile, file &cifFile);
-
-} // namespace pdbx
+#warning "Use of this file is deprecated, please use "cif++/pdb.hpp"

include/cif++/pdb/tls.hpp

@@ -26,26 +26,7 @@
 
 #pragma once
 
-#include "cif++/datablock.hpp"
+/// \file tls.hpp
+/// \deprecated This code has been moved to libpdb-redo
 
-#include <string>
-#include <tuple>
-#include <vector>
-
-namespace cif
-{
-
-struct tls_selection;
-struct tls_residue;
-
-struct tls_selection
-{
-	virtual ~tls_selection() {}
-	virtual void collect_residues(cif::datablock &db, std::vector<tls_residue> &residues, std::size_t indentLevel = 0) const = 0;
-	std::vector<std::tuple<std::string, int, int>> get_ranges(cif::datablock &db, bool pdbNamespace) const;
-};
-
-// Low level: get the selections
-std::unique_ptr<tls_selection> parse_tls_selection_details(const std::string &program, const std::string &selection);
-
-} // namespace cif
+#warning "This code has been moved to libpdb-redo"

include/cif++/point.hpp

@@ -26,8 +26,6 @@
 
 #pragma once
 
-#include "cif++/exports.hpp"
-
 #include <array>
 #include <cmath>
 #include <complex>
@@ -37,27 +35,55 @@
 
 #if __has_include(<clipper/core/coords.h>)
 #define HAVE_LIBCLIPPER 1
+#pragma GCC diagnostic push
+#pragma GCC diagnostic ignored "-Wignored-qualifiers"
 #include <clipper/core/coords.h>
+#pragma GCC diagnostic pop
 #endif
 
+/** \file point.hpp
+ *
+ * This file contains the definition for *cif::point* as well as
+ * lots of routines and classes that can manipulate points.
+ */
+
 namespace cif
 {
 
 // --------------------------------------------------------------------
 
+/// \brief Our value for Pi
 const double
 	kPI = 3.141592653589793238462643383279502884;
 
 // --------------------------------------------------------------------
-// A stripped down quaternion implementation, based on boost::math::quaternion
-// We use quaternions to do rotations in 3d space
+/**
+ * @brief A stripped down quaternion implementation, based on boost::math::quaternion
+ *
+ * We use quaternions to do rotations in 3d space. Quaternions are faster than
+ * matrix calculations and they also suffer less from drift caused by rounding
+ * errors.
+ *
+ * Like complex number, quaternions do have a meaningful notion of "real part",
+ * but unlike them there is no meaningful notion of "imaginary part".
+ * Instead there is an "unreal part" which itself is a quaternion, and usually
+ * nothing simpler (as opposed to the complex number case).
+ * However, for practicality, there are accessors for the other components
+ * (these are necessary for the templated copy constructor, for instance).
+ *
+ * @note Quaternion multiplication is *NOT* commutative;
+ * symbolically, "q *= rhs;" means "q = q * rhs;"
+ * and "q /= rhs;" means "q = q * inverse_of(rhs);"
+ */
 
 template <typename T>
 class quaternion_type
 {
   public:
+	/// \brief the value type of the elements, usually this is float
 	using value_type = T;
 
+	/// \brief constructor with the four members
 	constexpr explicit quaternion_type(value_type const &value_a = {}, value_type const &value_b = {}, value_type const &value_c = {}, value_type const &value_d = {})
 		: a(value_a)
 		, b(value_b)
@@ -66,6 +92,7 @@ class quaternion_type
 	{
 	}
 
+	/// \brief constructor taking two complex values as input
 	constexpr explicit quaternion_type(std::complex<value_type> const &z0, std::complex<value_type> const &z1 = std::complex<value_type>())
 		: a(z0.real())
 		, b(z0.imag())
@@ -74,9 +101,10 @@ class quaternion_type
 	{
 	}
 
-	constexpr quaternion_type(quaternion_type const &) = default;
-	constexpr quaternion_type(quaternion_type &&) = default;
+	constexpr quaternion_type(quaternion_type const &) = default; ///< Copy constructor
+	constexpr quaternion_type(quaternion_type &&) = default;      ///< Copy constructor
 
+	/// \brief Copy constructor accepting a quaternion with a different value_type
 	template <typename X>
 	constexpr explicit quaternion_type(quaternion_type<X> const &rhs)
 		: a(static_cast<value_type>(rhs.a))
@@ -87,24 +115,20 @@ class quaternion_type
 	}
 
 	// accessors
-	//
-	// Note:    Like complex number, quaternions do have a meaningful notion of "real part",
-	//            but unlike them there is no meaningful notion of "imaginary part".
-	//            Instead there is an "unreal part" which itself is a quaternion, and usually
-	//            nothing simpler (as opposed to the complex number case).
-	//            However, for practicality, there are accessors for the other components
-	//            (these are necessary for the templated copy constructor, for instance).
 
+	/// \brief See class description, return the *real* part of the quaternion
 	constexpr value_type real() const
 	{
 		return a;
 	}
 
+	/// \brief See class description, return the *unreal* part of the quaternion
 	constexpr quaternion_type unreal() const
 	{
 		return { 0, b, c, d };
 	}
 
+	/// \brief swap
 	constexpr void swap(quaternion_type &o)
 	{
 		std::swap(a, o.a);
@@ -115,6 +139,7 @@ class quaternion_type
 
 	// assignment operators
 
+	/// \brief Assignment operator accepting a quaternion with optionally another value_type
 	template <typename X>
 	constexpr quaternion_type &operator=(quaternion_type<X> const &rhs)
 	{
@@ -126,6 +151,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief Assignment operator
 	constexpr quaternion_type &operator=(quaternion_type const &rhs)
 	{
 		a = rhs.a;
@@ -136,6 +162,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief Assignment operator that sets the *real* part to @a rhs and the *unreal* parts to zero
 	constexpr quaternion_type &operator=(value_type const &rhs)
 	{
 		a = rhs;
@@ -145,6 +172,9 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief Assignment operator that sets the *real* part to the real part of @a rhs
+	/// and the first *unreal* part to the imaginary part of of @a rhs. The other *unreal*
+	// parts are set to zero.
 	constexpr quaternion_type &operator=(std::complex<value_type> const &rhs)
 	{
 		a = rhs.real();
@@ -156,17 +186,16 @@ class quaternion_type
 	}
 
 	// other assignment-related operators
-	//
-	// NOTE:    Quaternion multiplication is *NOT* commutative;
-	//            symbolically, "q *= rhs;" means "q = q * rhs;"
-	//            and "q /= rhs;" means "q = q * inverse_of(rhs);"
 
+	/// \brief operator += adding value @a rhs to the *real* part
 	constexpr quaternion_type &operator+=(value_type const &rhs)
 	{
 		a += rhs;
 		return *this;
 	}
 
+	/// \brief operator += adding the real part of @a rhs to the *real* part
+	/// and the imaginary part of @a rhs to the first *unreal* part
 	constexpr quaternion_type &operator+=(std::complex<value_type> const &rhs)
 	{
 		a += std::real(rhs);
@@ -174,6 +203,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief operator += adding the parts of @a rhs to the equivalent part of this
 	template <class X>
 	constexpr quaternion_type &operator+=(quaternion_type<X> const &rhs)
 	{
@@ -184,12 +214,15 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief operator -= subtracting value @a rhs from the *real* part
 	constexpr quaternion_type &operator-=(value_type const &rhs)
 	{
 		a -= rhs;
 		return *this;
 	}
 
+	/// \brief operator -= subtracting the real part of @a rhs from the *real* part
+	/// and the imaginary part of @a rhs from the first *unreal* part
 	constexpr quaternion_type &operator-=(std::complex<value_type> const &rhs)
 	{
 		a -= std::real(rhs);
@@ -197,6 +230,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief operator -= subtracting the parts of @a rhs from the equivalent part of this
 	template <class X>
 	constexpr quaternion_type &operator-=(quaternion_type<X> const &rhs)
 	{
@@ -207,6 +241,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief multiply all parts with value @a rhs
 	constexpr quaternion_type &operator*=(value_type const &rhs)
 	{
 		a *= rhs;
@@ -216,6 +251,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief multiply with complex number @a rhs
 	constexpr quaternion_type &operator*=(std::complex<value_type> const &rhs)
 	{
 		value_type ar = rhs.real();
@@ -225,13 +261,15 @@ class quaternion_type
 		return *this;
 	}
 
-	constexpr friend quaternion_type operator*(const quaternion_type &a, const quaternion_type &b)
+	/// \brief multiply @a a with @a b and return the result
+	friend constexpr quaternion_type operator*(const quaternion_type &a, const quaternion_type &b)
 	{
 		auto result = a;
 		result *= b;
 		return result;
 	}
 
+	/// \brief multiply with quaternion @a rhs
 	template <typename X>
 	constexpr quaternion_type &operator*=(quaternion_type<X> const &rhs)
 	{
@@ -245,6 +283,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief divide all parts by @a rhs
 	constexpr quaternion_type &operator/=(value_type const &rhs)
 	{
 		a /= rhs;
@@ -254,6 +293,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief divide by complex number @a rhs
 	constexpr quaternion_type &operator/=(std::complex<value_type> const &rhs)
 	{
 		value_type ar = rhs.real();
@@ -264,6 +304,7 @@ class quaternion_type
 		return *this;
 	}
 
+	/// \brief divide by quaternion @a rhs
 	template <typename X>
 	constexpr quaternion_type &operator/=(quaternion_type<X> const &rhs)
 	{
@@ -278,7 +319,8 @@ class quaternion_type
 		return *this;
 	}
 
-	constexpr friend quaternion_type normalize(quaternion_type q)
+	/// \brief normalise the values so that the length of the result is exactly 1
+	friend constexpr quaternion_type normalize(quaternion_type q)
 	{
 		std::valarray<value_type> t(4);
 
@@ -299,26 +341,30 @@ class quaternion_type
 		return q;
 	}
 
-	constexpr friend quaternion_type conj(quaternion_type q)
+	/// \brief return the conjugate of this
+	friend constexpr quaternion_type conj(quaternion_type q)
 	{
 		return quaternion_type{ +q.a, -q.b, -q.c, -q.d };
 	}
 
-	constexpr value_type get_a() const { return a; }
-	constexpr value_type get_b() const { return b; }
-	constexpr value_type get_c() const { return c; }
-	constexpr value_type get_d() const { return d; }
+	constexpr value_type get_a() const { return a; } ///< Return part a
+	constexpr value_type get_b() const { return b; } ///< Return part b
+	constexpr value_type get_c() const { return c; } ///< Return part c
+	constexpr value_type get_d() const { return d; } ///< Return part d
 
+	/// \brief compare with @a rhs
 	constexpr bool operator==(const quaternion_type &rhs) const
 	{
 		return a == rhs.a and b == rhs.b and c == rhs.c and d == rhs.d;
 	}
 
+	/// \brief compare with @a rhs
 	constexpr bool operator!=(const quaternion_type &rhs) const
 	{
 		return a != rhs.a or b != rhs.b or c != rhs.c or d != rhs.d;
 	}
 
+	/// \brief test for all zero values
 	constexpr operator bool() const
 	{
 		return a != 0 or b != 0 or c != 0 or d != 0;
@@ -328,6 +374,19 @@ class quaternion_type
 	value_type a, b, c, d;
 };
 
+/**
+ * @brief This code is similar to the code in boost so I copy the documentation as well:
+ *
+ * > spherical is a simple transposition of polar, it takes as inputs a (positive)
+ * > magnitude and a point on the hypersphere, given by three angles. The first of
+ * > these, theta has a natural range of -pi to +pi, and the other two have natural
+ * > ranges of -pi/2 to +pi/2 (as is the case with the usual spherical coordinates in
+ * > **R**<sup>3</sup>). Due to the many symmetries and periodicities, nothing untoward happens if
+ * > the magnitude is negative or the angles are outside their natural ranges. The
+ * > expected degeneracies (a magnitude of zero ignores the angles settings...) do
+ * > happen however.
+ */
+
 template <typename T>
 inline quaternion_type<T> spherical(T const &rho, T const &theta, T const &phi1, T const &phi2)
 {
@@ -345,24 +404,34 @@ inline quaternion_type<T> spherical(T const &rho, T const &theta, T const &phi1,
 	return result;
 }
 
+/// \brief By default we use the float version of a quaternion
 using quaternion = quaternion_type<float>;
 
 // --------------------------------------------------------------------
 
-//	point, a location with x, y and z coordinates as floating point.
-//	This one is derived from a tuple<float,float,float> so
-//	you can do things like:
-//
-//	float x, y, z;
-//	tie(x, y, z) = atom.loc();
+/**
+ * @brief 3D point: a location with x, y and z coordinates as floating point.
+ *
+ * Note that you can simply use structured binding to get access to the
+ * individual parts like so:
+ *
+ * @code{.cpp}
+ * float x, y, z;
+ * tie(x, y, z) = atom.get_location();
+ * @endcode
+ */
 
 template <typename F>
 struct point_type
 {
+	/// \brief the value type of the x, y and z members
 	using value_type = F;
 
-	value_type m_x, m_y, m_z;
+	value_type m_x, ///< The x part of the location
+		m_y,        ///< The y part of the location
+		m_z;        ///< The z part of the location
 
+	/// \brief default constructor, initialises the values to zero
 	constexpr point_type()
 		: m_x(0)
 		, m_y(0)
@@ -370,6 +439,7 @@ struct point_type
 	{
 	}
 
+	/// \brief constructor taking three values
 	constexpr point_type(value_type x, value_type y, value_type z)
 		: m_x(x)
 		, m_y(y)
@@ -377,6 +447,7 @@ struct point_type
 	{
 	}
 
+	/// \brief Copy constructor
 	template <typename PF>
 	constexpr point_type(const point_type<PF> &pt)
 		: m_x(static_cast<F>(pt.m_x))
@@ -385,12 +456,14 @@ struct point_type
 	{
 	}
 
+	/// \brief constructor taking a tuple of three values
 	constexpr point_type(const std::tuple<value_type, value_type, value_type> &pt)
 		: point_type(std::get<0>(pt), std::get<1>(pt), std::get<2>(pt))
 	{
 	}
 
 #if HAVE_LIBCLIPPER
+	/// \brief Construct a point using the values in clipper coordinate @a pt
 	constexpr point_type(const clipper::Coord_orth &pt)
 		: m_x(pt[0])
 		, m_y(pt[1])
@@ -398,6 +471,7 @@ struct point_type
 	{
 	}
 
+	/// \brief Assign a point using the values in clipper coordinate @a rhs
 	constexpr point_type &operator=(const clipper::Coord_orth &rhs)
 	{
 		m_x = rhs[0];
@@ -407,6 +481,7 @@ struct point_type
 	}
 #endif
 
+	/// \brief Assignment operator
 	template <typename PF>
 	constexpr point_type &operator=(const point_type<PF> &rhs)
 	{
@@ -416,18 +491,19 @@ struct point_type
 		return *this;
 	}
 
-	constexpr value_type &get_x() { return m_x; }
-	constexpr value_type get_x() const { return m_x; }
-	constexpr void set_x(value_type x) { m_x = x; }
+	constexpr value_type &get_x() { return m_x; }      ///< Get a reference to x
+	constexpr value_type get_x() const { return m_x; } ///< Get the value of x
+	constexpr void set_x(value_type x) { m_x = x; }    ///< Set the value of x to @a x
 
-	constexpr value_type &get_y() { return m_y; }
-	constexpr value_type get_y() const { return m_y; }
-	constexpr void set_y(value_type y) { m_y = y; }
+	constexpr value_type &get_y() { return m_y; }      ///< Get a reference to y
+	constexpr value_type get_y() const { return m_y; } ///< Get the value of y
+	constexpr void set_y(value_type y) { m_y = y; }    ///< Set the value of y to @a y
 
-	constexpr value_type &get_z() { return m_z; }
-	constexpr value_type get_z() const { return m_z; }
-	constexpr void set_z(value_type z) { m_z = z; }
+	constexpr value_type &get_z() { return m_z; }      ///< Get a reference to z
+	constexpr value_type get_z() const { return m_z; } ///< Get the value of z
+	constexpr void set_z(value_type z) { m_z = z; }    ///< Set the value of z to @a z
 
+	/// \brief add @a rhs
 	constexpr point_type &operator+=(const point_type &rhs)
 	{
 		m_x += rhs.m_x;
@@ -437,6 +513,7 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief add @a d to all members
 	constexpr point_type &operator+=(value_type d)
 	{
 		m_x += d;
@@ -446,6 +523,14 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief Add the points @a lhs and @a rhs and return the result
+	template <typename F2>
+	friend constexpr auto operator+(const point_type &lhs, const point_type<F2> &rhs)
+	{
+		return point_type<std::common_type_t<value_type, F2>>(lhs.m_x + rhs.m_x, lhs.m_y + rhs.m_y, lhs.m_z + rhs.m_z);
+	}
+
+	/// \brief subtract @a rhs
 	constexpr point_type &operator-=(const point_type &rhs)
 	{
 		m_x -= rhs.m_x;
@@ -455,6 +540,7 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief subtract @a d from all members
 	constexpr point_type &operator-=(value_type d)
 	{
 		m_x -= d;
@@ -464,6 +550,20 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief Subtract the points @a lhs and @a rhs and return the result
+	template <typename F2>
+	friend constexpr auto operator-(const point_type &lhs, const point_type<F2> &rhs)
+	{
+		return point_type<std::common_type_t<value_type, F2>>(lhs.m_x - rhs.m_x, lhs.m_y - rhs.m_y, lhs.m_z - rhs.m_z);
+	}
+
+	/// \brief Return the negative copy of @a pt
+	friend constexpr point_type operator-(const point_type &pt)
+	{
+		return point_type(-pt.m_x, -pt.m_y, -pt.m_z);
+	}
+
+	/// \brief multiply all members with @a rhs
 	constexpr point_type &operator*=(value_type rhs)
 	{
 		m_x *= rhs;
@@ -472,6 +572,21 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief multiply point @a pt with value @a f and return the result
+	template <typename F2>
+	friend constexpr auto operator*(const point_type &pt, F2 f)
+	{
+		return point_type<std::common_type_t<value_type, F2>>(pt.m_x * f, pt.m_y * f, pt.m_z * f);
+	}
+
+	/// \brief multiply point @a pt with value @a f and return the result
+	template <typename F2>
+	friend constexpr auto operator*(F2 f, const point_type &pt)
+	{
+		return point_type<std::common_type_t<value_type, F2>>(pt.m_x * f, pt.m_y * f, pt.m_z * f);
+	}
+
+	/// \brief divide all members by @a rhs
 	constexpr point_type &operator/=(value_type rhs)
 	{
 		m_x /= rhs;
@@ -480,6 +595,20 @@ struct point_type
 		return *this;
 	}
 
+	/// \brief divide point @a pt by value @a f and return the result
+	template <typename F2>
+	friend constexpr auto operator/(const point_type &pt, F2 f)
+	{
+		return point_type<std::common_type_t<value_type, F2>>(pt.m_x / f, pt.m_y / f, pt.m_z / f);
+	}
+
+	/**
+	 * @brief looking at this point as a vector, normalise it which
+	 * means dividing all members by the length making the length
+	 * effectively 1.
+	 *
+	 * @return The previous length of this vector
+	 */
 	constexpr value_type normalize()
 	{
 		auto length = m_x * m_x + m_y * m_y + m_z * m_z;
@@ -491,6 +620,7 @@ struct point_type
 		return length;
 	}
 
+	/// \brief Rotate this point using the quaterion @a q
 	constexpr void rotate(const quaternion &q)
 	{
 		quaternion_type<value_type> p(0, m_x, m_y, m_z);
@@ -502,6 +632,9 @@ struct point_type
 		m_z = p.get_d();
 	}
 
+	/// \brief Rotate this point using the quaterion @a q by first
+	/// moving the point to @a pivot and after rotating moving it
+	/// back
 	constexpr void rotate(const quaternion &q, point_type pivot)
 	{
 		operator-=(pivot);
@@ -510,97 +643,82 @@ struct point_type
 	}
 
 #if HAVE_LIBCLIPPER
+	/// \brief Make it possible to pass a point to clipper functions expecting a clipper coordinate
 	operator clipper::Coord_orth() const
 	{
 		return clipper::Coord_orth(m_x, m_y, m_z);
 	}
 #endif
 
+	/// \brief Allow access to this point as if it is a tuple of three const value_type's
 	constexpr operator std::tuple<const value_type &, const value_type &, const value_type &>() const
 	{
 		return std::make_tuple(std::ref(m_x), std::ref(m_y), std::ref(m_z));
 	}
 
+	/// \brief Allow access to this point as if it is a tuple of three value_type's
 	constexpr operator std::tuple<value_type &, value_type &, value_type &>()
 	{
 		return std::make_tuple(std::ref(m_x), std::ref(m_y), std::ref(m_z));
 	}
 
+#if defined(__cpp_impl_three_way_comparison)
+	/// \brief a default spaceship operator
+	constexpr auto operator<=>(const point_type &rhs) const = default;
+#else
+	/// \brief a default equals operator
 	constexpr bool operator==(const point_type &rhs) const
 	{
 		return m_x == rhs.m_x and m_y == rhs.m_y and m_z == rhs.m_z;
 	}
 
+	/// \brief a default not-equals operator
+	constexpr bool operator!=(const point_type &rhs) const
+	{
+		return not operator==(rhs);
+	}
+#endif
+
 	// consider point as a vector... perhaps I should rename point?
+
+	/// \brief looking at the point as if it is a vector, return the squared length
 	constexpr value_type length_sq() const
 	{
 		return m_x * m_x + m_y * m_y + m_z * m_z;
 	}
 
+	/// \brief looking at the point as if it is a vector, return the length
 	constexpr value_type length() const
 	{
-		return std::sqrt(m_x * m_x + m_y * m_y + m_z * m_z);
+		return std::sqrt(length_sq());
+	}
+
+	/// \brief Print out the point @a pt to @a os
+	friend std::ostream &operator<<(std::ostream &os, const point_type &pt)
+	{
+		os << '(' << pt.m_x << ',' << pt.m_y << ',' << pt.m_z << ')';
+		return os;
 	}
 };
 
+/// \brief By default we use points with float value_type
 using point = point_type<float>;
 
-template <typename F>
-inline constexpr std::ostream &operator<<(std::ostream &os, const point_type<F> &pt)
-{
-	os << '(' << pt.m_x << ',' << pt.m_y << ',' << pt.m_z << ')';
-	return os;
-}
-
-template <typename F>
-inline constexpr point_type<F> operator+(const point_type<F> &lhs, const point_type<F> &rhs)
-{
-	return point_type<F>(lhs.m_x + rhs.m_x, lhs.m_y + rhs.m_y, lhs.m_z + rhs.m_z);
-}
-
-template <typename F>
-inline constexpr point_type<F> operator-(const point_type<F> &lhs, const point_type<F> &rhs)
-{
-	return point_type<F>(lhs.m_x - rhs.m_x, lhs.m_y - rhs.m_y, lhs.m_z - rhs.m_z);
-}
-
-template <typename F>
-inline constexpr point_type<F> operator-(const point_type<F> &pt)
-{
-	return point_type<F>(-pt.m_x, -pt.m_y, -pt.m_z);
-}
-
-template <typename F>
-inline constexpr point_type<F> operator*(const point_type<F> &pt, F f)
-{
-	return point_type<F>(pt.m_x * f, pt.m_y * f, pt.m_z * f);
-}
-
-template <typename F>
-inline constexpr point_type<F> operator*(F f, const point_type<F> &pt)
-{
-	return point_type<F>(pt.m_x * f, pt.m_y * f, pt.m_z * f);
-}
-
-template <typename F>
-inline constexpr point_type<F> operator/(const point_type<F> &pt, F f)
-{
-	return point_type<F>(pt.m_x / f, pt.m_y / f, pt.m_z / f);
-}
-
 // --------------------------------------------------------------------
 // several standard 3d operations
 
-template <typename F>
-inline constexpr auto distance_squared(const point_type<F> &a, const point_type<F> &b)
+/// \brief return the squared distance between points @a a and @a b
+template <typename F1, typename F2>
+constexpr auto distance_squared(const point_type<F1> &a, const point_type<F2> &b)
 {
 	return (a.m_x - b.m_x) * (a.m_x - b.m_x) +
 	       (a.m_y - b.m_y) * (a.m_y - b.m_y) +
 	       (a.m_z - b.m_z) * (a.m_z - b.m_z);
 }
 
-template <typename F>
-inline constexpr auto distance(const point_type<F> &a, const point_type<F> &b)
+/// \brief return the distance between points @a a and @a b
+template <typename F1, typename F2>
+constexpr auto distance(const point_type<F1> &a, const point_type<F2> &b)
 {
 	return std::sqrt(
 		(a.m_x - b.m_x) * (a.m_x - b.m_x) +
@@ -608,20 +726,24 @@ inline constexpr auto distance(const point_type<F> &a, const point_type<F> &b)
 		(a.m_z - b.m_z) * (a.m_z - b.m_z));
 }
 
-template <typename F>
-inline constexpr auto dot_product(const point_type<F> &a, const point_type<F> &b)
+/// \brief return the dot product between the vectors @a a and @a b
+template <typename F1, typename F2>
+inline constexpr auto dot_product(const point_type<F1> &a, const point_type<F2> &b)
 {
 	return a.m_x * b.m_x + a.m_y * b.m_y + a.m_z * b.m_z;
 }
 
-template <typename F>
-inline constexpr point_type<F> cross_product(const point_type<F> &a, const point_type<F> &b)
+/// \brief return the cross product between the vectors @a a and @a b
+template <typename F1, typename F2>
+inline constexpr auto cross_product(const point_type<F1> &a, const point_type<F2> &b)
 {
-	return point_type<F>(a.m_y * b.m_z - b.m_y * a.m_z,
+	return point_type<std::common_type_t<F1, F2>>(
+		a.m_y * b.m_z - b.m_y * a.m_z,
 		a.m_z * b.m_x - b.m_z * a.m_x,
 		a.m_x * b.m_y - b.m_x * a.m_y);
 }
 
+/// \brief return the angle in degrees between the vectors from point @a p2 to @a p1 and @a p2 to @a p3
 template <typename F>
 constexpr auto angle(const point_type<F> &p1, const point_type<F> &p2, const point_type<F> &p3)
 {
@@ -631,6 +753,9 @@ constexpr auto angle(const point_type<F> &p1, const point_type<F> &p2, const poi
 	return std::acos(dot_product(v1, v2) / (v1.length() * v2.length())) * 180 / kPI;
 }
 
+/// \brief return the dihedral angle in degrees for the four points @a p1, @a p2, @a p3 and @a p4
+///
+/// See https://en.wikipedia.org/wiki/Dihedral_angle for an explanation of what a dihedral angle is
 template <typename F>
 constexpr auto dihedral_angle(const point_type<F> &p1, const point_type<F> &p2, const point_type<F> &p3, const point_type<F> &p4)
 {
@@ -658,6 +783,7 @@ constexpr auto dihedral_angle(const point_type<F> &p1, const point_type<F> &p2,
 	return result;
 }
 
+/// \brief return the cosinus angle for the four points @a p1, @a p2, @a p3 and @a p4
 template <typename F>
 constexpr auto cosinus_angle(const point_type<F> &p1, const point_type<F> &p2, const point_type<F> &p3, const point_type<F> &p4)
 {
@@ -669,6 +795,7 @@ constexpr auto cosinus_angle(const point_type<F> &p1, const point_type<F> &p2, c
 	return x > 0 ? dot_product(v12, v34) / std::sqrt(x) : 0;
 }
 
+/// \brief return the distance from point @a p to the line from @a l1 to @a l2
 template <typename F>
 constexpr auto distance_point_to_line(const point_type<F> &l1, const point_type<F> &l2, const point_type<F> &p)
 {
@@ -680,15 +807,19 @@ constexpr auto distance_point_to_line(const point_type<F> &l1, const point_type<
 }
 
 // --------------------------------------------------------------------
-// For e.g. simulated annealing, returns a new point that is moved in
-// a random direction with a distance randomly chosen from a normal
-// distribution with a stddev of offset.
-
+/**
+ * @brief For e.g. simulated annealing, returns a new point that is moved in
+ * a random direction with a distance randomly chosen from a normal
+ * distribution with a stddev of offset.
+ */
 point nudge(point p, float offset);
 
 // --------------------------------------------------------------------
 
+/// \brief Return a quaternion created from angle @a angle and axis @a axis
 quaternion construct_from_angle_axis(float angle, point axis);
+
+/// \brief Return a tuple of an angle and an axis for quaternion @a q
 std::tuple<double, point> quaternion_to_angle_axis(quaternion q);
 
 /// @brief Given four points and an angle, return the quaternion required to rotate
@@ -697,8 +828,12 @@ std::tuple<double, point> quaternion_to_angle_axis(quaternion q);
 quaternion construct_for_dihedral_angle(point p1, point p2, point p3, point p4,
 	float angle, float esd);
 
-point centroid(const std::vector<point> &Points);
-point center_points(std::vector<point> &Points);
+/// \brief Return the point that is the centroid of all the points in @a pts
+point centroid(const std::vector<point> &pts);
+
+/// \brief Move all the points in @a pts so that their centroid is at the origin
+/// (0, 0, 0) and return the offset used (the former centroid)
+point center_points(std::vector<point> &pts);
 
 /// \brief Returns how the two sets of points \a a and \b b can be aligned
 ///
@@ -712,38 +847,56 @@ quaternion align_points(const std::vector<point> &a, const std::vector<point> &b
 double RMSd(const std::vector<point> &a, const std::vector<point> &b);
 
 // --------------------------------------------------------------------
-// Helper class to generate evenly divided points on a sphere
-// we use a fibonacci sphere to calculate even distribution of the dots
-
+/**
+ * @brief Helper class to generate evenly divided points on a sphere
+ *
+ * We use a fibonacci sphere to calculate even distribution of the dots
+ *
+ * @tparam N The number of points on the sphere is 2 * N + 1
+ */
 template <int N>
 class spherical_dots
 {
   public:
+	/// \brief the number of points
 	constexpr static int P = 2 * N * 1;
 
+	/// \brief the *weight* of the fibonacci sphere
+	constexpr static double W = (4 * kPI) / P;
+
+	/// \brief the internal storage type
 	using array_type = typename std::array<point, P>;
+
+	/// \brief iterator type
 	using iterator = typename array_type::const_iterator;
 
+	/// \brief singleton instance
 	static spherical_dots &instance()
 	{
 		static spherical_dots sInstance;
 		return sInstance;
 	}
 
-	size_t size() const { return m_points.size(); }
+	/// \brief The number of points
+	std::size_t size() const { return P; }
+
+	/// \brief Access a point by index
 	const point operator[](uint32_t inIx) const { return m_points[inIx]; }
+
+	/// \brief iterator pointing to the first point
 	iterator begin() const { return m_points.begin(); }
+
+	/// \brief iterator pointing past the last point
 	iterator end() const { return m_points.end(); }
 
-	double weight() const { return m_weight; }
+	/// \brief return the *weight*,
+	double weight() const { return W; }
 
 	spherical_dots()
 	{
 		const double
 			kGoldenRatio = (1 + std::sqrt(5.0)) / 2;
 
-		m_weight = (4 * kPI) / P;
-
 		auto p = m_points.begin();
 
 		for (int32_t i = -N; i <= N; ++i)
@@ -761,7 +914,6 @@ class spherical_dots
 
   private:
 	array_type m_points;
-	double m_weight;
 };
 
 } // namespace cif

include/cif++/row.hpp

@@ -30,6 +30,51 @@
 
 #include <array>
 
+/**
+ * @file row.hpp
+ * 
+ * The class cif::row should be an opaque type. It is used to store the
+ * internal data per row in a category. You should use cif::row_handle
+ * to get access to the contents in a row.
+ * 
+ * One could think of rows as vectors of cif::item. But internally
+ * that's not the case.
+ * 
+ * You can access the values of stored items by name or index.
+ * The return value of operator[] is an cif::item_handle object.
+ * 
+ * @code {.cpp}
+ * cif::category &atom_site = my_db["atom_site"];
+ * cif::row_handle rh = atom_site.front();
+ * 
+ * // by name:
+ * std::string name = rh["label_atom_id"].as<std::string>();
+ * 
+ * // by index:
+ * uint16_t ix = atom_site.get_item_ix("label_atom_id");
+ * assert(rh[ix].as<std::string() == name);
+ * @endcode
+ * 
+ * There some template magic here to allow easy extracting of data
+ * from rows. This can be done using cif::tie e.g.:
+ * 
+ * @code {.cpp}
+ * std::string name;
+ * float x, y, z;
+ * 
+ * cif::tie(name, x, y, z) = rh.get("label_atom_id", "cartn_x", "cartn_y", "cartn_z");
+ * @endcode
+ * 
+ * However, a more modern way uses structured binding:
+ * 
+ * @code {.cpp}
+ * const auto &[name, x, y, z] = rh.get<std::string,float,float,float>("label_atom_id", "cartn_x", "cartn_y", "cartn_z");
+ * @endcode
+ * 
+ * 
+ * 
+ */
+
 namespace cif
 {
 
@@ -40,17 +85,17 @@ namespace detail
 	template <typename... C>
 	struct get_row_result
 	{
-		static constexpr size_t N = sizeof...(C);
+		static constexpr std::size_t N = sizeof...(C);
 
-		get_row_result(const row_handle &r, std::array<uint16_t, N> &&columns)
+		get_row_result(const row_handle &r, std::array<uint16_t, N> &&items)
 			: m_row(r)
-			, m_columns(std::move(columns))
+			, m_items(std::move(items))
 		{
 		}
 
 		const item_handle operator[](uint16_t ix) const
 		{
-			return m_row[m_columns[ix]];
+			return m_row[m_items[ix]];
 		}
 
 		template <typename... Ts, std::enable_if_t<N == sizeof...(Ts), int> = 0>
@@ -59,14 +104,14 @@ namespace detail
 			return get<Ts...>(std::index_sequence_for<Ts...>{});
 		}
 
-		template <typename... Ts, size_t... Is>
+		template <typename... Ts, std::size_t... Is>
 		std::tuple<Ts...> get(std::index_sequence<Is...>) const
 		{
-			return std::tuple<Ts...>{ m_row[m_columns[Is]].template as<Ts>()... };
+			return std::tuple<Ts...>{ m_row[m_items[Is]].template as<Ts>()... };
 		}
 
 		const row_handle &m_row;
-		std::array<uint16_t, N> m_columns;
+		std::array<uint16_t, N> m_items;
 	};
 
 	// we want to be able to tie some variables to a get_row_result, for this we use tiewraps
@@ -96,6 +141,8 @@ namespace detail
 
 } // namespace detail
 
+/// \brief similar to std::tie, assign values to each element in @a v from the 
+/// result of a get on a row_handle.
 template <typename... Ts>
 auto tie(Ts &...v)
 {
@@ -110,11 +157,17 @@ class row : public std::vector<item_value>
   public:
 	row() = default;
 
+	/**
+	 * @brief Return the item_value pointer for item at index @a ix
+	 */
 	item_value* get(uint16_t ix)
 	{
 		return ix < size() ? &data()[ix] : nullptr;
 	}
 
+	/**
+	 * @brief Return the const item_value pointer for item at index @a ix
+	 */
 	const item_value* get(uint16_t ix) const
 	{
 		return ix < size() ? &data()[ix] : nullptr;
@@ -150,10 +203,12 @@ class row : public std::vector<item_value>
 class row_handle
 {
   public:
+	/** @cond */
 	friend struct item_handle;
 	friend class category;
 	friend class category_index;
 	friend class row_initializer;
+	template <typename, typename...> friend class iterator_impl;
 
 	row_handle() = default;
 
@@ -163,86 +218,133 @@ class row_handle
 	row_handle &operator=(const row_handle &) = default;
 	row_handle &operator=(row_handle &&) = default;
 
+	/** @endcond */
+
+	/// \brief constructor taking a category @a cat and a row @a r
 	row_handle(const category &cat, const row &r)
 		: m_category(const_cast<category *>(&cat))
 		, m_row(const_cast<row *>(&r))
 	{
 	}
 
+	/// \brief return the category this row belongs to
 	const category &get_category() const
 	{
 		return *m_category;
 	}
 
+	/// \brief Return true if the row is empty or uninitialised
 	bool empty() const
 	{
 		return m_category == nullptr or m_row == nullptr;
 	}
 
+	/// \brief convenience method to test for empty()
 	explicit operator bool() const
 	{
 		return not empty();
 	}
 
-	item_handle operator[](uint16_t column_ix)
+	/// \brief return a cif::item_handle to the item in item @a item_ix
+	item_handle operator[](uint16_t item_ix)
 	{
-		return empty() ? item_handle::s_null_item : item_handle(column_ix, *this);
+		return empty() ? item_handle::s_null_item : item_handle(item_ix, *this);
 	}
 
-	const item_handle operator[](uint16_t column_ix) const
+	/// \brief return a const cif::item_handle to the item in item @a item_ix
+	const item_handle operator[](uint16_t item_ix) const
 	{
-		return empty() ? item_handle::s_null_item : item_handle(column_ix, const_cast<row_handle &>(*this));
+		return empty() ? item_handle::s_null_item : item_handle(item_ix, const_cast<row_handle &>(*this));
 	}
 
-	item_handle operator[](std::string_view column_name)
+	/// \brief return a cif::item_handle to the item in the item named @a item_name
+	item_handle operator[](std::string_view item_name)
 	{
-		return empty() ? item_handle::s_null_item : item_handle(add_column(column_name), *this);
+		return empty() ? item_handle::s_null_item : item_handle(add_item(item_name), *this);
 	}
 
-	const item_handle operator[](std::string_view column_name) const
+	/// \brief return a const cif::item_handle to the item in the item named @a item_name
+	const item_handle operator[](std::string_view item_name) const
 	{
-		return empty() ? item_handle::s_null_item : item_handle(get_column_ix(column_name), const_cast<row_handle &>(*this));
+		return empty() ? item_handle::s_null_item : item_handle(get_item_ix(item_name), const_cast<row_handle &>(*this));
 	}
 
+	/// \brief Return an object that can be used in combination with cif::tie
+	/// to assign the values for the items @a items
 	template <typename... C>
-	auto get(C... columns) const
+	auto get(C... items) const
 	{
-		return detail::get_row_result<C...>(*this, { get_column_ix(columns)... });
+		return detail::get_row_result<C...>(*this, { get_item_ix(items)... });
 	}
 
+	/// \brief Return a tuple of values of types @a Ts for the items @a items
 	template <typename... Ts, typename... C, std::enable_if_t<sizeof...(Ts) == sizeof...(C) and sizeof...(C) != 1, int> = 0>
-	std::tuple<Ts...> get(C... columns) const
+	std::tuple<Ts...> get(C... items) const
 	{
-		return detail::get_row_result<Ts...>(*this, { get_column_ix(columns)... });
+		return detail::get_row_result<Ts...>(*this, { get_item_ix(items)... });
 	}
 
+	/// \brief Get the value of item @a item cast to type @a T
 	template <typename T>
-	T get(const char *column) const
+	T get(const char *item) const
 	{
-		return operator[](get_column_ix(column)).template as<T>();
+		return operator[](get_item_ix(item)).template as<T>();
 	}
 
+	/// \brief Get the value of item @a item cast to type @a T
+	template <typename T>
+	T get(std::string_view item) const
+	{
+		return operator[](get_item_ix(item)).template as<T>();
+	}
+
+	/// \brief assign each of the items named in @a values to their respective value
 	void assign(const std::vector<item> &values)
 	{
 		for (auto &value : values)
 			assign(value, true);
 	}
 
+	/** \brief assign the value @a value to the item named @a name 
+	 * 
+	 * If updateLinked it true, linked records are updated as well.
+	 * That means that if item @a name is part of the link definition
+	 * and the link results in a linked record in another category
+	 * this record in the linked category is updated as well.
+	 * 
+	 * If validate is true, which is default, the assigned value is
+	 * checked to see if it conforms to the rules defined in the dictionary
+	 */
+
 	void assign(std::string_view name, std::string_view value, bool updateLinked, bool validate = true)
 	{
-		assign(add_column(name), value, updateLinked, validate);
+		assign(add_item(name), value, updateLinked, validate);
 	}
 
-	void assign(uint16_t column, std::string_view value, bool updateLinked, bool validate = true);
+	/** \brief assign the value @a value to item at index @a item
+	 * 
+	 * If updateLinked it true, linked records are updated as well.
+	 * That means that if item @a item is part of the link definition
+	 * and the link results in a linked record in another category
+	 * this record in the linked category is updated as well.
+	 * 
+	 * If validate is true, which is default, the assigned value is
+	 * checked to see if it conforms to the rules defined in the dictionary
+	 */
 
+	void assign(uint16_t item, std::string_view value, bool updateLinked, bool validate = true);
+
+	/// \brief compare two rows
 	bool operator==(const row_handle &rhs) const { return m_category == rhs.m_category and m_row == rhs.m_row; }
+
+	/// \brief compare two rows
 	bool operator!=(const row_handle &rhs) const { return m_category != rhs.m_category or m_row != rhs.m_row; }
 
   private:
-	uint16_t get_column_ix(std::string_view name) const;
-	std::string_view get_column_name(uint16_t ix) const;
+	uint16_t get_item_ix(std::string_view name) const;
+	std::string_view get_item_name(uint16_t ix) const;
 
-	uint16_t add_column(std::string_view name);
+	uint16_t add_item(std::string_view name);
 
 	row *get_row()
 	{
@@ -259,7 +361,7 @@ class row_handle
 		assign(i.name(), i.value(), updateLinked);
 	}
 
-	void swap(uint16_t column, row_handle &r);
+	void swap(uint16_t item, row_handle &r);
 
 	category *m_category = nullptr;
 	row *m_row = nullptr;
@@ -267,9 +369,17 @@ class row_handle
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief The class row_initializer is a list of cif::item's.
+ * 
+ * This class is used to construct new rows, it allows to
+ * group a list of item name and value pairs and pass it
+ * in one go to the constructing function.
+ */
 class row_initializer : public std::vector<item>
 {
   public:
+	/** @cond */
 	friend class category;
 
 	row_initializer() = default;
@@ -278,26 +388,38 @@ class row_initializer : public std::vector<item>
 	row_initializer &operator=(const row_initializer &) = default;
 	row_initializer &operator=(row_initializer &&) = default;
 
+	/** @endcond */
+
+	/// \brief constructor taking a std::initializer_list of items
 	row_initializer(std::initializer_list<item> items)
 		: std::vector<item>(items)
 	{
 	}
 
+	/// \brief constructor taking a range of items
 	template <typename ItemIter, std::enable_if_t<std::is_same_v<typename ItemIter::value_type, item>, int> = 0>
 	row_initializer(ItemIter b, ItemIter e)
 		: std::vector<item>(b, e)
 	{
 	}
 
+	/// \brief constructor taking the values of an existing row
 	row_initializer(row_handle rh);
 
+
+	/// \brief set the value for item name @a name to @a value
 	void set_value(std::string_view name, std::string_view value);
+
+	/// \brief set the value for item based on @a i
 	void set_value(const item &i)
 	{
 		set_value(i.name(), i.value());
 	}
 
+	/// \brief set the value for item name @a name to @a value, but only if the item did not have a value already
 	void set_value_if_empty(std::string_view name, std::string_view value);
+
+	/// \brief set the value for item @a i, but only if the item did not have a value already
 	void set_value_if_empty(const item &i)
 	{
 		set_value_if_empty(i.name(), i.value());

include/cif++/symmetry.hpp

@@ -38,15 +38,18 @@
 #include <compare>
 #endif
 
-/// \file cif++/symmetry.hpp
-/// This file contains code to do symmetry operations based on the
-/// operations as specified in the International Tables.
+/** \file cif++/symmetry.hpp
+ *
+ * This file contains code to do symmetry operations based on the
+ * operations as specified in the International Tables.
+ */
 
 namespace cif
 {
 
 // --------------------------------------------------------------------
 
+/// \brief Apply matrix transformation @a m on point @a pt and return the result
 inline point operator*(const matrix3x3<float> &m, const point &pt)
 {
 	return {
@@ -58,28 +61,40 @@ inline point operator*(const matrix3x3<float> &m, const point &pt)
 
 // --------------------------------------------------------------------
 
+/// \brief the space groups we know
 enum class space_group_name
 {
-	full,
-	xHM,
-	Hall
+	full, ///< The *full* spacegroup
+	xHM,  ///< The *xHM* spacegroup
+	Hall  ///< The *Hall* spacegroup
 };
 
+/// \brief For each known spacegroup we define a structure like this
 struct space_group
 {
-	const char *name;
-	const char *xHM;
-	const char *Hall;
-	int nr;
+	const char *name; ///< The name according to *full*
+	const char *xHM;  ///< The name according to *xHM*
+	const char *Hall; ///< The name according to *Hall*
+	int nr;           ///< The number for this spacegroup
 };
 
+/// \brief Global list of spacegroups
 extern CIFPP_EXPORT const space_group kSpaceGroups[];
+
+/// \brief Global for the size of the list of spacegroups
 extern CIFPP_EXPORT const std::size_t kNrOfSpaceGroups;
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief Helper class to efficiently pack the data that
+ * makes up a symmetry operation
+ *
+ */
+
 struct symop_data
 {
+	/// \brief constructor
 	constexpr symop_data(const std::array<int, 15> &data)
 		: m_packed((data[0] bitand 0x03ULL) << 34 bitor
 				   (data[1] bitand 0x03ULL) << 32 bitor
@@ -99,27 +114,32 @@ struct symop_data
 	{
 	}
 
+	/// \brief compare
 	bool operator==(const symop_data &rhs) const
 	{
 		return m_packed == rhs.m_packed;
 	}
 
+	/// \brief sorting order
 	bool operator<(const symop_data &rhs) const
 	{
 		return m_packed < rhs.m_packed;
 	}
 
+	/// \brief return an int representing the value stored in the two bits at offset @a offset
 	inline constexpr int unpack3(int offset) const
 	{
 		int result = (m_packed >> offset) bitand 0x03;
 		return result == 3 ? -1 : result;
 	}
 
+	/// \brief return an int representing the value stored in the three bits at offset @a offset
 	inline constexpr int unpack7(int offset) const
 	{
 		return (m_packed >> offset) bitand 0x07;
 	}
 
+	/// \brief return an array of 15 ints representing the values stored
 	constexpr std::array<int, 15> data() const
 	{
 		return {
@@ -154,8 +174,14 @@ struct symop_data
 	uint64_t m_packed;
 };
 
+/**
+ * @brief For each symmetry operator defined in the international tables
+ * we have an entry in this struct type. It contains the spacegroup
+ * number, the symmetry operations and the rotational number.
+ */
 struct symop_datablock
 {
+	/// \brief constructor
 	constexpr symop_datablock(int spacegroup, int rotational_number, const std::array<int, 15> &rt_data)
 		: m_v((spacegroup bitand 0xffffULL) << 48 bitor
 			  (rotational_number bitand 0xffULL) << 40 bitor
@@ -163,9 +189,9 @@ struct symop_datablock
 	{
 	}
 
-	uint16_t spacegroup() const { return m_v >> 48; }
-	symop_data symop() const { return symop_data(m_v); }
-	uint8_t rotational_number() const { return (m_v >> 40) bitand 0xff; }
+	uint16_t spacegroup() const { return m_v >> 48; }                     ///< Return the spacegroup
+	symop_data symop() const { return symop_data(m_v); }                  ///< Return the symmetry operation
+	uint8_t rotational_number() const { return (m_v >> 40) bitand 0xff; } ///< Return the rotational_number
 
   private:
 	uint64_t m_v;
@@ -173,7 +199,10 @@ struct symop_datablock
 
 static_assert(sizeof(symop_datablock) == sizeof(uint64_t), "Size of symop_data is wrong");
 
+/// \brief Global containing the list of known symmetry operations
 extern CIFPP_EXPORT const symop_datablock kSymopNrTable[];
+
+/// \brief Size of the list of known symmetry operations
 extern CIFPP_EXPORT const std::size_t kSymopNrTableSize;
 
 // --------------------------------------------------------------------
@@ -186,13 +215,21 @@ class spacegroup;
 class rtop;
 struct sym_op;
 
+/** @brief A class that encapsulates the symmetry operations as used in PDB files,
+ * i.e. a rotational number and a translation vector.
+ *
+ * The syntax in string format follows the syntax as used in mmCIF files, i.e.
+ * rotational number followed by underscore and the three translations where 5 is
+ * no movement.
+ *
+ * So the string 1_555 means no symmetry movement at all since the rotational number
+ * 1 always corresponds to the symmetry operation [x, y, z].
+ */
 
-/// @brief A class that encapsulates the symmetry operations as used in PDB files, i.e. a rotational number and a translation vector
-/// The syntax in string format follows the syntax as used in mmCIF files, i.e. rotational number followed by underscore and the
-/// three translations where 5 is no movement.
 struct sym_op
 {
   public:
+	/// \brief constructor
 	sym_op(uint8_t nr = 1, uint8_t ta = 5, uint8_t tb = 5, uint8_t tc = 5)
 		: m_nr(nr)
 		, m_ta(ta)
@@ -201,48 +238,69 @@ struct sym_op
 	{
 	}
 
+	/// \brief construct a sym_op based on the contents encoded in string @a s
 	explicit sym_op(std::string_view s);
 
+	/** @cond */
 	sym_op(const sym_op &) = default;
 	sym_op(sym_op &&) = default;
 	sym_op &operator=(const sym_op &) = default;
 	sym_op &operator=(sym_op &&) = default;
+	/** @endcond */
 
+	/// \brief return true if this sym_op is the identity operator
 	constexpr bool is_identity() const
 	{
 		return m_nr == 1 and m_ta == 5 and m_tb == 5 and m_tc == 5;
 	}
 
+	/// \brief quick test for unequal to identity
 	explicit constexpr operator bool() const
 	{
 		return not is_identity();
 	}
 
+	/// \brief return the content encoded in a string
 	std::string string() const;
 
 #if defined(__cpp_impl_three_way_comparison)
+	/// \brief a default spaceship operator
 	constexpr auto operator<=>(const sym_op &rhs) const = default;
 #else
+	/// \brief a default equals operator
 	constexpr bool operator==(const sym_op &rhs) const
 	{
 		return m_nr == rhs.m_nr and m_ta == rhs.m_ta and m_tb == rhs.m_tb and m_tc == rhs.m_tc;
 	}
 
+	/// \brief a default not-equals operator
 	constexpr bool operator!=(const sym_op &rhs) const
 	{
 		return not operator==(rhs);
 	}
 #endif
 
+	/// @cond
 	uint8_t m_nr;
 	uint8_t m_ta, m_tb, m_tc;
+	/// @endcond
 };
 
 static_assert(sizeof(sym_op) == 4, "Sym_op should be four bytes");
 
 namespace literals
 {
-	inline sym_op operator""_symop(const char *text, size_t length)
+	/**
+	 * @brief This operator allows you to write code like this:
+	 *
+	 * @code {.cpp}
+	 * using namespace cif::literals;
+	 *
+	 * cif::sym_op so = "1_555"_symop;
+	 * @endcode
+	 *
+	 */
+	inline sym_op operator""_symop(const char *text, std::size_t length)
 	{
 		return sym_op({ text, length });
 	}
@@ -251,17 +309,33 @@ namespace literals
 // --------------------------------------------------------------------
 // The transformation class
 
+/**
+ * @brief A class you can use to apply symmetry transformations on points
+ *
+ * Transformations consist of two operations, a matrix transformation which
+ * is often a rotation followed by a translation.
+ *
+ * In case the matrix transformation is a pure rotation a quaternion
+ * is created to do the actual calculations. That's faster and more
+ * precise.
+ */
 class transformation
 {
   public:
+	/// \brief constructor taking a symop_data object @a data
 	transformation(const symop_data &data);
+
+	/// \brief constructor taking a rotation matrix @a r and a translation vector @a t
 	transformation(const matrix3x3<float> &r, const cif::point &t);
 
+	/** @cond */
 	transformation(const transformation &) = default;
 	transformation(transformation &&) = default;
 	transformation &operator=(const transformation &) = default;
 	transformation &operator=(transformation &&) = default;
+	/** @endcond */
 
+	/// \brief operator() to perform the transformation on point @a pt and return the result
 	point operator()(point pt) const
 	{
 		if (m_q)
@@ -272,9 +346,13 @@ class transformation
 		return pt + m_translation;
 	}
 
+	/// \brief return a transformation object that is the result of applying @a rhs after @a lhs
 	friend transformation operator*(const transformation &lhs, const transformation &rhs);
+
+	/// \brief return the inverse transformation for @a t
 	friend transformation inverse(const transformation &t);
 
+	/// \brief return the inverse tranformation for this
 	transformation operator-() const
 	{
 		return inverse(*this);
@@ -283,7 +361,6 @@ class transformation
 	friend class spacegroup;
 
   private:
-
 	// Most rotation matrices provided by the International Tables
 	// are really rotation matrices, in those cases we can construct
 	// a quaternion. Unfortunately, that doesn't work for all of them
@@ -298,22 +375,32 @@ class transformation
 // --------------------------------------------------------------------
 // class cell
 
+/**
+ * @brief The cell class describes the dimensions and angles of a unit cell
+ * in a crystal
+ */
+
 class cell
 {
   public:
+	/// \brief constructor
 	cell(float a, float b, float c, float alpha = 90.f, float beta = 90.f, float gamma = 90.f);
+
+	/// \brief constructor that takes the appropriate values from the *cell* category in datablock @a db
 	cell(const datablock &db);
 
-	float get_a() const { return m_a; }
-	float get_b() const { return m_b; }
-	float get_c() const { return m_c; }
+	float get_a() const { return m_a; } ///< return dimension a
+	float get_b() const { return m_b; } ///< return dimension b
+	float get_c() const { return m_c; } ///< return dimension c
 
-	float get_alpha() const { return m_alpha; }
-	float get_beta() const { return m_beta; }
-	float get_gamma() const { return m_gamma; }
+	float get_alpha() const { return m_alpha; } ///< return angle alpha
+	float get_beta() const { return m_beta; }   ///< return angle beta
+	float get_gamma() const { return m_gamma; } ///< return angle gamma
 
-	matrix3x3<float> get_orthogonal_matrix() const { return m_orthogonal; }
-	matrix3x3<float> get_fractional_matrix() const { return m_fractional; }
+	float get_volume() const; ///< return the calculated volume for this cell
+
+	matrix3x3<float> get_orthogonal_matrix() const { return m_orthogonal; } ///< return the matrix to use to transform coordinates from fractional to orthogonal
+	matrix3x3<float> get_fractional_matrix() const { return m_fractional; } ///< return the matrix to use to transform coordinates from orthogonal to fractional
 
   private:
 	void init();
@@ -324,80 +411,112 @@ class cell
 
 // --------------------------------------------------------------------
 
+/// \brief Return the spacegroup number from the *symmetry* category in datablock @a db
 int get_space_group_number(const datablock &db);
+
+/// \brief Return the spacegroup number for spacegroup named @a spacegroup
 int get_space_group_number(std::string_view spacegroup);
+
+/// \brief Return the spacegroup number for spacegroup named @a spacegroup assuming space_group_name @a type
 int get_space_group_number(std::string_view spacegroup, space_group_name type);
 
+/**
+ * @brief class to encapsulate the list of transformations making up a spacegroup
+ *
+ */
 class spacegroup : public std::vector<transformation>
 {
   public:
+	/// \brief constructor using the information in the *symmetry* category in datablock @a db
 	spacegroup(const datablock &db)
 		: spacegroup(get_space_group_number(db))
 	{
-
 	}
 
+	/// \brief constructor using the spacegroup named @a name
 	spacegroup(std::string_view name)
 		: spacegroup(get_space_group_number(name))
 	{
 	}
 
+	/// \brief constructor using the spacegroup named @a name assuming space_group_name @a type
 	spacegroup(std::string_view name, space_group_name type)
 		: spacegroup(get_space_group_number(name, type))
 	{
 	}
 
+	/// \brief constructor using the spacegroup number @a nr
 	spacegroup(int nr);
 
-	int get_nr() const { return m_nr; }
-	std::string get_name() const;
+	int get_nr() const { return m_nr; } ///< Return the nr
+	std::string get_name() const;       ///< Return the name
+
+	/** \brief perform a spacegroup operation on point @a pt using
+	 * cell @a c and sym_op @a symop
+	 */
 
 	point operator()(const point &pt, const cell &c, sym_op symop) const;
 
+	/** \brief perform an inverse spacegroup operation on point @a pt using
+	 * cell @a c and sym_op @a symop
+	 */
 	point inverse(const point &pt, const cell &c, sym_op symop) const;
 
   private:
 	int m_nr;
-	size_t m_index;
+	std::size_t m_index;
 };
 
 // --------------------------------------------------------------------
-// A crystal combines a cell and a spacegroup.
+/**
+ * @brief A crystal combines a cell and a spacegroup.
+ *
+ * The information in cell and spacegroup together make up all
+ * information you need to do symmetry calculations in a crystal
+ */
 
 class crystal
 {
   public:
+	/// \brief constructor using the information found in datablock @a db
 	crystal(const datablock &db)
 		: m_cell(db)
 		, m_spacegroup(db)
 	{
 	}
 
+	/// \brief constructor using cell @a c and spacegroup @a sg
 	crystal(const cell &c, const spacegroup &sg)
 		: m_cell(c)
 		, m_spacegroup(sg)
 	{
 	}
 
+	/** @cond */
 	crystal(const crystal &) = default;
 	crystal(crystal &&) = default;
 	crystal &operator=(const crystal &) = default;
 	crystal &operator=(crystal &&) = default;
+	/** @endcond */
 
-	const cell &get_cell() const { return m_cell; }
-	const spacegroup &get_spacegroup() const { return m_spacegroup; }
+	const cell &get_cell() const { return m_cell; }                   ///< Return the cell
+	const spacegroup &get_spacegroup() const { return m_spacegroup; } ///< Return the spacegroup
 
+	/// \brief Return the symmetry copy of point @a pt using symmetry operation @a symop
 	point symmetry_copy(const point &pt, sym_op symop) const
 	{
 		return m_spacegroup(pt, m_cell, symop);
 	}
 
+	/// \brief Return the symmetry copy of point @a pt using the inverse of symmetry operation @a symop
 	point inverse_symmetry_copy(const point &pt, sym_op symop) const
 	{
 		return m_spacegroup.inverse(pt, m_cell, symop);
 	}
-	
-	std::tuple<float,point,sym_op> closest_symmetry_copy(point a, point b) const;
+
+	/// \brief Return a tuple consisting of distance, new location and symmetry operation
+	/// for the point @a b with respect to point @a a.
+	std::tuple<float, point, sym_op> closest_symmetry_copy(point a, point b) const;
 
   private:
 	cell m_cell;
@@ -407,11 +526,13 @@ class crystal
 // --------------------------------------------------------------------
 // Symmetry operations on points
 
+/// \brief convenience function returning the fractional point @a pt in orthogonal coordinates for cell @a c
 inline point orthogonal(const point &pt, const cell &c)
 {
 	return c.get_orthogonal_matrix() * pt;
 }
 
+/// \brief convenience function returning the orthogonal point @a pt in fractional coordinates for cell @a c
 inline point fractional(const point &pt, const cell &c)
 {
 	return c.get_fractional_matrix() * pt;

include/cif++/text.hpp

@@ -38,12 +38,47 @@
 #include <vector>
 
 #if __has_include(<experimental/type_traits>)
+
 #include <experimental/type_traits>
+namespace std_experimental = std::experimental;
+
 #else
-// sub optimal, but replicating the same code is worse
-#include <zeep/type-traits.hpp>
+
+// A quick hack to work around the missing is_detected in MSVC
+namespace std_experimental
+{
+
+namespace detail
+{
+	template <class AlwaysVoid, template <class...> class Op, class... Args>
+	struct detector
+	{
+		using value_t = std::false_type;
+	};
+
+	template <template <class...> class Op, class... Args>
+	struct detector<std::void_t<Op<Args...>>, Op, Args...>
+	{
+		using value_t = std::true_type;
+	};
+} // namespace detail
+
+template <template <class...> class Op, class... Args>
+using is_detected = typename detail::detector<void, Op, Args...>::value_t;
+
+template <template <class...> class Op, class... Args>
+const auto is_detected_v = is_detected<Op, Args...>::value;
+
+} // namespace std_experimental
+
 #endif
 
+/**
+ * \file text.hpp
+ *
+ * Various text manipulating routines
+ */
+
 namespace cif
 {
 
@@ -52,18 +87,40 @@ namespace cif
 // some basic utilities: Since we're using ASCII input only, we define for optimisation
 // our own case conversion routines.
 
+/// \brief return whether string @a is equal to string @a b ignoring changes in character case
 bool iequals(std::string_view a, std::string_view b);
+
+/// \brief compare string @a is to string @a b ignoring changes in character case
 int icompare(std::string_view a, std::string_view b);
 
+/// \brief return whether string @a is equal to string @a b ignoring changes in character case
 bool iequals(const char *a, const char *b);
+
+/// \brief compare string @a is to string @a b ignoring changes in character case
 int icompare(const char *a, const char *b);
 
+/// \brief convert the string @a s to lower case in situ
 void to_lower(std::string &s);
+
+/// \brief return a lower case copy of string @a s
 std::string to_lower_copy(std::string_view s);
 
+/// \brief convert the string @a s to upper case in situ
 void to_upper(std::string &s);
-// std::string toUpperCopy(const std::string &s);
 
+/**
+ * @brief Join the strings in the range [ @a a, @a e ) using
+ * @a sep as separator
+ *
+ * Example usage:
+ *
+ * @code {.cpp}
+ * std::vector<std::string> v{ "aap", "noot", "mies" };
+ *
+ * assert(cif::join(v.begin(), v.end(), ", ") == "aap, noot, mies");
+ * @endcode
+ *
+ */
 template <typename IterType>
 std::string join(IterType b, IterType e, std::string_view sep)
 {
@@ -91,12 +148,41 @@ std::string join(IterType b, IterType e, std::string_view sep)
 	return s.str();
 }
 
+/**
+ * @brief Join the strings in the array @a arr using @a sep as separator
+ *
+ * Example usage:
+ *
+ * @code {.cpp}
+ * std::list<std::string> v{ "aap", "noot", "mies" };
+ *
+ * assert(cif::join(v, ", ") == "aap, noot, mies");
+ * @endcode
+ *
+ */
 template <typename V>
 std::string join(const V &arr, std::string_view sep)
 {
 	return join(arr.begin(), arr.end(), sep);
 }
 
+/**
+ * @brief Split the string in @a s based on the characters in @a separators
+ *
+ * Each of the characters in @a separators induces a split.
+ *
+ * When suppress_empty is true, empty strings are not produced in the
+ * resulting array.
+ *
+ * Example:
+ *
+ * @code {.cpp}
+ * auto v = cif::split("aap:noot,,mies", ":,", true);
+ *
+ * assert(v == std::vector{"aap", "noot", "mies"});
+ * @endcode
+ *
+ */
 template <typename StringType = std::string_view>
 std::vector<StringType> split(std::string_view s, std::string_view separators, bool suppress_empty = false)
 {
@@ -124,15 +210,23 @@ std::vector<StringType> split(std::string_view s, std::string_view separators, b
 	return result;
 }
 
+/**
+ * @brief Replace all occurrences of @a what in string @a s with the string @a with
+ *
+ * The string @a with may be empty in which case each occurrence of @a what is simply
+ * deleted.
+ */
 void replace_all(std::string &s, std::string_view what, std::string_view with = {});
 
 #if defined(__cpp_lib_starts_ends_with)
 
+/// \brief return whether string @a s starts with @a with
 inline bool starts_with(std::string s, std::string_view with)
 {
 	return s.starts_with(with);
 }
 
+/// \brief return whether string @a s ends with @a with
 inline bool ends_with(std::string_view s, std::string_view with)
 {
 	return s.ends_with(with);
@@ -140,11 +234,13 @@ inline bool ends_with(std::string_view s, std::string_view with)
 
 #else
 
+/// \brief return whether string @a s starts with @a with
 inline bool starts_with(std::string s, std::string_view with)
 {
 	return s.compare(0, with.length(), with) == 0;
 }
 
+/// \brief return whether string @a s ends with @a with
 inline bool ends_with(std::string_view s, std::string_view with)
 {
 	return s.length() >= with.length() and s.compare(s.length() - with.length(), with.length(), with) == 0;
@@ -154,6 +250,7 @@ inline bool ends_with(std::string_view s, std::string_view with)
 
 #if defined(__cpp_lib_string_contains)
 
+/// \brief return whether string @a s contains @a q
 inline bool contains(std::string_view s, std::string_view q)
 {
 	return s.contains(q);
@@ -161,6 +258,7 @@ inline bool contains(std::string_view s, std::string_view q)
 
 #else
 
+/// \brief return whether string @a s contains @a q
 inline bool contains(std::string_view s, std::string_view q)
 {
 	return s.find(q) != std::string_view::npos;
@@ -168,33 +266,50 @@ inline bool contains(std::string_view s, std::string_view q)
 
 #endif
 
+/// \brief return whether string @a s contains @a q ignoring character case
 bool icontains(std::string_view s, std::string_view q);
 
+/// \brief trim white space at the start of string @a s in situ
 void trim_left(std::string &s);
+
+/// \brief trim white space at the end of string @a s in situ
 void trim_right(std::string &s);
+
+/// \brief trim white space at both the start and the end of string @a s in situ
 void trim(std::string &s);
 
+/// \brief return a string trimmed of white space at the start of string @a s
 std::string trim_left_copy(std::string_view s);
+
+/// \brief return a string trimmed of white space at the end of string @a s
 std::string trim_right_copy(std::string_view s);
+
+/// \brief return a string trimmed of white space at both the start and the end of string @a s
 std::string trim_copy(std::string_view s);
 
 // To make life easier, we also define iless and iset using iequals
 
+/// \brief an operator object you can use to compare strings ignoring their character case
 struct iless
 {
+	/// \brief return the result of icompare for @a a and @a b
 	bool operator()(const std::string &a, const std::string &b) const
 	{
 		return icompare(a, b) < 0;
 	}
 };
 
-typedef std::set<std::string, iless> iset;
+/// iset is a std::set of std::string but with a comparator that
+/// ignores character case.
+using iset = std::set<std::string, iless>;
 
 // --------------------------------------------------------------------
 // This really makes a difference, having our own tolower routines
 
+/// \brief global list containing the lower case version of each ASCII character
 extern CIFPP_EXPORT const uint8_t kCharToLowerMap[256];
 
+/// \brief a very fast tolower implementation
 inline char tolower(int ch)
 {
 	return static_cast<char>(kCharToLowerMap[static_cast<uint8_t>(ch)]);
@@ -202,22 +317,49 @@ inline char tolower(int ch)
 
 // --------------------------------------------------------------------
 
-std::tuple<std::string, std::string> split_tag_name(std::string_view tag);
+/** \brief return a tuple consisting of the category and item name for @a item_name
+ *
+ * The category name is stripped of its leading underscore character.
+ *
+ * If no dot character was found, the category name is empty. That's for
+ * cif 1.0 formatted data.
+ */
+
+[[deprecated("use split_item_name instead")]]
+std::tuple<std::string, std::string> split_tag_name(std::string_view item_name);
+
+
+/** \brief return a tuple consisting of the category and item name for @a item_name
+ *
+ * The category name is stripped of its leading underscore character.
+ *
+ * If no dot character was found, the category name is empty. That's for
+ * cif 1.0 formatted data.
+ */
+
+std::tuple<std::string, std::string> split_item_name(std::string_view item_name);
 
 // --------------------------------------------------------------------
-// generate a cif name, mainly used to generate asym_id's
 
+/// \brief generate a cif name, used e.g. to generate asym_id's
 std::string cif_id_for_number(int number);
 
 // --------------------------------------------------------------------
-//	custom wordwrapping routine
 
-std::vector<std::string> word_wrap(const std::string &text, size_t width);
+/** \brief custom word wrapping routine.
+ *
+ * Wrap the text in @a text based on a maximum line width @a width using
+ * a dynamic programming approach to get the most efficient filling of
+ * the space.
+ */
+std::vector<std::string> word_wrap(const std::string &text, std::size_t width);
 
 // --------------------------------------------------------------------
-/// std::from_chars for floating point types.
+/// \brief std::from_chars for floating point types.
+///
 /// These are optional, there's a selected_charconv class below that selects
-/// the best option to used based on support by the stl library
+/// the best option to use based on support by the stl library.
+///
 /// I.e. that in case of GNU < 12 (or something) the cif implementation will
 /// be used, all other cases will use the stl version.
 
@@ -236,12 +378,12 @@ std::from_chars_result from_chars(const char *first, const char *last, FloatType
 	} state = IntegerSign;
 	int sign = 1;
 	unsigned long long vi = 0;
-	long double f = 1;
+	int fl = 0, tz = 0;
 	int exponent_sign = 1;
 	int exponent = 0;
 	bool done = false;
 
-	while (not done and result.ec == std::errc())
+	while (not done and not (bool)result.ec)
 	{
 		char ch = result.ptr != last ? *result.ptr : 0;
 		++result.ptr;
@@ -285,7 +427,14 @@ std::from_chars_result from_chars(const char *first, const char *last, FloatType
 				if (ch >= '0' and ch <= '9')
 				{
 					vi = 10 * vi + (ch - '0');
-					f /= 10;
+
+					if (ch == '0')
+						tz += 1;
+					else
+					{
+						fl += tz + 1;
+						tz = 0;
+					}
 				}
 				else if (ch == 'e' or ch == 'E')
 					state = ExponentSign;
@@ -325,9 +474,12 @@ std::from_chars_result from_chars(const char *first, const char *last, FloatType
 		}
 	}
 
-	if (result.ec == std::errc())
+	if (not (bool)result.ec)
 	{
-		long double v = f * vi * sign;
+		while (tz-- > 0)
+			vi /= 10;
+
+		long double v = std::pow(10, -fl) * vi * sign;
 		if (exponent != 0)
 			v *= std::pow(10, exponent * exponent_sign);
 
@@ -342,6 +494,7 @@ std::from_chars_result from_chars(const char *first, const char *last, FloatType
 	return result;
 }
 
+/// \brief duplication of std::chars_format for deficient STL implementations
 enum class chars_format
 {
 	scientific = 1,
@@ -350,6 +503,7 @@ enum class chars_format
 	general = fixed | scientific
 };
 
+/// \brief a simplistic implementation of std::to_chars for old STL implementations
 template <typename FloatType, std::enable_if_t<std::is_floating_point_v<FloatType>, int> = 0>
 std::to_chars_result to_chars(char *first, char *last, FloatType &value, chars_format fmt)
 {
@@ -389,6 +543,7 @@ std::to_chars_result to_chars(char *first, char *last, FloatType &value, chars_f
 	return result;
 }
 
+/// \brief a simplistic implementation of std::to_chars for old STL implementations
 template <typename FloatType, std::enable_if_t<std::is_floating_point_v<FloatType>, int> = 0>
 std::to_chars_result to_chars(char *first, char *last, FloatType &value, chars_format fmt, int precision)
 {
@@ -428,38 +583,51 @@ std::to_chars_result to_chars(char *first, char *last, FloatType &value, chars_f
 	return result;
 }
 
+/// \brief class that uses our implementation of std::from_chars and std::to_chars
 template <typename T>
 struct my_charconv
 {
+	/// @brief Simply call our version of std::from_chars
 	static std::from_chars_result from_chars(const char *a, const char *b, T &d)
 	{
 		return cif::from_chars(a, b, d);
 	}
 
+	/// @brief Simply call our version of std::to_chars
 	static std::to_chars_result to_chars(char *first, char *last, T &value, chars_format fmt)
 	{
 		return cif::to_chars(first, last, value, fmt);
 	}
 };
 
+/// \brief class that uses the STL implementation of std::from_chars and std::to_chars
 template <typename T>
 struct std_charconv
 {
+	/// @brief Simply call std::from_chars
 	static std::from_chars_result from_chars(const char *a, const char *b, T &d)
 	{
 		return std::from_chars(a, b, d);
 	}
 
+	/// @brief Simply call std::to_chars
 	static std::to_chars_result to_chars(char *first, char *last, T &value, chars_format fmt)
 	{
 		return std::to_chars(first, last, value, fmt);
 	}
 };
 
+/// \brief helper to find a from_chars function
 template <typename T>
 using from_chars_function = decltype(std::from_chars(std::declval<const char *>(), std::declval<const char *>(), std::declval<T &>()));
 
+/**
+ * @brief Helper to select the best implementation of charconv based on availability of the
+ * function in the std:: namespace
+ *
+ * @tparam T The type for which we want to find a from_chars/to_chars function
+ */
 template <typename T>
-using selected_charconv = typename std::conditional_t<std::experimental::is_detected_v<from_chars_function, T>, std_charconv<T>, my_charconv<T>>;
+using selected_charconv = typename std::conditional_t<std_experimental::is_detected_v<from_chars_function, T>, std_charconv<T>, my_charconv<T>>;
 
 } // namespace cif

include/cif++/utilities.hpp

@@ -1,17 +1,17 @@
 /*-
  * SPDX-License-Identifier: BSD-2-Clause
- * 
+ *
  * Copyright (c) 2020 NKI/AVL, Netherlands Cancer Institute
- * 
+ *
  * Redistribution and use in source and binary forms, with or without
  * modification, are permitted provided that the following conditions are met:
- * 
+ *
  * 1. Redistributions of source code must retain the above copyright notice, this
  *    list of conditions and the following disclaimer
  * 2. Redistributions in binary form must reproduce the above copyright notice,
  *    this list of conditions and the following disclaimer in the documentation
  *    and/or other materials provided with the distribution.
- * 
+ *
  * THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND CONTRIBUTORS "AS IS" AND
  * ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT NOT LIMITED TO, THE IMPLIED
  * WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR PURPOSE ARE
@@ -29,11 +29,18 @@
 #include "cif++/exports.hpp"
 
 #include <filesystem>
+#include <iostream>
 
 #ifndef STDOUT_FILENO
+/// @brief For systems that lack this value
 #define STDOUT_FILENO 1
 #endif
 
+#ifndef STDERR_FILENO
+/// @brief For systems that lack this value
+#define STDERR_FILENO 2
+#endif
+
 #if _WIN32
 #include <io.h>
 #define isatty _isatty
@@ -49,123 +56,243 @@
 #define _SILENCE_CXX17_CODECVT_HEADER_DEPRECATION_WARNING 1
 #endif
 
+/** \file utilities.hpp
+ *
+ * This file contains code that is very generic in nature like a progress_bar
+ * and classes you can use to colourise output text.
+ */
+
 namespace cif
 {
 
+/**
+ * @brief The global variable VERBOSE contains the level of verbosity
+ * requested. A value of 0 is normal, with some output on error conditions.
+ * A value > 0 will result in more output, the higher the value, the more
+ * output. A value < 0 will make the library silent, even in error
+ * conditions.
+ */
 extern CIFPP_EXPORT int VERBOSE;
 
-// the git 'build' number
+/// return the git 'build' number
 std::string get_version_nr();
-// std::string get_version_date();
-
-// --------------------------------------------------------------------
-//	Code helping with terminal i/o
 
+/// return the width of the current output terminal, or 80 if it cannot be determined
 uint32_t get_terminal_width();
 
 // --------------------------------------------------------------------
-//	Path of the current executable
 
-std::string get_executable_path();
-
-// --------------------------------------------------------------------
-//	some manipulators to write coloured text to terminals
-
-enum StringColour
+namespace colour
 {
-	scBLACK = 0,
-	scRED,
-	scGREEN,
-	scYELLOW,
-	scBLUE,
-	scMAGENTA,
-	scCYAN,
-	scWHITE,
-	scNONE = 9
-};
-
-template <typename String, typename CharT>
-struct ColouredString
-{
-	static_assert(std::is_reference<String>::value or std::is_pointer<String>::value, "String type must be pointer or reference");
-
-	ColouredString(String s, StringColour fore, StringColour back, bool bold = true)
-		: m_s(s)
-		, m_fore(fore)
-		, m_back(back)
-		, m_bold(bold)
+	/// @brief The defined colours
+	enum colour_type
 	{
-	}
+		black = 0,
+		red,
+		green,
+		yellow,
+		blue,
+		magenta,
+		cyan,
+		white,
+		none = 9
+	};
 
-	ColouredString &operator=(const ColouredString &) = delete;
-
-	String m_s;
-	StringColour m_fore, m_back;
-	bool m_bold;
-};
-
-template <typename CharT, typename Traits>
-std::basic_ostream<CharT, Traits> &operator<<(std::basic_ostream<CharT, Traits> &os, const ColouredString<const CharT *, CharT> &s)
-{
-	if (isatty(STDOUT_FILENO))
+	/// @brief The defined styles
+	enum style_type
 	{
-		std::basic_ostringstream<CharT, Traits> ostr;
-		ostr << "\033[" << (30 + s.m_fore) << ';' << (s.m_bold ? "1" : "22") << ';' << (40 + s.m_back) << 'm'
-			 << s.m_s
-			 << "\033[0m";
+		bold = 1,
+		underlined = 4,
+		blink = 5,
+		inverse = 7,
+		regular = 22,
+	};
 
-		return os << ostr.str();
-	}
-	else
-		return os << s.m_s;
-}
-
-template <typename CharT, typename Traits, typename String>
-std::basic_ostream<CharT, Traits> &operator<<(std::basic_ostream<CharT, Traits> &os, const ColouredString<String, CharT> &s)
-{
-	if (isatty(STDOUT_FILENO))
+	namespace detail
 	{
-		std::basic_ostringstream<CharT, Traits> ostr;
-		ostr << "\033[" << (30 + s.m_fore) << ';' << (s.m_bold ? "1" : "22") << ';' << (40 + s.m_back) << 'm'
-			 << s.m_s
-			 << "\033[0m";
+		/**
+		 * @brief Struct for delimited strings.
+		 */
+		struct coloured_string_t
+		{
+			/**
+			 * @brief Construct a new coloured string t object
+			 */
+			coloured_string_t(std::string_view s, colour_type fc, colour_type bc, style_type st)
+				: m_str(s)
+				, m_fore_colour(static_cast<int>(fc) + 30)
+				, m_back_colour(static_cast<int>(bc) + 40)
+				, m_style(static_cast<int>(st))
+			{
+			}
 
-		return os << ostr.str();
-	}
-	else
-		return os << s.m_s;
-}
+			coloured_string_t &operator=(coloured_string_t &) = delete;
 
-template <typename CharT>
-inline auto coloured(const CharT *s, StringColour fore = scWHITE, StringColour back = scRED, bool bold = true)
+			/**
+			 * @brief Write out the string, either coloured or not
+			 */
+			template <typename char_type, typename traits_type>
+			friend std::basic_ostream<char_type, traits_type> &operator<<(
+				std::basic_ostream<char_type, traits_type> &os, const coloured_string_t &cs)
+			{
+				bool use_colour = false;
+
+				if (os.rdbuf() == std::cout.rdbuf() and isatty(STDOUT_FILENO))
+					use_colour = true;
+				else if (os.rdbuf() == std::cerr.rdbuf() and isatty(STDERR_FILENO))
+					use_colour = true;
+
+				if (use_colour)
+				{
+					os << "\033[" << cs.m_fore_colour << ';' << cs.m_style << ';' << cs.m_back_colour << 'm'
+					   << cs.m_str
+					   << "\033[0m";
+				}
+				else
+					os << cs.m_str;
+
+				return os;
+			}
+
+			/// @cond
+			std::string_view m_str;
+			int m_fore_colour, m_back_colour;
+			int m_style;
+			/// @endcond
+		};
+
+	} // namespace detail
+} // namespace colour
+
+/**
+ * @brief Manipulator for coloured strings.
+ * 
+ * When writing out text to the terminal it is often useful to have
+ * some of the text colourised. But only if the output is really a
+ * terminal since colouring text is done using escape sequences
+ * an if output is redirected to a file, these escape sequences end up
+ * in the file making the real text less easy to read.
+ *
+ * The code presented here is rather basic. It mimics the std::quoted
+ * manipulator in that it will colour a string with optionally
+ * requested colours and text style.
+ *
+ * Example:
+ *
+ * @code {.cpp}
+ * using namespace cif::colour;
+ * std::cout << cif::coloured("Hello, world!", white, red, bold) << '\n';
+ * @endcode
+ * @param str String to quote.
+ * @param fg Foreground (=text) colour to use
+ * @param bg Background colour to use
+ * @param st Text style to use
+ */
+
+template <typename T>
+	requires std::is_assignable_v<std::string_view, T>
+inline auto coloured(T str,
+	colour::colour_type fg, colour::colour_type bg = colour::colour_type::none,
+	colour::style_type st = colour::style_type::regular)
 {
-	return ColouredString<const CharT *, CharT>(s, fore, back, bold);
-}
-
-template <typename CharT, typename Traits, typename Alloc>
-inline auto coloured(const std::basic_string<CharT, Traits, Alloc> &s, StringColour fore = scWHITE, StringColour back = scRED, bool bold = true)
-{
-	return ColouredString<const std::basic_string<CharT, Traits, Alloc>, CharT>(s, fore, back, bold);
-}
-
-template <typename CharT, typename Traits, typename Alloc>
-inline auto coloured(std::basic_string<CharT, Traits, Alloc> &s, StringColour fore = scWHITE, StringColour back = scRED, bool bold = true)
-{
-	return ColouredString<std::basic_string<CharT, Traits, Alloc>, CharT>(s, fore, back, bold);
+	return colour::detail::coloured_string_t(str, fg, bg, st);
 }
 
 // --------------------------------------------------------------------
 //	A progress bar
 
+/**
+ * @brief A simple progress bar class for terminal based output
+ * 
+ * Using a progress bar is very convenient for the end user when
+ * you have long running code. It gives feed back on how fast an
+ * operation is performed and may give an indication how long it
+ * will take before it is finished.
+ * 
+ * Using this cif::progress_bar implementation is straightforward:
+ * 
+ * @code {.cpp}
+ * using namespace std::chrono_literals;
+ * 
+ * cif::progress_bar pb(10, "counting to ten");
+ * 
+ * for (int i = 1; i <= 10; ++i)
+ * {
+ *   pb.consumed(1);
+ *   std::this_thread::sleep_for(1s);
+ * }
+ * 
+ * @endcode
+ * 
+ * When the progress_bar is created, it first checks
+ * to see if stdout is to a real TTY and if the VERBOSE
+ * flag is not less than zero (quiet mode). If this passes
+ * a thread is started that waits for updates.
+ * 
+ * The first two seconds, nothing is written to the screen
+ * so if the work is finished within those two seconds
+ * the screen stays clean.
+ * 
+ * After this time, a progress bar is printed that may look
+ * like this:
+ * 
+ * @code
+ * step 3           ========================--------------------------------  40% ⢁
+ * @endcode
+ * 
+ * The first characters contain the initial action name or
+ * the message text if it was used afterwards.
+ * 
+ * The thermometer is made up with '=' and '-' characters.
+ * 
+ * A percentage is also shown and at the end there is a spinner
+ * that gives feedback that the program is really still working.
+ * 
+ * The progress bar is removed if the max has been reached
+ * or if the progress bar is destructed. If any output has
+ * been generated, the initial action is printed out along
+ * with the total time spent.
+ */
+
 class progress_bar
 {
   public:
+	/**
+	 * @brief Construct a new progress bar object
+	 * 
+	 * Progress ranges from 0 (zero) to @a inMax
+	 * 
+	 * The action in @a inAction is used for display
+	 * 
+	 * @param inMax The maximum value
+	 * @param inAction The description of what is
+	 * going on
+	 */
+
 	progress_bar(int64_t inMax, const std::string &inAction);
+
+	/**
+	 * @brief Destroy the progress bar object
+	 * 
+	 */
 	~progress_bar();
 
+	/**
+	 * @brief Notify the progress bar that @a inConsumed
+	 * should be added to the internal progress counter
+	 */
 	void consumed(int64_t inConsumed); // consumed is relative
+
+	/**
+	 * @brief Notify the progress bar that the internal
+	 * progress counter should be updated to @a inProgress
+	 */
 	void progress(int64_t inProgress); // progress is absolute
 
+	/**
+	 * @brief Replace the action string in the progress bar
+	 * with @a inMessage
+	 */
 	void message(const std::string &inMessage);
 
   private:
@@ -178,8 +305,66 @@ class progress_bar
 // --------------------------------------------------------------------
 // Resources
 
+/**
+ * @brief Load a resource from disk or the compiled in resources
+ * 
+ * @verbatim embed:rst
+.. note::
+
+   See the :doc:`documentation on resources </resources>` for more information.
+
+   @endverbatim
+ * 
+ * @param name The named resource to load
+ * @return std::unique_ptr<std::istream> A pointer to the std::istream or empty if not found
+ */
+
 std::unique_ptr<std::istream> load_resource(std::filesystem::path name);
+
+/**
+ * @brief Add a file specified by @a dataFile as the data for resource @a name
+ * 
+ * @verbatim embed:rst
+.. note::
+
+   See the :doc:`documentation on resources </resources>` for more information.
+
+   @endverbatim
+ * 
+ * @param name The name of the resource to specify
+ * @param dataFile Path to a file containing the data
+ */
+
 void add_file_resource(const std::string &name, std::filesystem::path dataFile);
+
+/**
+ * @brief List all the file resources added with cif::add_file_resource.
+ * 
+ * @param os The std::ostream to write the directories to
+ */
+
+void list_file_resources(std::ostream &os);
+
+/**
+ * @brief Add a directory to the list of search directories. This list is
+ * searched in a last-in-first-out order.
+ * 
+ * @verbatim embed:rst
+.. note::
+
+   See the :doc:`documentation on resources </resources>` for more information.
+
+   @endverbatim
+ */
+
 void add_data_directory(std::filesystem::path dataDir);
 
+/**
+ * @brief List all the data directories, for error reporting on missing resources.
+ * 
+ * @param os The std::ostream to write the directories to
+ */
+
+void list_data_directories(std::ostream &os);
+
 } // namespace cif

include/cif++/validate.hpp

@@ -28,51 +28,215 @@
 
 #include "cif++/text.hpp"
 
+#include <cassert>
 #include <filesystem>
 #include <list>
 #include <mutex>
+#include <system_error>
 #include <utility>
 
+/**
+ * @file validate.hpp
+ *
+ * Support for validating mmCIF files based on a dictionary. These dictionaries
+ * contain information about the categories and items therein, what they may
+ * contain and how this should be formatted. There's also information on links
+ * between parent and child categories.
+ *
+ */
+
 namespace cif
 {
 
 struct category_validator;
 
 // --------------------------------------------------------------------
+// New: error_code
 
-class validation_error : public std::exception
+/**
+ * @enum validation_error
+ *
+ * @brief A stronly typed class containing the error codes reported by @ref cif::validator and friends
+ */
+enum class validation_error
+{
+	value_does_not_match_rx = 1,      /**< The value of an item does not conform to the regular expression specified for it */
+	value_is_not_in_enumeration_list, /**< The value of an item is not in the list of values allowed */
+	not_a_known_primitive_type,       /**< The type is not a known primitive type */
+	undefined_category,               /**< Category has no definition in the dictionary */
+	unknown_item,                     /**< The item is not defined to be part of the category */
+	incorrect_item_validator,         /**< Incorrectly specified validator for item */
+	missing_mandatory_items,          /**< Missing mandatory items */
+	missing_key_items,                /**< An index could not be constructed due to missing key items */
+	item_not_allowed_in_category,     /**< Requested item allowed in category according to dictionary */
+	empty_file,                       /**< The file contains no datablocks */
+	empty_datablock,                  /**< The datablock contains no categories */
+	empty_category,                   /**< The category is empty */
+	not_valid_pdbx,                   /**< The file is not a valid PDBx file */
+};
+/**
+ * @brief The implementation for @ref validation_category error messages
+ *
+ */
+class validation_category_impl : public std::error_category
 {
   public:
-	validation_error(const std::string &msg);
-	validation_error(const std::string &cat, const std::string &item,
-		const std::string &msg);
-	const char *what() const noexcept { return m_msg.c_str(); }
-	std::string m_msg;
+	/**
+	 * @brief User friendly name
+	 *
+	 * @return const char*
+	 */
+
+	const char *name() const noexcept override
+	{
+		return "cif::validation";
+	}
+
+	/**
+	 * @brief Provide the error message as a string for the error code @a ev
+	 *
+	 * @param ev The error code
+	 * @return std::string
+	 */
+
+	std::string message(int ev) const override
+	{
+		switch (static_cast<validation_error>(ev))
+		{
+			case validation_error::value_does_not_match_rx:
+				return "Value in item does not match regular expression";
+			case validation_error::value_is_not_in_enumeration_list:
+				return "Value is not in the enumerated list of valid values";
+			case validation_error::not_a_known_primitive_type:
+				return "The type is not a known primitive type";
+			case validation_error::undefined_category:
+				return "Category has no definition in the dictionary";
+			case validation_error::unknown_item:
+				return "Item is not defined to be part of the category";
+			case validation_error::incorrect_item_validator:
+				return "Incorrectly specified validator for item";
+			case validation_error::missing_mandatory_items:
+				return "Missing mandatory items";
+			case validation_error::missing_key_items:
+				return "An index could not be constructed due to missing key items";
+			case validation_error::item_not_allowed_in_category:
+				return "Requested item not allowed in category according to dictionary";
+			case validation_error::empty_file:
+				return "The file contains no datablocks";
+			case validation_error::empty_datablock:
+				return "The datablock contains no categories";
+			case validation_error::empty_category:
+				return "The category is empty";
+			case validation_error::not_valid_pdbx:
+				return "The file is not a valid PDBx file";
+
+			default:
+				assert(false);
+				return "unknown error code";
+		}
+	}
+
+	/**
+	 * @brief Return whether two error codes are equivalent, always false in this case
+	 *
+	 */
+
+	bool equivalent(const std::error_code & /*code*/, int /*condition*/) const noexcept override
+	{
+		return false;
+	}
+};
+
+/**
+ * @brief Return the implementation for the validation_category
+ *
+ * @return std::error_category&
+ */
+inline std::error_category &validation_category()
+{
+	static validation_category_impl instance;
+	return instance;
+}
+
+inline std::error_code make_error_code(validation_error e)
+{
+	return std::error_code(static_cast<int>(e), validation_category());
+}
+
+inline std::error_condition make_error_condition(validation_error e)
+{
+	return std::error_condition(static_cast<int>(e), validation_category());
+}
+
+// --------------------------------------------------------------------
+
+class validation_exception : public std::runtime_error
+{
+  public:
+	validation_exception(validation_error err)
+		: validation_exception(make_error_code(err))
+	{
+	}
+
+	validation_exception(validation_error err, std::string_view category)
+		: validation_exception(make_error_code(err), category)
+	{
+	}
+
+	validation_exception(validation_error err, std::string_view category, std::string_view item)
+		: validation_exception(make_error_code(err), category, item)
+	{
+	}
+
+	validation_exception(std::error_code ec);
+
+	validation_exception(std::error_code ec, std::string_view category);
+
+	validation_exception(std::error_code ec, std::string_view category, std::string_view item);
 };
 
 // --------------------------------------------------------------------
 
+/** @brief the primitive types known */
 enum class DDL_PrimitiveType
 {
-	Char,
-	UChar,
-	Numb
+	Char,  ///< Text
+	UChar, ///< Text that is compared ignoring the character case
+	Numb   ///< Nummeric values
 };
 
+/// @brief Return the DDL_PrimitiveType encoded in @a s
 DDL_PrimitiveType map_to_primitive_type(std::string_view s);
 
+/// @brief Return the DDL_PrimitiveType encoded in @a s, error reporting variant
+DDL_PrimitiveType map_to_primitive_type(std::string_view s, std::error_code &ec) noexcept;
+
 struct regex_impl;
 
+/**
+ * @brief For each defined type in a dictionary a type_validator is created
+ *
+ * A type validator can check if the contents of an item are conforming the
+ * specification. The check is done using regular expressions.
+ *
+ * A type_validator can also be used to compare two values that conform to
+ * this type. Comparison is of course based on the primitive type.
+ *
+ */
 struct type_validator
 {
-	std::string m_name;
-	DDL_PrimitiveType m_primitive_type;
-	regex_impl *m_rx;
+	std::string m_name;                 ///< The name of the type
+	DDL_PrimitiveType m_primitive_type; ///< The primitive_type of the type
+	regex_impl *m_rx;                   ///< The regular expression for the type
 
 	type_validator() = delete;
+
+	/// @brief Constructor
 	type_validator(std::string_view name, DDL_PrimitiveType type, std::string_view rx);
 
 	type_validator(const type_validator &) = delete;
+
+	/// @brief Copy constructor
 	type_validator(type_validator &&rhs)
 		: m_name(std::move(rhs.m_name))
 		, m_primitive_type(rhs.m_primitive_type)
@@ -81,6 +245,8 @@ struct type_validator
 	}
 
 	type_validator &operator=(const type_validator &) = delete;
+
+	/// @brief Move constructor
 	type_validator &operator=(type_validator &&rhs)
 	{
 		m_name = std::move(rhs.m_name);
@@ -90,119 +256,211 @@ struct type_validator
 		return *this;
 	}
 
+	/// @brief Destructor
 	~type_validator();
 
+	/// @brief Return the sorting order
 	bool operator<(const type_validator &rhs) const
 	{
 		return icompare(m_name, rhs.m_name) < 0;
 	}
 
+	/// @brief Compare the contents of @a a and @a b based on the
+	/// primitive type of this type. A value of zero indicates the
+	/// values are equal. Less than zero means @a a sorts before @a b
+	/// and a value larger than zero likewise means the opposite
 	int compare(std::string_view a, std::string_view b) const;
 };
 
-struct item_validator
+/** @brief Item alias, items can be renamed over time
+ */
+
+struct item_alias
 {
-	std::string m_tag;
-	bool m_mandatory;
-	const type_validator *m_type;
-	cif::iset m_enums;
-	std::string m_default;
-	bool m_default_is_null;
-	category_validator *m_category = nullptr;
-
-	// ItemLinked is used for non-key links
-	struct item_link
+	item_alias(const std::string &alias_name, const std::string &dictionary, const std::string &version)
+		: m_name(alias_name)
+		, m_dict(dictionary)
+		, m_vers(version)
 	{
-		item_validator *m_parent;
-		std::string m_parent_item;
-		std::string m_child_item;
-	};
-
-	std::vector<item_link> mLinked;
-
-	bool operator<(const item_validator &rhs) const
-	{
-		return icompare(m_tag, rhs.m_tag) < 0;
 	}
 
-	bool operator==(const item_validator &rhs) const
-	{
-		return iequals(m_tag, rhs.m_tag);
-	}
+	item_alias(const item_alias &) = default;
+	item_alias &operator=(const item_alias &) = default;
 
-	void operator()(std::string_view value) const;
+	std::string m_name; ///< The alias_name
+	std::string m_dict; ///< The dictionary in which it was known
+	std::string m_vers; ///< The version of the dictionary
 };
 
+/**
+ * @brief An item_validator binds a type_validator to an item in
+ * a category along with other information found in the dictionary.
+ *
+ * mmCIF dictionaries may indicate an item is e.g. mandatory or
+ * consists of a certain list of allowed values. Even default
+ * values can be provided.
+ *
+ */
+struct item_validator
+{
+	std::string m_item_name;                  ///< The item name
+	bool m_mandatory;                         ///< Flag indicating this item is mandatory
+	const type_validator *m_type;             ///< The type for this item
+	cif::iset m_enums;                        ///< If filled, the set of allowed values
+	std::string m_default;                    ///< If filled, a default value for this item
+	category_validator *m_category = nullptr; ///< The category_validator this item_validator belongs to
+	std::vector<item_alias> m_aliases;        ///< The aliases for this item
+
+	/// @brief Compare based on the name
+	bool operator<(const item_validator &rhs) const
+	{
+		return icompare(m_item_name, rhs.m_item_name) < 0;
+	}
+
+	/// @brief Compare based on the name
+	bool operator==(const item_validator &rhs) const
+	{
+		return iequals(m_item_name, rhs.m_item_name);
+	}
+
+	/// @brief Validate the value in @a value for this item
+	/// Will throw a std::system_error exception if it fails
+	void operator()(std::string_view value) const;
+
+	/// @brief A more gentle version of value validation
+	bool validate_value(std::string_view value, std::error_code &ec) const noexcept;
+};
+
+/**
+ * @brief A validator for categories
+ *
+ * Categories can have a key, a set of items that in combination
+ * should be unique.
+ */
 struct category_validator
 {
-	std::string m_name;
-	std::vector<std::string> m_keys;
-	cif::iset m_groups;
-	cif::iset m_mandatory_fields;
-	std::set<item_validator> m_item_validators;
+	std::string m_name;                         ///< The name of the category
+	std::vector<std::string> m_keys;            ///< The list of items that make up the key
+	cif::iset m_groups;                         ///< The category groups this category belongs to
+	cif::iset m_mandatory_items;                ///< The mandatory items for this category
+	std::set<item_validator> m_item_validators; ///< The item validators for the items in this category
 
+	/// @brief return true if this category sorts before @a rhs
 	bool operator<(const category_validator &rhs) const
 	{
 		return icompare(m_name, rhs.m_name) < 0;
 	}
 
-	void addItemValidator(item_validator &&v);
+	/// @brief Add item_validator @a v to the list of item validators
+	void add_item_validator(item_validator &&v);
 
-	const item_validator *get_validator_for_item(std::string_view tag) const;
+	/// @brief Return the item_validator for item @a item_name, may return nullptr
+	const item_validator *get_validator_for_item(std::string_view item_name) const;
 
-	const std::set<item_validator> &item_validators() const
-	{
-		return m_item_validators;
-	}
+	/// @brief Return the item_validator for an item that has as alias name @a item_name, may return nullptr
+	const item_validator *get_validator_for_aliased_item(std::string_view item_name) const;
 };
 
+/**
+ * @brief A validator for links between categories
+ *
+ * Links are defined as a set of pairs of item names in a
+ * parent category and a corresponding item in a child
+ * category. This means that the size of m_parent_keys
+ * is always equal to the size of m_child_keys.
+ *
+ * Multiple links may be defined between two categories.
+ *
+ */
 struct link_validator
 {
-	int m_link_group_id;
-	std::string m_parent_category;
-	std::vector<std::string> m_parent_keys;
-	std::string m_child_category;
-	std::vector<std::string> m_child_keys;
-	std::string m_link_group_label;
+	int m_link_group_id;                    ///< The link group ID
+	std::string m_parent_category;          ///< The name of the parent category
+	std::vector<std::string> m_parent_keys; ///< The items in the parent category making up the set of linked items
+	std::string m_child_category;           ///< The name of the child category
+	std::vector<std::string> m_child_keys;  ///< The items in the child category making up the set of linked items
+	std::string m_link_group_label;         ///< The group label assigned to this link
 };
 
 // --------------------------------------------------------------------
 
+/**
+ * @brief The validator class combines all the link, category and item validator classes
+ *
+ */
 class validator
 {
   public:
+	/**
+	 * @brief Construct a new validator object
+	 *
+	 * @param name The name of the underlying dictionary
+	 */
 	validator(std::string_view name)
 		: m_name(name)
 	{
 	}
 
+	/// @brief destructor
 	~validator() = default;
 
 	validator(const validator &rhs) = delete;
 	validator &operator=(const validator &rhs) = delete;
 
+	/// @brief move constructor
 	validator(validator &&rhs) = default;
+
+	/// @brief move assignment operator
 	validator &operator=(validator &&rhs) = default;
 
 	friend class dictionary_parser;
 
+	/// @brief Add type_validator @a v to the list of type validators
 	void add_type_validator(type_validator &&v);
+
+	/// @brief Return the type validator for @a type_code, may return nullptr
 	const type_validator *get_validator_for_type(std::string_view type_code) const;
 
+	/// @brief Add category_validator @a v to the list of category validators
 	void add_category_validator(category_validator &&v);
+
+	/// @brief Return the category validator for @a category, may return nullptr
 	const category_validator *get_validator_for_category(std::string_view category) const;
 
+	/// @brief Add link_validator @a v to the list of link validators
 	void add_link_validator(link_validator &&v);
+
+	/// @brief Return the list of link validators for which the parent is @a category
 	std::vector<const link_validator *> get_links_for_parent(std::string_view category) const;
+
+	/// @brief Return the list of link validators for which the child is @a category
 	std::vector<const link_validator *> get_links_for_child(std::string_view category) const;
 
-	void report_error(const std::string &msg, bool fatal) const;
+	/// @brief Bottleneck function to report an error in validation
+	void report_error(validation_error err, bool fatal = true) const
+	{
+		report_error(make_error_code(err), fatal);
+	}
 
-	const std::string &name() const { return m_name; }
-	void set_name(const std::string &name) { m_name = name; }
+	/// @brief Bottleneck function to report an error in validation
+	void report_error(std::error_code ec, bool fatal = true) const;
 
-	const std::string &version() const { return m_version; }
-	void version(const std::string &version) { m_version = version; }
+	/// @brief Bottleneck function to report an error in validation
+	void report_error(validation_error err, std::string_view category,
+		std::string_view item, bool fatal = true) const
+	{
+		report_error(make_error_code(err), category, item, fatal);
+	}
+
+	/// @brief Bottleneck function to report an error in validation
+	void report_error(std::error_code ec, std::string_view category,
+		std::string_view item, bool fatal = true) const;
+
+	const std::string &name() const { return m_name; }        ///< Get the name of this validator
+	void set_name(const std::string &name) { m_name = name; } ///< Set the name of this validator
+
+	const std::string &version() const { return m_version; }              ///< Get the version of this validator
+	void set_version(const std::string &version) { m_version = version; } ///< Set the version of this validator
 
   private:
 	// name is fully qualified here:
@@ -217,21 +475,25 @@ class validator
 };
 
 // --------------------------------------------------------------------
+
+/**
+ * @brief Validators are globally unique objects, use the validator_factory
+ * class to construct them. This class is a singleton.
+ */
+
 class validator_factory
 {
   public:
-	static validator_factory &instance()
-	{
-		static validator_factory s_instance;
-		return s_instance;
-	}
+	/// @brief Return the singleton instance
+	static validator_factory &instance();
 
+	/// @brief Return the validator with name @a dictionary_name
 	const validator &operator[](std::string_view dictionary_name);
 
+	/// @brief Construct a new validator with name @a name from the data in @a is
 	const validator &construct_validator(std::string_view name, std::istream &is);
 
   private:
-
 	// --------------------------------------------------------------------
 
 	validator_factory() = default;

libcifpp.pc.in

@@ -1,13 +0,0 @@
-prefix=@prefix@
-exec_prefix=@exec_prefix@
-libdir=@libdir@
-includedir=@includedir@
-datalibdir=@datarootdir@/libcifpp
-
-Name: libcifpp
-Description: C++ library for the manipulation of mmCIF files.
-Version: @PACKAGE_VERSION@
-
-Requires: zlib
-Libs: -L${libdir} -lcifpp
-Cflags: -I${includedir} -pthread

regex/CMakeLists.txt

@@ -1,69 +0,0 @@
-# Copyright 2018 Mike Dev
-# Copyright 2019 Peter Dimov
-# Distributed under the Boost Software License, Version 1.0.
-# See accompanying file LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt
-
-# We support CMake 3.5, but prefer 3.16 policies and behavior
-cmake_minimum_required(VERSION 3.5...3.16)
-
-project(boost_regex VERSION "${BOOST_SUPERPROJECT_VERSION}" LANGUAGES CXX)
-
-add_library(boost_regex INTERFACE)
-add_library(Boost::regex ALIAS boost_regex)
-
-target_include_directories(boost_regex INTERFACE include)
-
-option(BOOST_REGEX_STANDALONE "Boost.Regex: Enable Standalone Mode (i.e. no Boost dependencies)")
-
-if(NOT BOOST_REGEX_STANDALONE)
-
-   target_link_libraries(boost_regex
-     INTERFACE
-       Boost::config
-       Boost::throw_exception
-       Boost::predef
-       Boost::assert
-   )
-
-else()
-
-   target_compile_definitions(boost_regex
-     INTERFACE BOOST_REGEX_STANDALONE
-   )
-
-endif()
-
-find_package(ICU COMPONENTS data i18n uc QUIET)
-#option(BOOST_REGEX_ENABLE_ICU "Boost.Regex: enable ICU support" ${ICU_FOUND})
-
-if(ICU_FOUND)
-
-   add_library(boost_regex_icu INTERFACE)
-   add_library(Boost::regex_icu ALIAS boost_regex_icu)
-
-   target_include_directories(boost_regex_icu INTERFACE include)
-
-   if(NOT BOOST_REGEX_STANDALONE)
-
-      target_link_libraries(boost_regex_icu
-        INTERFACE
-          Boost::config
-          Boost::throw_exception
-          Boost::predef
-          Boost::assert
-      )
-
-   else()
-
-      target_compile_definitions(boost_regex_icu
-        INTERFACE BOOST_REGEX_STANDALONE
-      )
-
-   endif()
-
-   find_package(ICU COMPONENTS data i18n uc REQUIRED)
-
-   target_link_libraries(boost_regex_icu INTERFACE ICU::data ICU::i18n ICU::uc)
-
-endif()
-

regex/LICENSE_1_0.txt

@@ -1,23 +0,0 @@
-Boost Software License - Version 1.0 - August 17th, 2003
-
-Permission is hereby granted, free of charge, to any person or organization
-obtaining a copy of the software and accompanying documentation covered by
-this license (the "Software") to use, reproduce, display, distribute,
-execute, and transmit the Software, and to prepare derivative works of the
-Software, and to permit third-parties to whom the Software is furnished to
-do so, all subject to the following:
-
-The copyright notices in the Software and this entire statement, including
-the above license grant, this restriction and the following disclaimer,
-must be included in all copies of the Software, in whole or in part, and
-all derivative works of the Software, unless such copies or derivative
-works are solely in the form of machine-executable object code generated by
-a source language processor.
-
-THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
-IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
-FITNESS FOR A PARTICULAR PURPOSE, TITLE AND NON-INFRINGEMENT. IN NO EVENT
-SHALL THE COPYRIGHT HOLDERS OR ANYONE DISTRIBUTING THE SOFTWARE BE LIABLE
-FOR ANY DAMAGES OR OTHER LIABILITY, WHETHER IN CONTRACT, TORT OR OTHERWISE,
-ARISING FROM, OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER
-DEALINGS IN THE SOFTWARE.

regex/README.md

@@ -1,47 +0,0 @@
-Boost Regex Library
-============================
-
-The Boost Regex library provides regular expression support for C++, this library is the ancestor to std::regex and still goes beyond
-and offers some advantages to, the standard version.
-
-The full documentation is available on [boost.org](http://www.boost.org/doc/libs/release/libs/regex/index.html).
-
-## Standalone Mode ##
-
- This library may now be used in "standalone" mode without the rest of the Boost C++ libraries, in order to do this you must either:
-
-* Have a C++17 compiler that supports __has_include, in this case if <boost/config.hpp> is not present then the library will automoatically enter standalone mode. Or:
-* Define BOOST_REGEX_STANDALONE when building.
-
-The main difference between the 2 modes, is that when Boost.Config is present the library will automatically configure itself around various compiler defects. In particular in order to use the library with exception support turned off, you will either need a copy of Boost.Config in your include path, or else manually define BOOST_NO_EXCEPTIONS when building.
-
-In any event, to obtain a standalone version of this library, simply download a .zip of the "master" branch of this repository.
-
-## Support, bugs and feature requests ##
-
-Bugs and feature requests can be reported through the [Gitub issue tracker](https://github.com/boostorg/regex/issues)
-(see [open issues](https://github.com/boostorg/regex/issues) and
-[closed issues](https://github.com/boostorg/regex/issues?utf8=%E2%9C%93&q=is%3Aissue+is%3Aclosed)).
-
-You can submit your changes through a [pull request](https://github.com/boostorg/regex/pulls).
-
-There is no mailing-list specific to Boost Regex, although you can use the general-purpose Boost [mailing-list](http://lists.boost.org/mailman/listinfo.cgi/boost-users) using the tag [regex].
-
-
-## Development ##
-
-Clone the whole boost project, which includes the individual Boost projects as submodules ([see boost+git doc](https://github.com/boostorg/boost/wiki/Getting-Started)): 
-
-    git clone https://github.com/boostorg/boost
-    cd boost
-    git submodule update --init
-
-The Boost Regex Library is located in `libs/regex/`. 
-
-### Running tests ###
-First, make sure you are in `libs/regex/test`. 
-You can either run all the tests listed in `Jamfile.v2` or run a single test:
-
-    ../../../b2                        <- run all tests
-    ../../../b2 regex_regress          <- single test
-

regex/include/boost/cregex.hpp

@@ -1,43 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org/libs/regex for most recent version.
-  *   FILE         cregex.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares POSIX API functions
-  *                + boost::RegEx high level wrapper.
-  */
-
-#ifndef BOOST_RE_CREGEX_HPP
-#define BOOST_RE_CREGEX_HPP
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#include <boost/regex/config.hpp>
-#endif
-
-#ifdef BOOST_REGEX_CXX03
-#include <boost/regex/v4/cregex.hpp>
-#else
-#include <boost/regex/v5/cregex.hpp>
-#endif
-
-#endif /* include guard */
-
-
-
-
-
-
-
-
-
-

regex/include/boost/regex.h

@@ -1,100 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2000
- * Dr John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org/libs/regex for documentation.
-  *   FILE         regex.h
-  *   VERSION      3.12
-  *   DESCRIPTION: Declares POSIX API functions
-  */
-
-#ifndef BOOST_RE_REGEX_H
-#define BOOST_RE_REGEX_H
-
-#include <boost/cregex.hpp>
-
-/*
-*  add using declarations to bring POSIX API functions into
-* global scope, only if this is C++ (and not C).
-*/
-#ifdef __cplusplus
-
-using boost::regoff_t;
-using boost::regex_tA;
-using boost::regmatch_t;
-using boost::REG_BASIC;
-using boost::REG_EXTENDED;
-using boost::REG_ICASE;
-using boost::REG_NOSUB;
-using boost::REG_NEWLINE;
-using boost::REG_NOSPEC;
-using boost::REG_PEND;
-using boost::REG_DUMP;
-using boost::REG_NOCOLLATE;
-using boost::REG_ESCAPE_IN_LISTS;
-using boost::REG_NEWLINE_ALT;
-using boost::REG_PERL;
-using boost::REG_AWK;
-using boost::REG_GREP;
-using boost::REG_EGREP;
-using boost::REG_ASSERT;
-using boost::REG_INVARG;
-using boost::REG_ATOI;
-using boost::REG_ITOA;
-
-using boost::REG_NOTBOL;
-using boost::REG_NOTEOL;
-using boost::REG_STARTEND;
-
-using boost::reg_comp_flags;
-using boost::reg_exec_flags;
-using boost::regcompA;
-using boost::regerrorA;
-using boost::regexecA;
-using boost::regfreeA;
-
-#ifndef BOOST_NO_WREGEX
-using boost::regcompW;
-using boost::regerrorW;
-using boost::regexecW;
-using boost::regfreeW;
-using boost::regex_tW;
-#endif
-
-using boost::REG_NOERROR;
-using boost::REG_NOMATCH;
-using boost::REG_BADPAT;
-using boost::REG_ECOLLATE;
-using boost::REG_ECTYPE;
-using boost::REG_EESCAPE;
-using boost::REG_ESUBREG;
-using boost::REG_EBRACK;
-using boost::REG_EPAREN;
-using boost::REG_EBRACE;
-using boost::REG_BADBR;
-using boost::REG_ERANGE;
-using boost::REG_ESPACE;
-using boost::REG_BADRPT;
-using boost::REG_EEND;
-using boost::REG_ESIZE;
-using boost::REG_ERPAREN;
-using boost::REG_EMPTY;
-using boost::REG_E_MEMORY;
-using boost::REG_E_UNKNOWN;
-using boost::reg_errcode_t;
-
-#endif /* __cplusplus */
-
-#endif /* BOOST_RE_REGEX_H */
-
-
-
-

regex/include/boost/regex.hpp

@@ -1,41 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org/libs/regex for documentation.
-  *   FILE         regex.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares boost::basic_regex<> and associated
-  *                functions and classes. This header is the main
-  *                entry point for the template regex code.
-  */
-
-
-/* start with C compatibility API */
-
-#ifndef BOOST_RE_REGEX_HPP
-#define BOOST_RE_REGEX_HPP
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#include <boost/regex/config.hpp>
-#endif
-
-#ifdef BOOST_REGEX_CXX03
-#include <boost/regex/v4/regex.hpp>
-#else
-#include <boost/regex/v5/regex.hpp>
-#endif
-
-#endif  // include
-
-
-
-

regex/include/boost/regex/config.hpp

@@ -1,480 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the
- * Boost Software License, Version 1.0. (See accompanying file
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         config.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: regex extended config setup.
-  */
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#define BOOST_REGEX_CONFIG_HPP
-
-#if !((__cplusplus >= 201103L) || (defined(_MSC_VER) && (_MSC_VER >= 1600)) || defined(BOOST_REGEX_CXX03))
-#  define BOOST_REGEX_CXX03
-#endif
-
-#if defined(BOOST_REGEX_RECURSIVE) && !defined(BOOST_REGEX_CXX03)
-#  define BOOST_REGEX_CXX03
-#endif
-
-#if defined(__has_include)
-#if !defined(BOOST_REGEX_STANDALONE) && !__has_include(<boost/version.hpp>)
-#define BOOST_REGEX_STANDALONE
-#endif
-#endif
-
-/*
- * Borland C++ Fix/error check
- * this has to go *before* we include any std lib headers:
- */
-#if defined(__BORLANDC__) && !defined(__clang__)
-#  include <boost/regex/config/borland.hpp>
-#endif
-#ifndef BOOST_REGEX_STANDALONE
-#include <boost/version.hpp>
-#endif
-
-/*************************************************************************
-*
-* Asserts:
-*
-*************************************************************************/
-
-#ifdef BOOST_REGEX_STANDALONE
-#include <cassert>
-#  define BOOST_REGEX_ASSERT(x) assert(x)
-#else
-#include <boost/assert.hpp>
-#  define BOOST_REGEX_ASSERT(x) BOOST_ASSERT(x)
-#endif
-
-/*****************************************************************************
- *
- *  Include all the headers we need here:
- *
- ****************************************************************************/
-
-#ifdef __cplusplus
-
-#  ifndef BOOST_REGEX_USER_CONFIG
-#     define BOOST_REGEX_USER_CONFIG <boost/regex/user.hpp>
-#  endif
-
-#  include BOOST_REGEX_USER_CONFIG
-
-#ifndef BOOST_REGEX_STANDALONE
-#  include <boost/config.hpp>
-#  include <boost/predef.h>
-#endif
-
-#else
-   /*
-    * C build,
-    * don't include <boost/config.hpp> because that may
-    * do C++ specific things in future...
-    */
-#  include <stdlib.h>
-#  include <stddef.h>
-#  ifdef _MSC_VER
-#     define BOOST_MSVC _MSC_VER
-#  endif
-#endif
-
-
-/****************************************************************************
-*
-* Legacy support:
-*
-*******************************************************************************/
-
-#if defined(BOOST_NO_STD_LOCALE) || defined(BOOST_NO_CXX11_HDR_MUTEX) || defined(BOOST_NO_CXX11_HDR_TYPE_TRAITS) \
-   || defined(BOOST_NO_CXX11_HDR_ATOMIC) || defined(BOOST_NO_CXX11_ALLOCATOR) || defined(BOOST_NO_CXX11_SMART_PTR) \
-   || defined(BOOST_NO_CXX11_STATIC_ASSERT) || defined(BOOST_NO_NOEXCEPT)
-#ifndef BOOST_REGEX_CXX03
-#  define BOOST_REGEX_CXX03
-#endif
-#endif
-
-/*****************************************************************************
- *
- *  Boilerplate regex config options:
- *
- ****************************************************************************/
-
-/* Obsolete macro, use BOOST_VERSION instead: */
-#define BOOST_RE_VERSION 500
-
-/* fix: */
-#if defined(_UNICODE) && !defined(UNICODE)
-#define UNICODE
-#endif
-
-#define BOOST_REGEX_JOIN(X, Y) BOOST_REGEX_DO_JOIN(X, Y)
-#define BOOST_REGEX_DO_JOIN(X, Y) BOOST_REGEX_DO_JOIN2(X,Y)
-#define BOOST_REGEX_DO_JOIN2(X, Y) X##Y
-
-#ifdef BOOST_FALLTHROUGH
-#  define BOOST_REGEX_FALLTHROUGH BOOST_FALLTHROUGH
-#else
-
-#if defined(__clang__) && (__cplusplus >= 201103L) && defined(__has_warning)
-#  if __has_feature(cxx_attributes) && __has_warning("-Wimplicit-fallthrough")
-#    define BOOST_REGEX_FALLTHROUGH [[clang::fallthrough]]
-#  endif
-#endif
-#if !defined(BOOST_REGEX_FALLTHROUGH) && defined(__INTEL_COMPILER) && (__INTEL_COMPILER >= 1800) && (__cplusplus >= 201703)
-#  define BOOST_REGEX_FALLTHROUGH [[fallthrough]]
-#endif
-#if !defined(BOOST_REGEX_FALLTHROUGH) && defined(__GNUC__) && (__GNUC__ >= 7)
-#  define BOOST_REGEX_FALLTHROUGH __attribute__((fallthrough))
-#endif
-
-#if !defined(BOOST_REGEX_FALLTHROUGH)
-#  define BOOST_REGEX_FALLTHROUGH
-#endif
-#endif
-
-#ifdef BOOST_NORETURN
-#  define BOOST_REGEX_NORETURN BOOST_NORETURN
-#else
-#  define BOOST_REGEX_NORETURN
-#endif
-
-
-/*
-* Define a macro for the namespace that details are placed in, this includes the Boost
-* version number to avoid mismatched header and library versions:
-*/
-#define BOOST_REGEX_DETAIL_NS BOOST_REGEX_JOIN(re_detail_, BOOST_RE_VERSION)
-
-/*
- * Fix for gcc prior to 3.4: std::ctype<wchar_t> doesn't allow
- * masks to be combined, for example:
- * std::use_facet<std::ctype<wchar_t> >.is(std::ctype_base::lower|std::ctype_base::upper, L'a');
- * returns *false*.
- */
-#if defined(__GLIBCPP__) && defined(BOOST_REGEX_CXX03)
-#  define BOOST_REGEX_BUGGY_CTYPE_FACET
-#endif
-
-/*
- * If there isn't good enough wide character support then there will
- * be no wide character regular expressions:
- */
-#if (defined(BOOST_NO_CWCHAR) || defined(BOOST_NO_CWCTYPE) || defined(BOOST_NO_STD_WSTRING))
-#  if !defined(BOOST_NO_WREGEX)
-#     define BOOST_NO_WREGEX
-#  endif
-#else
-#  if defined(__sgi) && (defined(__SGI_STL_PORT) || defined(_STLPORT_VERSION))
-      /* STLPort on IRIX is misconfigured: <cwctype> does not compile
-       * as a temporary fix include <wctype.h> instead and prevent inclusion
-       * of STLPort version of <cwctype> */
-#     include <wctype.h>
-#     define __STLPORT_CWCTYPE
-#     define _STLP_CWCTYPE
-#  endif
-
-#if defined(__cplusplus) && defined(BOOST_REGEX_CXX03)
-#  include <boost/regex/config/cwchar.hpp>
-#endif
-
-#endif
-
-/*
- * If Win32 support has been disabled for boost in general, then
- * it is for regex in particular:
- */
-#if defined(BOOST_DISABLE_WIN32) && !defined(BOOST_REGEX_NO_W32)
-#  define BOOST_REGEX_NO_W32
-#endif
-
-/* disable our own file-iterators and mapfiles if we can't
- * support them: */
-#if defined(_WIN32)
-#  if defined(BOOST_REGEX_NO_W32) || BOOST_PLAT_WINDOWS_RUNTIME
-#    define BOOST_REGEX_NO_FILEITER
-#  endif
-#else /* defined(_WIN32) */
-#  if !defined(BOOST_HAS_DIRENT_H)
-#    define BOOST_REGEX_NO_FILEITER
-#  endif
-#endif
-
-/* backwards compatibitity: */
-#if defined(BOOST_RE_NO_LIB)
-#  define BOOST_REGEX_NO_LIB
-#endif
-
-#if defined(__GNUC__) && !defined(_MSC_VER) && (defined(_WIN32) || defined(__CYGWIN__))
-/* gcc on win32 has problems if you include <windows.h>
-   (sporadically generates bad code). */
-#  define BOOST_REGEX_NO_W32
-#endif
-#if defined(__COMO__) && !defined(BOOST_REGEX_NO_W32) && !defined(_MSC_EXTENSIONS)
-#  define BOOST_REGEX_NO_W32
-#endif
-
-#ifdef BOOST_REGEX_STANDALONE
-#  if defined(_MSC_VER) && !defined(__clang__) && !defined(__GNUC__)
-#     define BOOST_REGEX_MSVC _MSC_VER
-#endif
-#elif defined(BOOST_MSVC)
-#  define BOOST_REGEX_MSVC BOOST_MSVC
-#endif
-
-
-/*****************************************************************************
- *
- *  Set up dll import/export options:
- *
- ****************************************************************************/
-
-#if (defined(BOOST_REGEX_DYN_LINK) || defined(BOOST_ALL_DYN_LINK)) && !defined(BOOST_REGEX_STATIC_LINK) && defined(BOOST_SYMBOL_IMPORT)
-#  if defined(BOOST_REGEX_SOURCE)
-#     define BOOST_REGEX_BUILD_DLL
-#     define BOOST_REGEX_DECL BOOST_SYMBOL_EXPORT
-#  else
-#     define BOOST_REGEX_DECL BOOST_SYMBOL_IMPORT
-#  endif
-#else
-#  define BOOST_REGEX_DECL
-#endif
-
-#ifdef BOOST_REGEX_CXX03
-#if !defined(BOOST_REGEX_NO_LIB) && !defined(BOOST_REGEX_SOURCE) && !defined(BOOST_ALL_NO_LIB) && defined(__cplusplus)
-#  define BOOST_LIB_NAME boost_regex
-#  if defined(BOOST_REGEX_DYN_LINK) || defined(BOOST_ALL_DYN_LINK)
-#     define BOOST_DYN_LINK
-#  endif
-#  ifdef BOOST_REGEX_DIAG
-#     define BOOST_LIB_DIAGNOSTIC
-#  endif
-#  include <boost/config/auto_link.hpp>
-#endif
-#endif
-
-/*****************************************************************************
- *
- *  Set up function call type:
- *
- ****************************************************************************/
-
-#if defined(_MSC_VER) && defined(_MSC_EXTENSIONS)
-#if defined(_DEBUG) || defined(__MSVC_RUNTIME_CHECKS) || defined(_MANAGED) || defined(BOOST_REGEX_NO_FASTCALL)
-#  define BOOST_REGEX_CALL __cdecl
-#else
-#  define BOOST_REGEX_CALL __fastcall
-#endif
-#  define BOOST_REGEX_CCALL __cdecl
-#endif
-
-#if defined(__BORLANDC__) && !defined(BOOST_DISABLE_WIN32)
-#if defined(__clang__)
-#  define BOOST_REGEX_CALL __cdecl
-#  define BOOST_REGEX_CCALL __cdecl
-#else
-#  define BOOST_REGEX_CALL __fastcall
-#  define BOOST_REGEX_CCALL __stdcall
-#endif
-#endif
-
-#ifndef BOOST_REGEX_CALL
-#  define BOOST_REGEX_CALL
-#endif
-#ifndef BOOST_REGEX_CCALL
-#define BOOST_REGEX_CCALL
-#endif
-
-/*****************************************************************************
- *
- *  Set up localisation model:
- *
- ****************************************************************************/
-
-/* backwards compatibility: */
-#ifdef BOOST_RE_LOCALE_C
-#  define BOOST_REGEX_USE_C_LOCALE
-#endif
-
-#ifdef BOOST_RE_LOCALE_CPP
-#  define BOOST_REGEX_USE_CPP_LOCALE
-#endif
-
-#if defined(__CYGWIN__)
-#  define BOOST_REGEX_USE_C_LOCALE
-#endif
-
-/* use C++ locale when targeting windows store */
-#if BOOST_PLAT_WINDOWS_RUNTIME
-#  define BOOST_REGEX_USE_CPP_LOCALE
-#  define BOOST_REGEX_NO_WIN32_LOCALE
-#endif
-
-/* Win32 defaults to native Win32 locale: */
-#if defined(_WIN32) && \
-    !defined(BOOST_REGEX_USE_WIN32_LOCALE) && \
-    !defined(BOOST_REGEX_USE_C_LOCALE) && \
-    !defined(BOOST_REGEX_USE_CPP_LOCALE) && \
-    !defined(BOOST_REGEX_NO_W32) && \
-    !defined(BOOST_REGEX_NO_WIN32_LOCALE)
-#  define BOOST_REGEX_USE_WIN32_LOCALE
-#endif
-/* otherwise use C++ locale if supported: */
-#if !defined(BOOST_REGEX_USE_WIN32_LOCALE) && !defined(BOOST_REGEX_USE_C_LOCALE) && !defined(BOOST_REGEX_USE_CPP_LOCALE) && !defined(BOOST_NO_STD_LOCALE)
-#  define BOOST_REGEX_USE_CPP_LOCALE
-#endif
-/* otherwise use C locale: */
-#if !defined(BOOST_REGEX_USE_WIN32_LOCALE) && !defined(BOOST_REGEX_USE_C_LOCALE) && !defined(BOOST_REGEX_USE_CPP_LOCALE)
-#  define BOOST_REGEX_USE_C_LOCALE
-#endif
-
-#ifndef BOOST_REGEX_MAX_STATE_COUNT
-#  define BOOST_REGEX_MAX_STATE_COUNT 100000000
-#endif
-
-
-/*****************************************************************************
- *
- *  Error Handling for exception free compilers:
- *
- ****************************************************************************/
-
-#ifdef BOOST_NO_EXCEPTIONS
-/*
- * If there are no exceptions then we must report critical-errors
- * the only way we know how; by terminating.
- */
-#include <stdexcept>
-#include <string>
-#include <boost/throw_exception.hpp>
-
-#  define BOOST_REGEX_NOEH_ASSERT(x)\
-if(0 == (x))\
-{\
-   std::string s("Error: critical regex++ failure in: ");\
-   s.append(#x);\
-   std::runtime_error e(s);\
-   boost::throw_exception(e);\
-}
-#else
-/*
- * With exceptions then error handling is taken care of and
- * there is no need for these checks:
- */
-#  define BOOST_REGEX_NOEH_ASSERT(x)
-#endif
-
-
-/*****************************************************************************
- *
- *  Stack protection under MS Windows:
- *
- ****************************************************************************/
-
-#if !defined(BOOST_REGEX_NO_W32) && !defined(BOOST_REGEX_V3)
-#  if(defined(_WIN32) || defined(_WIN64) || defined(_WINCE)) \
-        && !(defined(__GNUC__) || defined(__BORLANDC__) && defined(__clang__)) \
-        && !(defined(__BORLANDC__) && (__BORLANDC__ >= 0x600)) \
-        && !(defined(__MWERKS__) && (__MWERKS__ <= 0x3003))
-#     define BOOST_REGEX_HAS_MS_STACK_GUARD
-#  endif
-#elif defined(BOOST_REGEX_HAS_MS_STACK_GUARD)
-#  undef BOOST_REGEX_HAS_MS_STACK_GUARD
-#endif
-
-#if defined(__cplusplus) && defined(BOOST_REGEX_HAS_MS_STACK_GUARD)
-
-namespace boost{
-namespace BOOST_REGEX_DETAIL_NS{
-
-BOOST_REGEX_DECL void BOOST_REGEX_CALL reset_stack_guard_page();
-
-}
-}
-
-#endif
-
-
-/*****************************************************************************
- *
- *  Algorithm selection and configuration.
- *  These options are now obsolete for C++11 and later (regex v5).
- *
- ****************************************************************************/
-
-#if !defined(BOOST_REGEX_RECURSIVE) && !defined(BOOST_REGEX_NON_RECURSIVE)
-#  if defined(BOOST_REGEX_HAS_MS_STACK_GUARD) && !defined(_STLP_DEBUG) && !defined(__STL_DEBUG) && !(defined(_MSC_VER) && (_MSC_VER >= 1400)) && defined(BOOST_REGEX_CXX03)
-#     define BOOST_REGEX_RECURSIVE
-#  else
-#     define BOOST_REGEX_NON_RECURSIVE
-#  endif
-#endif
-
-#ifdef BOOST_REGEX_NON_RECURSIVE
-#  ifdef BOOST_REGEX_RECURSIVE
-#     error "Can't set both BOOST_REGEX_RECURSIVE and BOOST_REGEX_NON_RECURSIVE"
-#  endif
-#  ifndef BOOST_REGEX_BLOCKSIZE
-#     define BOOST_REGEX_BLOCKSIZE 4096
-#  endif
-#  if BOOST_REGEX_BLOCKSIZE < 512
-#     error "BOOST_REGEX_BLOCKSIZE must be at least 512"
-#  endif
-#  ifndef BOOST_REGEX_MAX_BLOCKS
-#     define BOOST_REGEX_MAX_BLOCKS 1024
-#  endif
-#  ifdef BOOST_REGEX_HAS_MS_STACK_GUARD
-#     undef BOOST_REGEX_HAS_MS_STACK_GUARD
-#  endif
-#  ifndef BOOST_REGEX_MAX_CACHE_BLOCKS
-#     define BOOST_REGEX_MAX_CACHE_BLOCKS 16
-#  endif
-#endif
-
-
-/*****************************************************************************
- *
- *  Diagnostics:
- *
- ****************************************************************************/
-
-#ifdef BOOST_REGEX_CONFIG_INFO
-BOOST_REGEX_DECL void BOOST_REGEX_CALL print_regex_library_info();
-#endif
-
-#if defined(BOOST_REGEX_DIAG)
-#  pragma message ("BOOST_REGEX_DECL" BOOST_STRINGIZE(=BOOST_REGEX_DECL))
-#  pragma message ("BOOST_REGEX_CALL" BOOST_STRINGIZE(=BOOST_REGEX_CALL))
-#  pragma message ("BOOST_REGEX_CCALL" BOOST_STRINGIZE(=BOOST_REGEX_CCALL))
-#ifdef BOOST_REGEX_USE_C_LOCALE
-#  pragma message ("Using C locale in regex traits class")
-#elif BOOST_REGEX_USE_CPP_LOCALE
-#  pragma message ("Using C++ locale in regex traits class")
-#else
-#  pragma message ("Using Win32 locale in regex traits class")
-#endif
-#if defined(BOOST_REGEX_DYN_LINK) || defined(BOOST_ALL_DYN_LINK)
-#  pragma message ("Dynamic linking enabled")
-#endif
-#if defined(BOOST_REGEX_NO_LIB) || defined(BOOST_ALL_NO_LIB)
-#  pragma message ("Auto-linking disabled")
-#endif
-#ifdef BOOST_REGEX_NO_EXTERNAL_TEMPLATES
-#  pragma message ("Extern templates disabled")
-#endif
-
-#endif
-
-#endif
-

regex/include/boost/regex/config/borland.hpp

@@ -1,72 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         boost/regex/config/borland.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: regex borland-specific config setup.
-  */
-
-
-#if defined(__BORLANDC__) && !defined(__clang__)
-#  if (__BORLANDC__ == 0x550) || (__BORLANDC__ == 0x551)
-      // problems with std::basic_string and dll RTL:
-#     if defined(_RTLDLL) && defined(_RWSTD_COMPILE_INSTANTIATE)
-#        ifdef BOOST_REGEX_BUILD_DLL
-#           error _RWSTD_COMPILE_INSTANTIATE must not be defined when building regex++ as a DLL
-#        else
-#           pragma message("Defining _RWSTD_COMPILE_INSTANTIATE when linking to the DLL version of the RTL may produce memory corruption problems in std::basic_string, as a result of separate versions of basic_string's static data in the RTL and you're exe/dll: be warned!!")
-#        endif
-#     endif
-#     ifndef _RTLDLL
-         // this is harmless for a staic link:
-#        define _RWSTD_COMPILE_INSTANTIATE
-#     endif
-      // external templates cause problems for some reason:
-#     define BOOST_REGEX_NO_EXTERNAL_TEMPLATES
-#  endif
-#  if (__BORLANDC__ <= 0x540) && !defined(BOOST_REGEX_NO_LIB) && !defined(_NO_VCL)
-      // C++ Builder 4 and earlier, we can't tell whether we should be using
-      // the VCL runtime or not, do a static link instead:
-#     define BOOST_REGEX_STATIC_LINK
-#  endif
-   //
-   // VCL support:
-   // if we're building a console app then there can't be any VCL (can there?)
-#  if !defined(__CONSOLE__) && !defined(_NO_VCL)
-#     define BOOST_REGEX_USE_VCL
-#  endif
-   //
-   // if this isn't Win32 then don't automatically select link
-   // libraries:
-   //
-#  ifndef _Windows
-#     ifndef BOOST_REGEX_NO_LIB
-#        define BOOST_REGEX_NO_LIB
-#     endif
-#     ifndef BOOST_REGEX_STATIC_LINK
-#        define BOOST_REGEX_STATIC_LINK
-#     endif
-#  endif
-
-#if __BORLANDC__ < 0x600
-//
-// string workarounds:
-//
-#include <cstring>
-#undef strcmp
-#undef strcpy
-#endif
-
-#endif
-
-

regex/include/boost/regex/config/cwchar.hpp

@@ -1,207 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         boost/regex/config/cwchar.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: regex wide character string fixes.
-  */
-
-#ifndef BOOST_REGEX_CONFIG_CWCHAR_HPP
-#define BOOST_REGEX_CONFIG_CWCHAR_HPP
-
-#include <cwchar>
-#include <cwctype>
-#include <boost/config.hpp>
-
-#if defined(__STD_RWCOMPILER_H__) || defined(_RWSTD_VER)
-// apparently this is required for the RW STL on Linux:
-#undef iswalnum
-#undef iswalpha
-#undef iswblank
-#undef iswcntrl
-#undef iswdigit
-#undef iswgraph
-#undef iswlower
-#undef iswprint
-#undef iswprint
-#undef iswpunct
-#undef iswspace
-#undef iswupper
-#undef iswxdigit
-#undef iswctype
-#undef towlower
-#undef towupper
-#undef towctrans
-#undef wctrans
-#undef wctype
-#endif
-
-namespace std{
-
-#ifndef BOOST_NO_STDC_NAMESPACE
-extern "C"{
-#endif
-
-#ifdef iswalnum
-inline int (iswalnum)(wint_t i)
-{ return iswalnum(i); }
-#undef iswalnum
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswalnum;
-#endif
-
-#ifdef iswalpha
-inline int (iswalpha)(wint_t i)
-{ return iswalpha(i); }
-#undef iswalpha
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswalpha;
-#endif
-
-#ifdef iswcntrl
-inline int (iswcntrl)(wint_t i)
-{ return iswcntrl(i); }
-#undef iswcntrl
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswcntrl;
-#endif
-
-#ifdef iswdigit
-inline int (iswdigit)(wint_t i)
-{ return iswdigit(i); }
-#undef iswdigit
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswdigit;
-#endif
-
-#ifdef iswgraph
-inline int (iswgraph)(wint_t i)
-{ return iswgraph(i); }
-#undef iswgraph
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswgraph;
-#endif
-
-#ifdef iswlower
-inline int (iswlower)(wint_t i)
-{ return iswlower(i); }
-#undef iswlower
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswlower;
-#endif
-
-#ifdef iswprint
-inline int (iswprint)(wint_t i)
-{ return iswprint(i); }
-#undef iswprint
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswprint;
-#endif
-
-#ifdef iswpunct
-inline int (iswpunct)(wint_t i)
-{ return iswpunct(i); }
-#undef iswpunct
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswpunct;
-#endif
-
-#ifdef iswspace
-inline int (iswspace)(wint_t i)
-{ return iswspace(i); }
-#undef iswspace
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswspace;
-#endif
-
-#ifdef iswupper
-inline int (iswupper)(wint_t i)
-{ return iswupper(i); }
-#undef iswupper
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswupper;
-#endif
-
-#ifdef iswxdigit
-inline int (iswxdigit)(wint_t i)
-{ return iswxdigit(i); }
-#undef iswxdigit
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::iswxdigit;
-#endif
-
-#ifdef towlower
-inline wint_t (towlower)(wint_t i)
-{ return towlower(i); }
-#undef towlower
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::towlower;
-#endif
-
-#ifdef towupper
-inline wint_t (towupper)(wint_t i)
-{ return towupper(i); }
-#undef towupper
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using :: towupper;
-#endif
-
-#ifdef wcscmp
-inline int (wcscmp)(const wchar_t *p1, const wchar_t *p2)
-{ return wcscmp(p1,p2); }
-#undef wcscmp
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::wcscmp;
-#endif
-
-#ifdef wcscoll
-inline int (wcscoll)(const wchar_t *p1, const wchar_t *p2)
-{ return wcscoll(p1,p2); }
-#undef wcscoll
-#elif defined(BOOST_NO_STDC_NAMESPACE) && !defined(UNDER_CE)
-using ::wcscoll;
-#endif
-
-#ifdef wcscpy
-inline wchar_t *(wcscpy)(wchar_t *p1, const wchar_t *p2)
-{ return wcscpy(p1,p2); }
-#undef wcscpy
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::wcscpy;
-#endif
-
-#ifdef wcslen
-inline size_t (wcslen)(const wchar_t *p)
-{ return wcslen(p); }
-#undef wcslen
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::wcslen;
-#endif
-
-#ifdef wcsxfrm
-size_t wcsxfrm(wchar_t *p1, const wchar_t *p2, size_t s)
-{ return wcsxfrm(p1,p2,s); }
-#undef wcsxfrm
-#elif defined(BOOST_NO_STDC_NAMESPACE)
-using ::wcsxfrm;
-#endif
-
-
-#ifndef BOOST_NO_STDC_NAMESPACE
-} // extern "C"
-#endif
-
-} // namespace std
-
-#endif
-

regex/include/boost/regex/icu.hpp

@@ -1,30 +0,0 @@
-/*
- *
- * Copyright (c) 2020
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         icu.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Unicode regular expressions on top of the ICU Library.
-  */
-
-#ifndef BOOST_REGEX_ICU_HPP
-#define BOOST_REGEX_ICU_HPP
-
-#include <boost/regex/config.hpp>
-
-#ifdef BOOST_REGEX_CXX03
-#include <boost/regex/v4/icu.hpp>
-#else
-#include <boost/regex/v5/icu.hpp>
-#endif
-
-#endif

regex/include/boost/regex/mfc.hpp

@@ -1,186 +0,0 @@
-/*
- *
- * Copyright (c) 2004
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         mfc.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Overloads and helpers for using MFC/ATL string types with Boost.Regex.
-  */
-
-#ifndef BOOST_REGEX_MFC_HPP
-#define BOOST_REGEX_MFC_HPP
-
-#include <atlsimpstr.h>
-#include <boost/regex.hpp>
-
-namespace boost{
-
-//
-// define the types used for TCHAR's:
-typedef basic_regex<TCHAR> tregex;
-typedef match_results<TCHAR const*> tmatch;
-typedef regex_iterator<TCHAR const*> tregex_iterator;
-typedef regex_token_iterator<TCHAR const*> tregex_token_iterator;
-
-// Obsolete. Remove
-#define SIMPLE_STRING_PARAM class B, bool b
-#define SIMPLE_STRING_ARG_LIST B, b
-
-//
-// define regex creation functions:
-//
-template <class B, bool b>
-inline basic_regex<B> 
-make_regex(const ATL::CSimpleStringT<B, b>& s, ::boost::regex_constants::syntax_option_type f = boost::regex_constants::normal)
-{
-   basic_regex<B> result(s.GetString(), s.GetString() + s.GetLength(), f);
-   return result;
-}
-//
-// regex_match overloads:
-//
-template <class B, bool b, class A, class T>
-inline bool regex_match(const ATL::CSimpleStringT<B, b>& s,
-                 match_results<const B*, A>& what,
-                 const basic_regex<B, T>& e,
-                 boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   return ::boost::regex_match(s.GetString(),
-                               s.GetString() + s.GetLength(),
-                               what,
-                               e,
-                               f);
-}
-
-template <class B, bool b, class T>
-inline bool regex_match(const ATL::CSimpleStringT<B, b>& s,
-                 const basic_regex<B, T>& e,
-                 boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   return ::boost::regex_match(s.GetString(),
-                               s.GetString() + s.GetLength(),
-                               e,
-                               f);
-}
-//
-// regex_search overloads:
-//
-template <class B, bool b, class A, class T>
-inline bool regex_search(const ATL::CSimpleStringT<B, b>& s,
-                 match_results<const B*, A>& what,
-                 const basic_regex<B, T>& e,
-                 boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   return ::boost::regex_search(s.GetString(),
-                               s.GetString() + s.GetLength(),
-                               what,
-                               e,
-                               f);
-}
-
-template <class B, bool b, class T>
-inline bool regex_search(const ATL::CSimpleStringT<B, b>& s,
-                 const basic_regex<B, T>& e,
-                 boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   return ::boost::regex_search(s.GetString(),
-                               s.GetString() + s.GetLength(),
-                               e,
-                               f);
-}
-//
-// regex_iterator creation:
-//
-template <class B, bool b>
-inline regex_iterator<B const*> 
-make_regex_iterator(const ATL::CSimpleStringT<B, b>& s, const basic_regex<B>& e, ::boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   regex_iterator<B const*> result(s.GetString(), s.GetString() + s.GetLength(), e, f);
-   return result;
-}
-
-template <class B, bool b>
-inline regex_token_iterator<B const*> 
-   make_regex_token_iterator(const ATL::CSimpleStringT<B, b>& s, const basic_regex<B>& e, int sub = 0, ::boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   regex_token_iterator<B const*> result(s.GetString(), s.GetString() + s.GetLength(), e, sub, f);
-   return result;
-}
-
-template <class B, bool b>
-inline regex_token_iterator<B const*> 
-make_regex_token_iterator(const ATL::CSimpleStringT<B, b>& s, const basic_regex<B>& e, const std::vector<int>& subs, ::boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   regex_token_iterator<B const*> result(s.GetString(), s.GetString() + s.GetLength(), e, subs, f);
-   return result;
-}
-
-template <class B, bool b, std::size_t N>
-inline regex_token_iterator<B const*> 
-make_regex_token_iterator(const ATL::CSimpleStringT<B, b>& s, const basic_regex<B>& e, const int (& subs)[N], ::boost::regex_constants::match_flag_type f = boost::regex_constants::match_default)
-{
-   regex_token_iterator<B const*> result(s.GetString(), s.GetString() + s.GetLength(), e, subs, f);
-   return result;
-}
-
-template <class OutputIterator, class BidirectionalIterator, class traits,
-          class B, bool b>
-OutputIterator regex_replace(OutputIterator out,
-                           BidirectionalIterator first,
-                           BidirectionalIterator last,
-                           const basic_regex<B, traits>& e,
-                           const ATL::CSimpleStringT<B, b>& fmt,
-                           match_flag_type flags = match_default)
-{
-   return ::boost::regex_replace(out, first, last, e, fmt.GetString(), flags);
-}
-
-namespace BOOST_REGEX_DETAIL_NS{
-
-template <class B, bool b>
-class mfc_string_out_iterator
-{
-   ATL::CSimpleStringT<B, b>* out;
-public:
-   mfc_string_out_iterator(ATL::CSimpleStringT<B, b>& s) : out(&s) {}
-   mfc_string_out_iterator& operator++() { return *this; }
-   mfc_string_out_iterator& operator++(int) { return *this; }
-   mfc_string_out_iterator& operator*() { return *this; }
-   mfc_string_out_iterator& operator=(B v) 
-   { 
-      out->AppendChar(v); 
-      return *this; 
-   }
-   typedef std::ptrdiff_t difference_type;
-   typedef B value_type;
-   typedef value_type* pointer;
-   typedef value_type& reference;
-   typedef std::output_iterator_tag iterator_category;
-};
-
-}
-
-template <class traits, class B, bool b>
-ATL::CSimpleStringT<B, b> regex_replace(const ATL::CSimpleStringT<B, b>& s,
-                            const basic_regex<B, traits>& e,
-                            const ATL::CSimpleStringT<B, b>& fmt,
-                            match_flag_type flags = match_default)
-{
-   ATL::CSimpleStringT<B, b> result(s.GetManager());
-   BOOST_REGEX_DETAIL_NS::mfc_string_out_iterator<B, b> i(result);
-   regex_replace(i, s.GetString(), s.GetString() + s.GetLength(), e, fmt.GetString(), flags);
-   return result;
-}
-
-} // namespace boost.
-
-#endif

regex/include/boost/regex/pattern_except.hpp

@@ -1,32 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         pattern_except.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares pattern-matching exception classes.
-  */
-
-#ifndef BOOST_RE_PAT_EXCEPT_HPP
-#define BOOST_RE_PAT_EXCEPT_HPP
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#include <boost/regex/config.hpp>
-#endif
-
-#ifdef BOOST_REGEX_CXX03
-#include <boost/regex/v4/pattern_except.hpp>
-#else
-#include <boost/regex/v5/pattern_except.hpp>
-#endif
-
-#endif

regex/include/boost/regex/pending/object_cache.hpp

@@ -1,29 +0,0 @@
-/*
- *
- * Copyright (c) 2004
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         object_cache.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Implements a generic object cache.
-  */
-
-#ifndef BOOST_REGEX_OBJECT_CACHE_HPP
-#define BOOST_REGEX_OBJECT_CACHE_HPP
-
-#include <boost/regex/config.hpp>
-#ifdef BOOST_REGEX_CXX03
-#include <boost/regex/v4/object_cache.hpp>
-#else
-#include <boost/regex/v5/object_cache.hpp>
-#endif
-
-#endif

regex/include/boost/regex/pending/static_mutex.hpp

@@ -1,182 +0,0 @@
-/*
- *
- * Copyright (c) 2004
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         static_mutex.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares static_mutex lock type, there are three different
-  *                implementations: POSIX pthreads, WIN32 threads, and portable,
-  *                these are described in more detail below.
-  */
-
-#ifndef BOOST_REGEX_STATIC_MUTEX_HPP
-#define BOOST_REGEX_STATIC_MUTEX_HPP
-
-#include <boost/config.hpp>
-#include <boost/regex/config.hpp> // dll import/export options.
-
-#ifdef BOOST_HAS_PTHREADS
-#include <pthread.h>
-#endif
-
-#if defined(BOOST_HAS_PTHREADS) && defined(PTHREAD_MUTEX_INITIALIZER)
-//
-// pthreads version:
-// simple wrap around a pthread_mutex_t initialized with
-// PTHREAD_MUTEX_INITIALIZER.
-//
-namespace boost{
-
-class static_mutex;
-
-#define BOOST_STATIC_MUTEX_INIT { PTHREAD_MUTEX_INITIALIZER, }
-
-class BOOST_REGEX_DECL scoped_static_mutex_lock
-{
-public:
-   scoped_static_mutex_lock(static_mutex& mut, bool lk = true);
-   ~scoped_static_mutex_lock();
-   inline bool locked()const
-   {
-      return m_have_lock;
-   }
-   inline operator void const*()const
-   {
-      return locked() ? this : 0;
-   }
-   void lock();
-   void unlock();
-private:
-   static_mutex& m_mutex;
-   bool m_have_lock;
-};
-
-class static_mutex
-{
-public:
-   typedef scoped_static_mutex_lock scoped_lock;
-   pthread_mutex_t m_mutex;
-};
-
-} // namespace boost
-#elif defined(BOOST_HAS_WINTHREADS)
-//
-// Win32 version:
-// Use a 32-bit int as a lock, along with a test-and-set
-// implementation using InterlockedCompareExchange.
-//
-
-#include <boost/cstdint.hpp>
-
-namespace boost{
-
-class BOOST_REGEX_DECL scoped_static_mutex_lock;
-
-class static_mutex
-{
-public:
-   typedef scoped_static_mutex_lock scoped_lock;
-   boost::int32_t m_mutex;
-};
-
-#define BOOST_STATIC_MUTEX_INIT { 0, }
-
-class BOOST_REGEX_DECL scoped_static_mutex_lock
-{
-public:
-   scoped_static_mutex_lock(static_mutex& mut, bool lk = true);
-   ~scoped_static_mutex_lock();
-   operator void const*()const
-   {
-      return locked() ? this : 0;
-   }
-   bool locked()const
-   {
-      return m_have_lock;
-   }
-   void lock();
-   void unlock();
-private:
-   static_mutex& m_mutex;
-   bool m_have_lock;
-   scoped_static_mutex_lock(const scoped_static_mutex_lock&);
-   scoped_static_mutex_lock& operator=(const scoped_static_mutex_lock&);
-};
-
-} // namespace
-
-#else
-//
-// Portable version of a static mutex based on Boost.Thread library:
-// This has to use a single mutex shared by all instances of static_mutex
-// because boost::call_once doesn't alow us to pass instance information
-// down to the initialisation proceedure.  In fact the initialisation routine
-// may need to be called more than once - but only once per instance.
-//
-// Since this preprocessor path is almost never taken, we hide these header
-// dependencies so that build tools don't find them.
-//
-#define BOOST_REGEX_H1 <boost/thread/once.hpp>
-#define BOOST_REGEX_H2 <boost/thread/recursive_mutex.hpp>
-#define BOOST_REGEX_H3 <boost/thread/lock_types.hpp>
-#include BOOST_REGEX_H1
-#include BOOST_REGEX_H2
-#include BOOST_REGEX_H3
-#undef BOOST_REGEX_H1
-#undef BOOST_REGEX_H2
-#undef BOOST_REGEX_H3
-
-namespace boost{
-
-class BOOST_REGEX_DECL scoped_static_mutex_lock;
-extern "C" BOOST_REGEX_DECL void boost_regex_free_static_mutex();
-
-class BOOST_REGEX_DECL static_mutex
-{
-public:
-   typedef scoped_static_mutex_lock scoped_lock;
-   static void init();
-   static boost::recursive_mutex* m_pmutex;
-   static boost::once_flag m_once;
-};
-
-#define BOOST_STATIC_MUTEX_INIT {  }
-
-class BOOST_REGEX_DECL scoped_static_mutex_lock
-{
-public:
-   scoped_static_mutex_lock(static_mutex& mut, bool lk = true);
-   ~scoped_static_mutex_lock();
-   operator void const*()const;
-   bool locked()const;
-   void lock();
-   void unlock();
-private:
-   boost::unique_lock<boost::recursive_mutex>* m_plock;
-   bool m_have_lock;
-};
-
-inline scoped_static_mutex_lock::operator void const*()const
-{
-   return locked() ? this : 0;
-}
-
-inline bool scoped_static_mutex_lock::locked()const
-{
-   return m_have_lock;
-}
-
-} // namespace
-
-#endif
-
-#endif

regex/include/boost/regex/pending/unicode_iterator.hpp

@@ -1,32 +0,0 @@
-/*
- *
- * Copyright (c) 2020
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         unicode_iterator.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Iterator adapters for converting between different Unicode encodings.
-  */
-
-#ifndef BOOST_REGEX_PENDING_UNICODE_ITERATOR_HPP
-#define BOOST_REGEX_PENDING_UNICODE_ITERATOR_HPP
-
-#include <boost/regex/config.hpp>
-
-#if defined(BOOST_REGEX_CXX03)
-#include <boost/regex/v4/unicode_iterator.hpp>
-#else
-#include <boost/regex/v5/unicode_iterator.hpp>
-#endif
-
-
-#endif // BOOST_REGEX_PENDING_UNICODE_ITERATOR_HPP
-

regex/include/boost/regex/regex_traits.hpp

@@ -1,39 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         regex_traits.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares regular expression traits classes.
-  */
-
-#ifndef BOOST_REGEX_TRAITS_HPP
-#define BOOST_REGEX_TRAITS_HPP
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#  include <boost/regex/config.hpp>
-#endif
-
-#  ifndef BOOST_REGEX_TRAITS_HPP_INCLUDED
-#ifdef BOOST_REGEX_CXX03
-#     include <boost/regex/v4/regex_traits.hpp>
-#else
-#     include <boost/regex/v5/regex_traits.hpp>
-#endif
-#  endif
-
-#endif // include
-
-
-
-
-

regex/include/boost/regex/user.hpp

@@ -1,95 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         user.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: User settable options.
-  */
-
-// define if you want the regex library to use the C locale
-// even on Win32:
-// #define BOOST_REGEX_USE_C_LOCALE
-
-// define this is you want the regex library to use the C++
-// locale:
-// #define BOOST_REGEX_USE_CPP_LOCALE
-
-// define this if the runtime library is a dll, and you
-// want BOOST_REGEX_DYN_LINK to set up dll exports/imports
-// with __declspec(dllexport)/__declspec(dllimport.)
-// #define BOOST_REGEX_HAS_DLL_RUNTIME
-
-// define this if you want to dynamically link to regex,
-// if the runtime library is also a dll (Probably Win32 specific,
-// and has no effect unless BOOST_REGEX_HAS_DLL_RUNTIME is set):
-// #define BOOST_REGEX_DYN_LINK
-
-// define this if you don't want the lib to automatically
-// select its link libraries:
-// #define BOOST_REGEX_NO_LIB
-
-// define this if templates with switch statements cause problems:
-// #define BOOST_REGEX_NO_TEMPLATE_SWITCH_MERGE
- 
-// define this to disable Win32 support when available:
-// #define BOOST_REGEX_NO_W32
-
-// define this if bool is not a real type:
-// #define BOOST_REGEX_NO_BOOL
-
-// define this if no template instances are to be placed in
-// the library rather than users object files:
-// #define BOOST_REGEX_NO_EXTERNAL_TEMPLATES
-
-// define this if the forward declarations in regex_fwd.hpp
-// cause more problems than they are worth:
-// #define BOOST_REGEX_NO_FWD
-
-// define this if your compiler supports MS Windows structured
-// exception handling.
-// #define BOOST_REGEX_HAS_MS_STACK_GUARD
-
-// define this if you want to use the recursive algorithm
-// even if BOOST_REGEX_HAS_MS_STACK_GUARD is not defined.
-// NOTE: OBSOLETE!!
-// #define BOOST_REGEX_RECURSIVE
-
-// define this if you want to use the non-recursive
-// algorithm, even if the recursive version would be the default.
-// NOTE: OBSOLETE!!
-// #define BOOST_REGEX_NON_RECURSIVE
-
-// define this if you want to set the size of the memory blocks
-// used by the non-recursive algorithm.
-// #define BOOST_REGEX_BLOCKSIZE 4096
-
-// define this if you want to set the maximum number of memory blocks
-// used by the non-recursive algorithm.
-// #define BOOST_REGEX_MAX_BLOCKS 1024
-
-// define this if you want to set the maximum number of memory blocks
-// cached by the non-recursive algorithm: Normally this is 16, but can be 
-// higher if you have multiple threads all using boost.regex, or lower 
-// if you don't want boost.regex to cache memory.
-// #define BOOST_REGEX_MAX_CACHE_BLOCKS 16
-
-// define this if you want to be able to access extended capture
-// information in your sub_match's (caution this will slow things
-// down quite a bit).
-// #define BOOST_REGEX_MATCH_EXTRA
-
-// define this if you want to enable support for Unicode via ICU.
-// #define BOOST_HAS_ICU
-
-// define this if you want regex to use __cdecl calling convensions, even when __fastcall is available:
-// #define BOOST_REGEX_NO_FASTCALL

regex/include/boost/regex/v5/basic_regex.hpp

@@ -1,734 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2004 John Maddock
- * Copyright 2011 Garmin Ltd. or its subsidiaries
- *
- * Distributed under the Boost Software License, Version 1.0.
- * (See accompanying file LICENSE_1_0.txt or copy at
- * http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org/ for most recent version.
-  *   FILE         basic_regex.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares template class basic_regex.
-  */
-
-#ifndef BOOST_REGEX_V5_BASIC_REGEX_HPP
-#define BOOST_REGEX_V5_BASIC_REGEX_HPP
-
-#include <vector>
-
-namespace boost{
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable : 4251)
-#if BOOST_REGEX_MSVC < 1700
-#     pragma warning(disable : 4231)
-#endif
-#if BOOST_REGEX_MSVC < 1600
-#pragma warning(disable : 4660)
-#endif
-#if BOOST_REGEX_MSVC < 1910
-#pragma warning(disable:4800)
-#endif
-#endif
-
-namespace BOOST_REGEX_DETAIL_NS{
-
-//
-// forward declaration, we will need this one later:
-//
-template <class charT, class traits>
-class basic_regex_parser;
-
-template <class I>
-void bubble_down_one(I first, I last)
-{
-   if(first != last)
-   {
-      I next = last - 1;
-      while((next != first) && (*next < *(next-1)))
-      {
-         (next-1)->swap(*next);
-         --next;
-      }
-   }
-}
-
-static const int hash_value_mask = 1 << (std::numeric_limits<int>::digits - 1);
-
-template <class Iterator>
-inline int hash_value_from_capture_name(Iterator i, Iterator j)
-{
-   std::size_t r = 0;
-   while (i != j)
-   {
-      r ^= *i + 0x9e3779b9 + (r << 6) + (r >> 2);
-      ++i;
-   }
-   r %= ((std::numeric_limits<int>::max)());
-   return static_cast<int>(r) | hash_value_mask;
-}
-
-class named_subexpressions
-{
-public:
-   struct name
-   {
-      template <class charT>
-      name(const charT* i, const charT* j, int idx)
-         : index(idx) 
-      { 
-         hash = hash_value_from_capture_name(i, j); 
-      }
-      name(int h, int idx)
-         : index(idx), hash(h)
-      { 
-      }
-      int index;
-      int hash;
-      bool operator < (const name& other)const
-      {
-         return hash < other.hash;
-      }
-      bool operator == (const name& other)const
-      {
-         return hash == other.hash; 
-      }
-      void swap(name& other)
-      {
-         std::swap(index, other.index);
-         std::swap(hash, other.hash);
-      }
-   };
-
-   typedef std::vector<name>::const_iterator const_iterator;
-   typedef std::pair<const_iterator, const_iterator> range_type;
-
-   named_subexpressions(){}
-
-   template <class charT>
-   void set_name(const charT* i, const charT* j, int index)
-   {
-      m_sub_names.push_back(name(i, j, index));
-      bubble_down_one(m_sub_names.begin(), m_sub_names.end());
-   }
-   template <class charT>
-   int get_id(const charT* i, const charT* j)const
-   {
-      name t(i, j, 0);
-      typename std::vector<name>::const_iterator pos = std::lower_bound(m_sub_names.begin(), m_sub_names.end(), t);
-      if((pos != m_sub_names.end()) && (*pos == t))
-      {
-         return pos->index;
-      }
-      return -1;
-   }
-   template <class charT>
-   range_type equal_range(const charT* i, const charT* j)const
-   {
-      name t(i, j, 0);
-      return std::equal_range(m_sub_names.begin(), m_sub_names.end(), t);
-   }
-   int get_id(int h)const
-   {
-      name t(h, 0);
-      std::vector<name>::const_iterator pos = std::lower_bound(m_sub_names.begin(), m_sub_names.end(), t);
-      if((pos != m_sub_names.end()) && (*pos == t))
-      {
-         return pos->index;
-      }
-      return -1;
-   }
-   range_type equal_range(int h)const
-   {
-      name t(h, 0);
-      return std::equal_range(m_sub_names.begin(), m_sub_names.end(), t);
-   }
-private:
-   std::vector<name> m_sub_names;
-};
-
-//
-// class regex_data:
-// represents the data we wish to expose to the matching algorithms.
-//
-template <class charT, class traits>
-struct regex_data : public named_subexpressions
-{
-   typedef regex_constants::syntax_option_type   flag_type;
-   typedef std::size_t                           size_type;  
-
-   regex_data(const ::std::shared_ptr<
-      ::boost::regex_traits_wrapper<traits> >& t) 
-      : m_ptraits(t), m_flags(0), m_status(0), m_expression(0), m_expression_len(0),
-         m_mark_count(0), m_first_state(0), m_restart_type(0),
-         m_startmap{ 0 },
-         m_can_be_null(0), m_word_mask(0), m_has_recursions(false), m_disable_match_any(false) {}
-   regex_data() 
-      : m_ptraits(new ::boost::regex_traits_wrapper<traits>()), m_flags(0), m_status(0), m_expression(0), m_expression_len(0), 
-         m_mark_count(0), m_first_state(0), m_restart_type(0), 
-      m_startmap{ 0 },
-         m_can_be_null(0), m_word_mask(0), m_has_recursions(false), m_disable_match_any(false) {}
-
-   ::std::shared_ptr<
-      ::boost::regex_traits_wrapper<traits>
-      >                        m_ptraits;                 // traits class instance
-   flag_type                   m_flags;                   // flags with which we were compiled
-   int                         m_status;                  // error code (0 implies OK).
-   const charT*                m_expression;              // the original expression
-   std::ptrdiff_t              m_expression_len;          // the length of the original expression
-   size_type                   m_mark_count;              // the number of marked sub-expressions
-   BOOST_REGEX_DETAIL_NS::re_syntax_base*  m_first_state;             // the first state of the machine
-   unsigned                    m_restart_type;            // search optimisation type
-   unsigned char               m_startmap[1 << CHAR_BIT]; // which characters can start a match
-   unsigned int                m_can_be_null;             // whether we can match a null string
-   BOOST_REGEX_DETAIL_NS::raw_storage      m_data;                    // the buffer in which our states are constructed
-   typename traits::char_class_type    m_word_mask;       // mask used to determine if a character is a word character
-   std::vector<
-      std::pair<
-      std::size_t, std::size_t> > m_subs;                 // Position of sub-expressions within the *string*.
-   bool                        m_has_recursions;          // whether we have recursive expressions;
-   bool                        m_disable_match_any;       // when set we need to disable the match_any flag as it causes different/buggy behaviour.
-};
-//
-// class basic_regex_implementation
-// pimpl implementation class for basic_regex.
-//
-template <class charT, class traits>
-class basic_regex_implementation
-   : public regex_data<charT, traits>
-{
-public:
-   typedef regex_constants::syntax_option_type   flag_type;
-   typedef std::ptrdiff_t                        difference_type;
-   typedef std::size_t                           size_type; 
-   typedef typename traits::locale_type          locale_type;
-   typedef const charT*                          const_iterator;
-
-   basic_regex_implementation(){}
-   basic_regex_implementation(const ::std::shared_ptr<
-      ::boost::regex_traits_wrapper<traits> >& t)
-      : regex_data<charT, traits>(t) {}
-   void assign(const charT* arg_first,
-                          const charT* arg_last,
-                          flag_type f)
-   {
-      regex_data<charT, traits>* pdat = this;
-      basic_regex_parser<charT, traits> parser(pdat);
-      parser.parse(arg_first, arg_last, f);
-   }
-
-   locale_type  imbue(locale_type l)
-   { 
-      return this->m_ptraits->imbue(l); 
-   }
-   locale_type  getloc()const
-   { 
-      return this->m_ptraits->getloc(); 
-   }
-   std::basic_string<charT>  str()const
-   {
-      std::basic_string<charT> result;
-      if(this->m_status == 0)
-         result = std::basic_string<charT>(this->m_expression, this->m_expression_len);
-      return result;
-   }
-   const_iterator  expression()const
-   {
-      return this->m_expression;
-   }
-   std::pair<const_iterator, const_iterator>  subexpression(std::size_t n)const
-   {
-      const std::pair<std::size_t, std::size_t>& pi = this->m_subs.at(n);
-      std::pair<const_iterator, const_iterator> p(expression() + pi.first, expression() + pi.second);
-      return p;
-   }
-   //
-   // begin, end:
-   const_iterator  begin()const
-   { 
-      return (this->m_status ? 0 : this->m_expression); 
-   }
-   const_iterator  end()const
-   { 
-      return (this->m_status ? 0 : this->m_expression + this->m_expression_len); 
-   }
-   flag_type  flags()const
-   {
-      return this->m_flags;
-   }
-   size_type  size()const
-   {
-      return this->m_expression_len;
-   }
-   int  status()const
-   {
-      return this->m_status;
-   }
-   size_type  mark_count()const
-   {
-      return this->m_mark_count - 1;
-   }
-   const BOOST_REGEX_DETAIL_NS::re_syntax_base* get_first_state()const
-   {
-      return this->m_first_state;
-   }
-   unsigned get_restart_type()const
-   {
-      return this->m_restart_type;
-   }
-   const unsigned char* get_map()const
-   {
-      return this->m_startmap;
-   }
-   const ::boost::regex_traits_wrapper<traits>& get_traits()const
-   {
-      return *(this->m_ptraits);
-   }
-   bool can_be_null()const
-   {
-      return this->m_can_be_null;
-   }
-   const regex_data<charT, traits>& get_data()const
-   {
-      basic_regex_implementation<charT, traits> const* p = this;
-      return *static_cast<const regex_data<charT, traits>*>(p);
-   }
-};
-
-} // namespace BOOST_REGEX_DETAIL_NS
-//
-// class basic_regex:
-// represents the compiled
-// regular expression:
-//
-
-#ifdef BOOST_REGEX_NO_FWD
-template <class charT, class traits = regex_traits<charT> >
-#else
-template <class charT, class traits >
-#endif
-class basic_regex : public regbase
-{
-public:
-   // typedefs:
-   typedef std::size_t                           traits_size_type;
-   typedef typename traits::string_type          traits_string_type;
-   typedef charT                                 char_type;
-   typedef traits                                traits_type;
-
-   typedef charT                                 value_type;
-   typedef charT&                                reference;
-   typedef const charT&                          const_reference;
-   typedef const charT*                          const_iterator;
-   typedef const_iterator                        iterator;
-   typedef std::ptrdiff_t                        difference_type;
-   typedef std::size_t                           size_type;   
-   typedef regex_constants::syntax_option_type   flag_type;
-   // locale_type
-   // placeholder for actual locale type used by the
-   // traits class to localise *this.
-   typedef typename traits::locale_type          locale_type;
-   
-public:
-   explicit basic_regex(){}
-   explicit basic_regex(const charT* p, flag_type f = regex_constants::normal)
-   {
-      assign(p, f);
-   }
-   basic_regex(const charT* p1, const charT* p2, flag_type f = regex_constants::normal)
-   {
-      assign(p1, p2, f);
-   }
-   basic_regex(const charT* p, size_type len, flag_type f)
-   {
-      assign(p, len, f);
-   }
-   basic_regex(const basic_regex& that)
-      : m_pimpl(that.m_pimpl) {}
-   ~basic_regex(){}
-   basic_regex&  operator=(const basic_regex& that)
-   {
-      return assign(that);
-   }
-   basic_regex&  operator=(const charT* ptr)
-   {
-      return assign(ptr);
-   }
-
-   //
-   // assign:
-   basic_regex& assign(const basic_regex& that)
-   { 
-      m_pimpl = that.m_pimpl;
-      return *this; 
-   }
-   basic_regex& assign(const charT* p, flag_type f = regex_constants::normal)
-   {
-      return assign(p, p + traits::length(p), f);
-   }
-   basic_regex& assign(const charT* p, size_type len, flag_type f)
-   {
-      return assign(p, p + len, f);
-   }
-private:
-   basic_regex& do_assign(const charT* p1,
-                          const charT* p2,
-                          flag_type f);
-public:
-   basic_regex& assign(const charT* p1,
-                          const charT* p2,
-                          flag_type f = regex_constants::normal)
-   {
-      return do_assign(p1, p2, f);
-   }
-
-   template <class ST, class SA>
-   unsigned int  set_expression(const std::basic_string<charT, ST, SA>& p, flag_type f = regex_constants::normal)
-   { 
-      return set_expression(p.data(), p.data() + p.size(), f); 
-   }
-
-   template <class ST, class SA>
-   explicit basic_regex(const std::basic_string<charT, ST, SA>& p, flag_type f = regex_constants::normal)
-   { 
-      assign(p, f); 
-   }
-
-   template <class InputIterator>
-   basic_regex(InputIterator arg_first, InputIterator arg_last, flag_type f = regex_constants::normal)
-   {
-      typedef typename traits::string_type seq_type;
-      seq_type a(arg_first, arg_last);
-      if(!a.empty())
-         assign(static_cast<const charT*>(&*a.begin()), static_cast<const charT*>(&*a.begin() + a.size()), f);
-      else
-         assign(static_cast<const charT*>(0), static_cast<const charT*>(0), f);
-   }
-
-   template <class ST, class SA>
-   basic_regex&  operator=(const std::basic_string<charT, ST, SA>& p)
-   {
-      return assign(p.data(), p.data() + p.size(), regex_constants::normal);
-   }
-
-   template <class string_traits, class A>
-   basic_regex&  assign(
-       const std::basic_string<charT, string_traits, A>& s,
-       flag_type f = regex_constants::normal)
-   {
-      return assign(s.data(), s.data() + s.size(), f);
-   }
-
-   template <class InputIterator>
-   basic_regex&  assign(InputIterator arg_first,
-                          InputIterator arg_last,
-                          flag_type f = regex_constants::normal)
-   {
-      typedef typename traits::string_type seq_type;
-      seq_type a(arg_first, arg_last);
-      if(a.size())
-      {
-         const charT* p1 = &*a.begin();
-         const charT* p2 = &*a.begin() + a.size();
-         return assign(p1, p2, f);
-      }
-      return assign(static_cast<const charT*>(0), static_cast<const charT*>(0), f);
-   }
-
-   //
-   // locale:
-   locale_type  imbue(locale_type l);
-   locale_type  getloc()const
-   { 
-      return m_pimpl.get() ? m_pimpl->getloc() : locale_type(); 
-   }
-   //
-   // getflags:
-   // retained for backwards compatibility only, "flags"
-   // is now the preferred name:
-   flag_type  getflags()const
-   { 
-      return flags();
-   }
-   flag_type  flags()const
-   { 
-      return m_pimpl.get() ? m_pimpl->flags() : 0;
-   }
-   //
-   // str:
-   std::basic_string<charT>  str()const
-   {
-      return m_pimpl.get() ? m_pimpl->str() : std::basic_string<charT>();
-   }
-   //
-   // begin, end, subexpression:
-   std::pair<const_iterator, const_iterator>  subexpression(std::size_t n)const
-   {
-#ifdef BOOST_REGEX_STANDALONE
-      if (!m_pimpl.get())
-         throw std::logic_error("Can't access subexpressions in an invalid regex.");
-#else
-      if(!m_pimpl.get())
-         boost::throw_exception(std::logic_error("Can't access subexpressions in an invalid regex."));
-#endif
-      return m_pimpl->subexpression(n);
-   }
-   const_iterator  begin()const
-   { 
-      return (m_pimpl.get() ? m_pimpl->begin() : 0); 
-   }
-   const_iterator  end()const
-   { 
-      return (m_pimpl.get() ? m_pimpl->end() : 0); 
-   }
-   //
-   // swap:
-   void  swap(basic_regex& that)throw()
-   {
-      m_pimpl.swap(that.m_pimpl);
-   }
-   //
-   // size:
-   size_type  size()const
-   { 
-      return (m_pimpl.get() ? m_pimpl->size() : 0); 
-   }
-   //
-   // max_size:
-   size_type  max_size()const
-   { 
-      return UINT_MAX; 
-   }
-   //
-   // empty:
-   bool  empty()const
-   { 
-      return (m_pimpl.get() ? 0 != m_pimpl->status() : true); 
-   }
-
-   size_type  mark_count()const 
-   { 
-      return (m_pimpl.get() ? m_pimpl->mark_count() : 0); 
-   }
-
-   int status()const
-   {
-      return (m_pimpl.get() ? m_pimpl->status() : regex_constants::error_empty);
-   }
-
-   int  compare(const basic_regex& that) const
-   {
-      if(m_pimpl.get() == that.m_pimpl.get())
-         return 0;
-      if(!m_pimpl.get())
-         return -1;
-      if(!that.m_pimpl.get())
-         return 1;
-      if(status() != that.status())
-         return status() - that.status();
-      if(flags() != that.flags())
-         return flags() - that.flags();
-      return str().compare(that.str());
-   }
-   bool  operator==(const basic_regex& e)const
-   { 
-      return compare(e) == 0; 
-   }
-   bool  operator != (const basic_regex& e)const
-   { 
-      return compare(e) != 0; 
-   }
-   bool  operator<(const basic_regex& e)const
-   { 
-      return compare(e) < 0; 
-   }
-   bool  operator>(const basic_regex& e)const
-   { 
-      return compare(e) > 0; 
-   }
-   bool  operator<=(const basic_regex& e)const
-   { 
-      return compare(e) <= 0; 
-   }
-   bool  operator>=(const basic_regex& e)const
-   { 
-      return compare(e) >= 0; 
-   }
-
-   //
-   // The following are deprecated as public interfaces
-   // but are available for compatibility with earlier versions.
-   const charT*  expression()const 
-   { 
-      return (m_pimpl.get() && !m_pimpl->status() ? m_pimpl->expression() : 0); 
-   }
-   unsigned int  set_expression(const charT* p1, const charT* p2, flag_type f = regex_constants::normal)
-   {
-      assign(p1, p2, f | regex_constants::no_except);
-      return status();
-   }
-   unsigned int  set_expression(const charT* p, flag_type f = regex_constants::normal) 
-   { 
-      assign(p, f | regex_constants::no_except); 
-      return status();
-   }
-   unsigned int  error_code()const
-   {
-      return status();
-   }
-   //
-   // private access methods:
-   //
-   const BOOST_REGEX_DETAIL_NS::re_syntax_base* get_first_state()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->get_first_state();
-   }
-   unsigned get_restart_type()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->get_restart_type();
-   }
-   const unsigned char* get_map()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->get_map();
-   }
-   const ::boost::regex_traits_wrapper<traits>& get_traits()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->get_traits();
-   }
-   bool can_be_null()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->can_be_null();
-   }
-   const BOOST_REGEX_DETAIL_NS::regex_data<charT, traits>& get_data()const
-   {
-      BOOST_REGEX_ASSERT(0 != m_pimpl.get());
-      return m_pimpl->get_data();
-   }
-   std::shared_ptr<BOOST_REGEX_DETAIL_NS::named_subexpressions > get_named_subs()const
-   {
-      return m_pimpl;
-   }
-
-private:
-   std::shared_ptr<BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits> > m_pimpl;
-};
-
-//
-// out of line members;
-// these are the only members that mutate the basic_regex object,
-// and are designed to provide the strong exception guarantee
-// (in the event of a throw, the state of the object remains unchanged).
-//
-template <class charT, class traits>
-basic_regex<charT, traits>& basic_regex<charT, traits>::do_assign(const charT* p1,
-                        const charT* p2,
-                        flag_type f)
-{
-   std::shared_ptr<BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits> > temp;
-   if(!m_pimpl.get())
-   {
-      temp = std::shared_ptr<BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits> >(new BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits>());
-   }
-   else
-   {
-      temp = std::shared_ptr<BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits> >(new BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits>(m_pimpl->m_ptraits));
-   }
-   temp->assign(p1, p2, f);
-   temp.swap(m_pimpl);
-   return *this;
-}
-
-template <class charT, class traits>
-typename basic_regex<charT, traits>::locale_type  basic_regex<charT, traits>::imbue(locale_type l)
-{ 
-   std::shared_ptr<BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits> > temp(new BOOST_REGEX_DETAIL_NS::basic_regex_implementation<charT, traits>());
-   locale_type result = temp->imbue(l);
-   temp.swap(m_pimpl);
-   return result;
-}
-
-//
-// non-members:
-//
-template <class charT, class traits>
-void swap(basic_regex<charT, traits>& e1, basic_regex<charT, traits>& e2)
-{
-   e1.swap(e2);
-}
-
-template <class charT, class traits, class traits2>
-std::basic_ostream<charT, traits>& 
-   operator << (std::basic_ostream<charT, traits>& os, 
-                const basic_regex<charT, traits2>& e)
-{
-   return (os << e.str());
-}
-
-//
-// class reg_expression:
-// this is provided for backwards compatibility only,
-// it is deprecated, no not use!
-//
-#ifdef BOOST_REGEX_NO_FWD
-template <class charT, class traits = regex_traits<charT> >
-#else
-template <class charT, class traits >
-#endif
-class reg_expression : public basic_regex<charT, traits>
-{
-public:
-   typedef typename basic_regex<charT, traits>::flag_type flag_type;
-   typedef typename basic_regex<charT, traits>::size_type size_type;
-   explicit reg_expression(){}
-   explicit reg_expression(const charT* p, flag_type f = regex_constants::normal)
-      : basic_regex<charT, traits>(p, f){}
-   reg_expression(const charT* p1, const charT* p2, flag_type f = regex_constants::normal)
-      : basic_regex<charT, traits>(p1, p2, f){}
-   reg_expression(const charT* p, size_type len, flag_type f)
-      : basic_regex<charT, traits>(p, len, f){}
-   reg_expression(const reg_expression& that)
-      : basic_regex<charT, traits>(that) {}
-   ~reg_expression(){}
-   reg_expression&  operator=(const reg_expression& that)
-   {
-      return this->assign(that);
-   }
-
-   template <class ST, class SA>
-   explicit reg_expression(const std::basic_string<charT, ST, SA>& p, flag_type f = regex_constants::normal)
-   : basic_regex<charT, traits>(p, f)
-   { 
-   }
-
-   template <class InputIterator>
-   reg_expression(InputIterator arg_first, InputIterator arg_last, flag_type f = regex_constants::normal)
-   : basic_regex<charT, traits>(arg_first, arg_last, f)
-   {
-   }
-
-   template <class ST, class SA>
-   reg_expression&  operator=(const std::basic_string<charT, ST, SA>& p)
-   {
-      this->assign(p);
-      return *this;
-   }
-
-};
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning (pop)
-#endif
-
-} // namespace boost
-
-#endif

regex/include/boost/regex/v5/c_regex_traits.hpp

@@ -1,474 +0,0 @@
-/*
- *
- * Copyright (c) 2004
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         c_regex_traits.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares regular expression traits class that wraps the global C locale.
-  */
-
-#ifndef BOOST_C_REGEX_TRAITS_HPP_INCLUDED
-#define BOOST_C_REGEX_TRAITS_HPP_INCLUDED
-
-#include <boost/regex/config.hpp>
-#include <boost/regex/v5/regex_workaround.hpp>
-#include <cctype>
-
-namespace boost{
-
-   namespace BOOST_REGEX_DETAIL_NS {
-
-      enum
-      {
-         char_class_space = 1 << 0,
-         char_class_print = 1 << 1,
-         char_class_cntrl = 1 << 2,
-         char_class_upper = 1 << 3,
-         char_class_lower = 1 << 4,
-         char_class_alpha = 1 << 5,
-         char_class_digit = 1 << 6,
-         char_class_punct = 1 << 7,
-         char_class_xdigit = 1 << 8,
-         char_class_alnum = char_class_alpha | char_class_digit,
-         char_class_graph = char_class_alnum | char_class_punct,
-         char_class_blank = 1 << 9,
-         char_class_word = 1 << 10,
-         char_class_unicode = 1 << 11,
-         char_class_horizontal = 1 << 12,
-         char_class_vertical = 1 << 13
-      };
-
-   }
-
-template <class charT>
-struct c_regex_traits;
-
-template<>
-struct c_regex_traits<char>
-{
-   c_regex_traits(){}
-   typedef char char_type;
-   typedef std::size_t size_type;
-   typedef std::string string_type;
-   struct locale_type{};
-   typedef std::uint32_t char_class_type;
-
-   static size_type length(const char_type* p) 
-   { 
-      return (std::strlen)(p); 
-   }
-
-   char translate(char c) const 
-   { 
-      return c; 
-   }
-   char translate_nocase(char c) const 
-   { 
-      return static_cast<char>((std::tolower)(static_cast<unsigned char>(c))); 
-   }
-
-   static string_type  transform(const char* p1, const char* p2);
-   static string_type  transform_primary(const char* p1, const char* p2);
-
-   static char_class_type  lookup_classname(const char* p1, const char* p2);
-   static string_type  lookup_collatename(const char* p1, const char* p2);
-
-   static bool  isctype(char, char_class_type);
-   static int  value(char, int);
-
-   locale_type imbue(locale_type l)
-   { return l; }
-   locale_type getloc()const
-   { return locale_type(); }
-
-private:
-   // this type is not copyable:
-   c_regex_traits(const c_regex_traits&);
-   c_regex_traits& operator=(const c_regex_traits&);
-};
-
-#ifndef BOOST_NO_WREGEX
-template<>
-struct c_regex_traits<wchar_t>
-{
-   c_regex_traits(){}
-   typedef wchar_t char_type;
-   typedef std::size_t size_type;
-   typedef std::wstring string_type;
-   struct locale_type{};
-   typedef std::uint32_t char_class_type;
-
-   static size_type length(const char_type* p) 
-   { 
-      return (std::wcslen)(p); 
-   }
-
-   wchar_t translate(wchar_t c) const 
-   { 
-      return c; 
-   }
-   wchar_t translate_nocase(wchar_t c) const 
-   { 
-      return (std::towlower)(c); 
-   }
-
-   static string_type  transform(const wchar_t* p1, const wchar_t* p2);
-   static string_type  transform_primary(const wchar_t* p1, const wchar_t* p2);
-
-   static char_class_type  lookup_classname(const wchar_t* p1, const wchar_t* p2);
-   static string_type  lookup_collatename(const wchar_t* p1, const wchar_t* p2);
-
-   static bool  isctype(wchar_t, char_class_type);
-   static int  value(wchar_t, int);
-
-   locale_type imbue(locale_type l)
-   { return l; }
-   locale_type getloc()const
-   { return locale_type(); }
-
-private:
-   // this type is not copyable:
-   c_regex_traits(const c_regex_traits&);
-   c_regex_traits& operator=(const c_regex_traits&);
-};
-
-#endif // BOOST_NO_WREGEX
-
-inline c_regex_traits<char>::string_type  c_regex_traits<char>::transform(const char* p1, const char* p2)
-{
-   std::string result(10, ' ');
-   std::size_t s = result.size();
-   std::size_t r;
-   std::string src(p1, p2);
-   while (s < (r = std::strxfrm(&*result.begin(), src.c_str(), s)))
-   {
-#if defined(_CPPLIB_VER)
-      //
-      // A bug in VC11 and 12 causes the program to hang if we pass a null-string
-      // to std::strxfrm, but only for certain locales :-(
-      // Probably effects Intel and Clang or any compiler using the VC std library (Dinkumware).
-      //
-      if (r == INT_MAX)
-      {
-         result.erase();
-         result.insert(result.begin(), static_cast<char>(0));
-         return result;
-      }
-#endif
-      result.append(r - s + 3, ' ');
-      s = result.size();
-   }
-   result.erase(r);
-   return result;
-}
-
-inline c_regex_traits<char>::string_type  c_regex_traits<char>::transform_primary(const char* p1, const char* p2)
-{
-   static char s_delim;
-   static const int s_collate_type = ::boost::BOOST_REGEX_DETAIL_NS::find_sort_syntax(static_cast<c_regex_traits<char>*>(0), &s_delim);
-   std::string result;
-   //
-   // What we do here depends upon the format of the sort key returned by
-   // sort key returned by this->transform:
-   //
-   switch (s_collate_type)
-   {
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_C:
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_unknown:
-      // the best we can do is translate to lower case, then get a regular sort key:
-   {
-      result.assign(p1, p2);
-      for (std::string::size_type i = 0; i < result.size(); ++i)
-         result[i] = static_cast<char>((std::tolower)(static_cast<unsigned char>(result[i])));
-      result = transform(&*result.begin(), &*result.begin() + result.size());
-      break;
-   }
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_fixed:
-   {
-      // get a regular sort key, and then truncate it:
-      result = transform(p1, p2);
-      result.erase(s_delim);
-      break;
-   }
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_delim:
-      // get a regular sort key, and then truncate everything after the delim:
-      result = transform(p1, p2);
-      if ((!result.empty()) && (result[0] == s_delim))
-         break;
-      std::size_t i;
-      for (i = 0; i < result.size(); ++i)
-      {
-         if (result[i] == s_delim)
-            break;
-      }
-      result.erase(i);
-      break;
-   }
-   if (result.empty())
-      result = std::string(1, char(0));
-   return result;
-}
-
-inline c_regex_traits<char>::char_class_type  c_regex_traits<char>::lookup_classname(const char* p1, const char* p2)
-{
-   using namespace BOOST_REGEX_DETAIL_NS;
-   static const char_class_type masks[] =
-   {
-      0,
-      char_class_alnum,
-      char_class_alpha,
-      char_class_blank,
-      char_class_cntrl,
-      char_class_digit,
-      char_class_digit,
-      char_class_graph,
-      char_class_horizontal,
-      char_class_lower,
-      char_class_lower,
-      char_class_print,
-      char_class_punct,
-      char_class_space,
-      char_class_space,
-      char_class_upper,
-      char_class_unicode,
-      char_class_upper,
-      char_class_vertical,
-      char_class_alnum | char_class_word,
-      char_class_alnum | char_class_word,
-      char_class_xdigit,
-   };
-
-   int idx = ::boost::BOOST_REGEX_DETAIL_NS::get_default_class_id(p1, p2);
-   if (idx < 0)
-   {
-      std::string s(p1, p2);
-      for (std::string::size_type i = 0; i < s.size(); ++i)
-         s[i] = static_cast<char>((std::tolower)(static_cast<unsigned char>(s[i])));
-      idx = ::boost::BOOST_REGEX_DETAIL_NS::get_default_class_id(&*s.begin(), &*s.begin() + s.size());
-   }
-   BOOST_REGEX_ASSERT(std::size_t(idx) + 1u < sizeof(masks) / sizeof(masks[0]));
-   return masks[idx + 1];
-}
-
-inline bool  c_regex_traits<char>::isctype(char c, char_class_type mask)
-{
-   using namespace BOOST_REGEX_DETAIL_NS;
-   return
-      ((mask & char_class_space) && (std::isspace)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_print) && (std::isprint)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_cntrl) && (std::iscntrl)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_upper) && (std::isupper)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_lower) && (std::islower)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_alpha) && (std::isalpha)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_digit) && (std::isdigit)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_punct) && (std::ispunct)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_xdigit) && (std::isxdigit)(static_cast<unsigned char>(c)))
-      || ((mask & char_class_blank) && (std::isspace)(static_cast<unsigned char>(c)) && !::boost::BOOST_REGEX_DETAIL_NS::is_separator(c))
-      || ((mask & char_class_word) && (c == '_'))
-      || ((mask & char_class_vertical) && (::boost::BOOST_REGEX_DETAIL_NS::is_separator(c) || (c == '\v')))
-      || ((mask & char_class_horizontal) && (std::isspace)(static_cast<unsigned char>(c)) && !::boost::BOOST_REGEX_DETAIL_NS::is_separator(c) && (c != '\v'));
-}
-
-inline c_regex_traits<char>::string_type  c_regex_traits<char>::lookup_collatename(const char* p1, const char* p2)
-{
-   std::string s(p1, p2);
-   s = ::boost::BOOST_REGEX_DETAIL_NS::lookup_default_collate_name(s);
-   if (s.empty() && (p2 - p1 == 1))
-      s.append(1, *p1);
-   return s;
-}
-
-inline int  c_regex_traits<char>::value(char c, int radix)
-{
-   char b[2] = { c, '\0', };
-   char* ep;
-   int result = std::strtol(b, &ep, radix);
-   if (ep == b)
-      return -1;
-   return result;
-}
-
-#ifndef BOOST_NO_WREGEX
-
-inline c_regex_traits<wchar_t>::string_type  c_regex_traits<wchar_t>::transform(const wchar_t* p1, const wchar_t* p2)
-{
-   std::size_t r;
-   std::size_t s = 10;
-   std::wstring src(p1, p2);
-   std::wstring result(s, L' ');
-   while (s < (r = std::wcsxfrm(&*result.begin(), src.c_str(), s)))
-   {
-#if defined(_CPPLIB_VER)
-      //
-      // A bug in VC11 and 12 causes the program to hang if we pass a null-string
-      // to std::strxfrm, but only for certain locales :-(
-      // Probably effects Intel and Clang or any compiler using the VC std library (Dinkumware).
-      //
-      if (r == INT_MAX)
-      {
-         result.erase();
-         result.insert(result.begin(), static_cast<wchar_t>(0));
-         return result;
-      }
-#endif
-      result.append(r - s + 3, L' ');
-      s = result.size();
-   }
-   result.erase(r);
-   return result;
-}
-
-inline c_regex_traits<wchar_t>::string_type  c_regex_traits<wchar_t>::transform_primary(const wchar_t* p1, const wchar_t* p2)
-{
-   static wchar_t s_delim;
-   static const int s_collate_type = ::boost::BOOST_REGEX_DETAIL_NS::find_sort_syntax(static_cast<const c_regex_traits<wchar_t>*>(0), &s_delim);
-   std::wstring result;
-   //
-   // What we do here depends upon the format of the sort key returned by
-   // sort key returned by this->transform:
-   //
-   switch (s_collate_type)
-   {
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_C:
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_unknown:
-      // the best we can do is translate to lower case, then get a regular sort key:
-   {
-      result.assign(p1, p2);
-      for (std::wstring::size_type i = 0; i < result.size(); ++i)
-         result[i] = (std::towlower)(result[i]);
-      result = c_regex_traits<wchar_t>::transform(&*result.begin(), &*result.begin() + result.size());
-      break;
-   }
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_fixed:
-   {
-      // get a regular sort key, and then truncate it:
-      result = c_regex_traits<wchar_t>::transform(&*result.begin(), &*result.begin() + result.size());
-      result.erase(s_delim);
-      break;
-   }
-   case ::boost::BOOST_REGEX_DETAIL_NS::sort_delim:
-      // get a regular sort key, and then truncate everything after the delim:
-      result = c_regex_traits<wchar_t>::transform(&*result.begin(), &*result.begin() + result.size());
-      if ((!result.empty()) && (result[0] == s_delim))
-         break;
-      std::size_t i;
-      for (i = 0; i < result.size(); ++i)
-      {
-         if (result[i] == s_delim)
-            break;
-      }
-      result.erase(i);
-      break;
-   }
-   if (result.empty())
-      result = std::wstring(1, char(0));
-   return result;
-}
-
-inline c_regex_traits<wchar_t>::char_class_type  c_regex_traits<wchar_t>::lookup_classname(const wchar_t* p1, const wchar_t* p2)
-{
-   using namespace BOOST_REGEX_DETAIL_NS;
-   static const char_class_type masks[] =
-   {
-      0,
-      char_class_alnum,
-      char_class_alpha,
-      char_class_blank,
-      char_class_cntrl,
-      char_class_digit,
-      char_class_digit,
-      char_class_graph,
-      char_class_horizontal,
-      char_class_lower,
-      char_class_lower,
-      char_class_print,
-      char_class_punct,
-      char_class_space,
-      char_class_space,
-      char_class_upper,
-      char_class_unicode,
-      char_class_upper,
-      char_class_vertical,
-      char_class_alnum | char_class_word,
-      char_class_alnum | char_class_word,
-      char_class_xdigit,
-   };
-
-   int idx = ::boost::BOOST_REGEX_DETAIL_NS::get_default_class_id(p1, p2);
-   if (idx < 0)
-   {
-      std::wstring s(p1, p2);
-      for (std::wstring::size_type i = 0; i < s.size(); ++i)
-         s[i] = (std::towlower)(s[i]);
-      idx = ::boost::BOOST_REGEX_DETAIL_NS::get_default_class_id(&*s.begin(), &*s.begin() + s.size());
-   }
-   BOOST_REGEX_ASSERT(idx + 1 < static_cast<int>(sizeof(masks) / sizeof(masks[0])));
-   return masks[idx + 1];
-}
-
-inline bool  c_regex_traits<wchar_t>::isctype(wchar_t c, char_class_type mask)
-{
-   using namespace BOOST_REGEX_DETAIL_NS;
-   return
-      ((mask & char_class_space) && (std::iswspace)(c))
-      || ((mask & char_class_print) && (std::iswprint)(c))
-      || ((mask & char_class_cntrl) && (std::iswcntrl)(c))
-      || ((mask & char_class_upper) && (std::iswupper)(c))
-      || ((mask & char_class_lower) && (std::iswlower)(c))
-      || ((mask & char_class_alpha) && (std::iswalpha)(c))
-      || ((mask & char_class_digit) && (std::iswdigit)(c))
-      || ((mask & char_class_punct) && (std::iswpunct)(c))
-      || ((mask & char_class_xdigit) && (std::iswxdigit)(c))
-      || ((mask & char_class_blank) && (std::iswspace)(c) && !::boost::BOOST_REGEX_DETAIL_NS::is_separator(c))
-      || ((mask & char_class_word) && (c == '_'))
-      || ((mask & char_class_unicode) && (c & ~static_cast<wchar_t>(0xff)))
-      || ((mask & char_class_vertical) && (::boost::BOOST_REGEX_DETAIL_NS::is_separator(c) || (c == L'\v')))
-      || ((mask & char_class_horizontal) && (std::iswspace)(c) && !::boost::BOOST_REGEX_DETAIL_NS::is_separator(c) && (c != L'\v'));
-}
-
-inline c_regex_traits<wchar_t>::string_type  c_regex_traits<wchar_t>::lookup_collatename(const wchar_t* p1, const wchar_t* p2)
-{
-   std::string name;
-   // Usual msvc warning suppression does not work here with std::string template constructor.... use a workaround instead:
-   for (const wchar_t* pos = p1; pos != p2; ++pos)
-      name.push_back((char)*pos);
-   name = ::boost::BOOST_REGEX_DETAIL_NS::lookup_default_collate_name(name);
-   if (!name.empty())
-      return string_type(name.begin(), name.end());
-   if (p2 - p1 == 1)
-      return string_type(1, *p1);
-   return string_type();
-}
-
-inline int  c_regex_traits<wchar_t>::value(wchar_t c, int radix)
-{
-#ifdef BOOST_BORLANDC
-   // workaround for broken wcstol:
-   if ((std::iswxdigit)(c) == 0)
-      return -1;
-#endif
-   wchar_t b[2] = { c, '\0', };
-   wchar_t* ep;
-   int result = std::wcstol(b, &ep, radix);
-   if (ep == b)
-      return -1;
-   return result;
-}
-
-#endif
-
-}
-
-#endif
-
-
-

regex/include/boost/regex/v5/char_regex_traits.hpp

@@ -1,59 +0,0 @@
-/*
- *
- * Copyright (c) 2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the
- * Boost Software License, Version 1.0. (See accompanying file
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         char_regex_traits.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares deprecated traits classes char_regex_traits<>.
-  */
-
-
-#ifndef BOOST_REGEX_V5_CHAR_REGEX_TRAITS_HPP
-#define BOOST_REGEX_V5_CHAR_REGEX_TRAITS_HPP
-
-namespace boost{
-
-namespace deprecated{
-//
-// class char_regex_traits_i
-// provides case insensitive traits classes (deprecated):
-template <class charT>
-class char_regex_traits_i : public regex_traits<charT> {};
-
-template<>
-class char_regex_traits_i<char> : public regex_traits<char>
-{
-public:
-   typedef char char_type;
-   typedef unsigned char uchar_type;
-   typedef unsigned int size_type;
-   typedef regex_traits<char> base_type;
-
-};
-
-#ifndef BOOST_NO_WREGEX
-template<>
-class char_regex_traits_i<wchar_t> : public regex_traits<wchar_t>
-{
-public:
-   typedef wchar_t char_type;
-   typedef unsigned short uchar_type;
-   typedef unsigned int size_type;
-   typedef regex_traits<wchar_t> base_type;
-
-};
-#endif
-} // namespace deprecated
-} // namespace boost
-
-#endif // include
-

regex/include/boost/regex/v5/cregex.hpp

@@ -1,195 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the
- * Boost Software License, Version 1.0. (See accompanying file
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         cregex.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares POSIX API functions
-  *                + boost::RegEx high level wrapper.
-  */
-
-#ifndef BOOST_RE_CREGEX_HPP_INCLUDED
-#define BOOST_RE_CREGEX_HPP_INCLUDED
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#include <boost/regex/config.hpp>
-#endif
-#include <boost/regex/v5/match_flags.hpp>
-#include <boost/regex/v5/error_type.hpp>
-
-#ifndef BOOST_REGEX_STANDALONE
-#if !defined(BOOST_REGEX_NO_LIB) && !defined(BOOST_REGEX_SOURCE) && !defined(BOOST_ALL_NO_LIB) && defined(__cplusplus)
-#  define BOOST_LIB_NAME boost_regex
-#  if defined(BOOST_REGEX_DYN_LINK) || defined(BOOST_ALL_DYN_LINK)
-#     define BOOST_DYN_LINK
-#  endif
-#  ifdef BOOST_REGEX_DIAG
-#     define BOOST_LIB_DIAGNOSTIC
-#  endif
-#  include <boost/config/auto_link.hpp>
-#endif
-#endif
-
-#ifdef __cplusplus
-#include <cstddef>
-#else
-#include <stddef.h>
-#endif
-
-/* include these defs only for POSIX compatablity */
-#ifdef __cplusplus
-namespace boost{
-extern "C" {
-#endif
-
-#if defined(__cplusplus)
-typedef std::ptrdiff_t regoff_t;
-typedef std::size_t regsize_t;
-#else
-typedef ptrdiff_t regoff_t;
-typedef size_t regsize_t;
-#endif
-
-typedef struct
-{
-   unsigned int re_magic;
-#ifdef __cplusplus
-   std::size_t  re_nsub;      /* number of parenthesized subexpressions */
-#else
-   size_t re_nsub; 
-#endif
-   const char*  re_endp;       /* end pointer for REG_PEND */
-   void* guts;                /* none of your business :-) */
-   match_flag_type eflags;        /* none of your business :-) */
-} regex_tA;
-
-#ifndef BOOST_NO_WREGEX
-typedef struct
-{
-   unsigned int re_magic;
-#ifdef __cplusplus
-   std::size_t  re_nsub;         /* number of parenthesized subexpressions */
-#else
-   size_t re_nsub;
-#endif
-   const wchar_t* re_endp;       /* end pointer for REG_PEND */
-   void* guts;                   /* none of your business :-) */
-   match_flag_type eflags;           /* none of your business :-) */
-} regex_tW;
-#endif
-
-typedef struct
-{
-   regoff_t rm_so;      /* start of match */
-   regoff_t rm_eo;      /* end of match */
-} regmatch_t;
-
-/* regcomp() flags */
-typedef enum{
-   REG_BASIC = 0000,
-   REG_EXTENDED = 0001,
-   REG_ICASE = 0002,
-   REG_NOSUB = 0004,
-   REG_NEWLINE = 0010,
-   REG_NOSPEC = 0020,
-   REG_PEND = 0040,
-   REG_DUMP = 0200,
-   REG_NOCOLLATE = 0400,
-   REG_ESCAPE_IN_LISTS = 01000,
-   REG_NEWLINE_ALT = 02000,
-   REG_PERLEX = 04000,
-
-   REG_PERL = REG_EXTENDED | REG_NOCOLLATE | REG_ESCAPE_IN_LISTS | REG_PERLEX,
-   REG_AWK = REG_EXTENDED | REG_ESCAPE_IN_LISTS,
-   REG_GREP = REG_BASIC | REG_NEWLINE_ALT,
-   REG_EGREP = REG_EXTENDED | REG_NEWLINE_ALT,
-
-   REG_ASSERT = 15,
-   REG_INVARG = 16,
-   REG_ATOI = 255,   /* convert name to number (!) */
-   REG_ITOA = 0400   /* convert number to name (!) */
-} reg_comp_flags;
-
-/* regexec() flags */
-typedef enum{
-   REG_NOTBOL =    00001,
-   REG_NOTEOL =    00002,
-   REG_STARTEND =  00004
-} reg_exec_flags;
-
-/*
- * POSIX error codes:
- */
-typedef unsigned reg_error_t;
-typedef reg_error_t reg_errcode_t;  /* backwards compatibility */
-
-static const reg_error_t REG_NOERROR = 0;   /* Success.  */
-static const reg_error_t REG_NOMATCH = 1;   /* Didn't find a match (for regexec).  */
-
-  /* POSIX regcomp return error codes.  (In the order listed in the
-     standard.)  */
-static const reg_error_t REG_BADPAT = 2;    /* Invalid pattern.  */
-static const reg_error_t REG_ECOLLATE = 3;  /* Undefined collating element.  */
-static const reg_error_t REG_ECTYPE = 4;    /* Invalid character class name.  */
-static const reg_error_t REG_EESCAPE = 5;   /* Trailing backslash.  */
-static const reg_error_t REG_ESUBREG = 6;   /* Invalid back reference.  */
-static const reg_error_t REG_EBRACK = 7;    /* Unmatched left bracket.  */
-static const reg_error_t REG_EPAREN = 8;    /* Parenthesis imbalance.  */
-static const reg_error_t REG_EBRACE = 9;    /* Unmatched \{.  */
-static const reg_error_t REG_BADBR = 10;    /* Invalid contents of \{\}.  */
-static const reg_error_t REG_ERANGE = 11;   /* Invalid range end.  */
-static const reg_error_t REG_ESPACE = 12;   /* Ran out of memory.  */
-static const reg_error_t REG_BADRPT = 13;   /* No preceding re for repetition op.  */
-static const reg_error_t REG_EEND = 14;     /* unexpected end of expression */
-static const reg_error_t REG_ESIZE = 15;    /* expression too big */
-static const reg_error_t REG_ERPAREN = 8;   /* = REG_EPAREN : unmatched right parenthesis */
-static const reg_error_t REG_EMPTY = 17;    /* empty expression */
-static const reg_error_t REG_E_MEMORY = 15; /* = REG_ESIZE : out of memory */
-static const reg_error_t REG_ECOMPLEXITY = 18; /* complexity too high */
-static const reg_error_t REG_ESTACK = 19;   /* out of stack space */
-static const reg_error_t REG_E_PERL = 20;   /* Perl (?...) error */
-static const reg_error_t REG_E_UNKNOWN = 21; /* unknown error */
-static const reg_error_t REG_ENOSYS = 21;   /* = REG_E_UNKNOWN : Reserved. */
-
-BOOST_REGEX_DECL int BOOST_REGEX_CCALL regcompA(regex_tA*, const char*, int);
-BOOST_REGEX_DECL regsize_t BOOST_REGEX_CCALL regerrorA(int, const regex_tA*, char*, regsize_t);
-BOOST_REGEX_DECL int BOOST_REGEX_CCALL regexecA(const regex_tA*, const char*, regsize_t, regmatch_t*, int);
-BOOST_REGEX_DECL void BOOST_REGEX_CCALL regfreeA(regex_tA*);
-
-#ifndef BOOST_NO_WREGEX
-BOOST_REGEX_DECL int BOOST_REGEX_CCALL regcompW(regex_tW*, const wchar_t*, int);
-BOOST_REGEX_DECL regsize_t BOOST_REGEX_CCALL regerrorW(int, const regex_tW*, wchar_t*, regsize_t);
-BOOST_REGEX_DECL int BOOST_REGEX_CCALL regexecW(const regex_tW*, const wchar_t*, regsize_t, regmatch_t*, int);
-BOOST_REGEX_DECL void BOOST_REGEX_CCALL regfreeW(regex_tW*);
-#endif
-
-#ifdef UNICODE
-#define regcomp regcompW
-#define regerror regerrorW
-#define regexec regexecW
-#define regfree regfreeW
-#define regex_t regex_tW
-#else
-#define regcomp regcompA
-#define regerror regerrorA
-#define regexec regexecA
-#define regfree regfreeA
-#define regex_t regex_tA
-#endif
-
-#ifdef __cplusplus
-} /* extern "C" */
-} /* namespace */
-#endif
-
-#endif /* include guard */
-

regex/include/boost/regex/v5/error_type.hpp

@@ -1,59 +0,0 @@
-/*
- *
- * Copyright (c) 2003-2005
- * John Maddock
- *
- * Use, modification and distribution are subject to the
- * Boost Software License, Version 1.0. (See accompanying file
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         error_type.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares regular expression error type enumerator.
-  */
-
-#ifndef BOOST_REGEX_ERROR_TYPE_HPP
-#define BOOST_REGEX_ERROR_TYPE_HPP
-
-#ifdef __cplusplus
-namespace boost{
-#endif
-
-#ifdef __cplusplus
-namespace regex_constants{
-
-enum error_type{
-
-   error_ok = 0,         /* not used */
-   error_no_match = 1,   /* not used */
-   error_bad_pattern = 2,
-   error_collate = 3,
-   error_ctype = 4,
-   error_escape = 5,
-   error_backref = 6,
-   error_brack = 7,
-   error_paren = 8,
-   error_brace = 9,
-   error_badbrace = 10,
-   error_range = 11,
-   error_space = 12,
-   error_badrepeat = 13,
-   error_end = 14,    /* not used */
-   error_size = 15,
-   error_right_paren = 16,  /* not used */
-   error_empty = 17,
-   error_complexity = 18,
-   error_stack = 19,
-   error_perl_extension = 20,
-   error_unknown = 21
-};
-
-}
-}
-#endif /* __cplusplus */
-
-#endif

regex/include/boost/regex/v5/iterator_category.hpp

@@ -1,84 +0,0 @@
-/*
- *
- * Copyright (c) 2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         regex_match.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Iterator traits for selecting an iterator type as
-  *                an integral constant expression.
-  */
-
-
-#ifndef BOOST_REGEX_ITERATOR_CATEGORY_HPP
-#define BOOST_REGEX_ITERATOR_CATEGORY_HPP
-
-#include <iterator>
-#include <type_traits>
-
-namespace boost{
-namespace detail{
-
-template <class I>
-struct is_random_imp
-{
-private:
-   typedef typename std::iterator_traits<I>::iterator_category cat;
-public:
-   static const bool value = (std::is_convertible<cat*, std::random_access_iterator_tag*>::value);
-};
-
-template <class I>
-struct is_random_pointer_imp
-{
-   static const bool value = true;
-};
-
-template <bool is_pointer_type>
-struct is_random_imp_selector
-{
-   template <class I>
-   struct rebind
-   {
-      typedef is_random_imp<I> type;
-   };
-};
-
-template <>
-struct is_random_imp_selector<true>
-{
-   template <class I>
-   struct rebind
-   {
-      typedef is_random_pointer_imp<I> type;
-   };
-};
-
-}
-
-template <class I>
-struct is_random_access_iterator
-{
-private:
-   typedef detail::is_random_imp_selector< std::is_pointer<I>::value> selector;
-   typedef typename selector::template rebind<I> bound_type;
-   typedef typename bound_type::type answer;
-public:
-   static const bool value = answer::value;
-};
-
-template <class I>
-const bool is_random_access_iterator<I>::value;
-
-}
-
-#endif
-

regex/include/boost/regex/v5/iterator_traits.hpp

@@ -1,32 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         iterator_traits.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares iterator traits workarounds.
-  */
-
-#ifndef BOOST_REGEX_V5_ITERATOR_TRAITS_HPP
-#define BOOST_REGEX_V5_ITERATOR_TRAITS_HPP
-
-namespace boost{
-namespace BOOST_REGEX_DETAIL_NS{
-
-template <class T>
-struct regex_iterator_traits : public std::iterator_traits<T> {};
-
-} // namespace BOOST_REGEX_DETAIL_NS
-} // namespace boost
-
-#endif
-

regex/include/boost/regex/v5/match_flags.hpp

@@ -1,156 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         match_flags.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares match_flags type.
-  */
-
-#ifndef BOOST_REGEX_V5_MATCH_FLAGS
-#define BOOST_REGEX_V5_MATCH_FLAGS
-
-#ifdef __cplusplus
-#  include <cstdint>
-#endif
-
-#ifdef __cplusplus
-namespace boost{
-   namespace regex_constants{
-#endif
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#if BOOST_REGEX_MSVC >= 1800
-#pragma warning(disable : 26812)
-#endif
-#endif
-
-typedef enum _match_flags
-{
-   match_default = 0,
-   match_not_bol = 1,                                /* first is not start of line */
-   match_not_eol = match_not_bol << 1,               /* last is not end of line */
-   match_not_bob = match_not_eol << 1,               /* first is not start of buffer */
-   match_not_eob = match_not_bob << 1,               /* last is not end of buffer */
-   match_not_bow = match_not_eob << 1,               /* first is not start of word */
-   match_not_eow = match_not_bow << 1,               /* last is not end of word */
-   match_not_dot_newline = match_not_eow << 1,       /* \n is not matched by '.' */
-   match_not_dot_null = match_not_dot_newline << 1,  /* '\0' is not matched by '.' */
-   match_prev_avail = match_not_dot_null << 1,       /* *--first is a valid expression */
-   match_init = match_prev_avail << 1,               /* internal use */
-   match_any = match_init << 1,                      /* don't care what we match */
-   match_not_null = match_any << 1,                  /* string can't be null */
-   match_continuous = match_not_null << 1,           /* each grep match must continue from */
-                                                     /* uninterrupted from the previous one */
-   match_partial = match_continuous << 1,            /* find partial matches */
-   
-   match_stop = match_partial << 1,                  /* stop after first match (grep) V3 only */
-   match_not_initial_null = match_stop,              /* don't match initial null, V4 only */
-   match_all = match_stop << 1,                      /* must find the whole of input even if match_any is set */
-   match_perl = match_all << 1,                      /* Use perl matching rules */
-   match_posix = match_perl << 1,                    /* Use POSIX matching rules */
-   match_nosubs = match_posix << 1,                  /* don't trap marked subs */
-   match_extra = match_nosubs << 1,                  /* include full capture information for repeated captures */
-   match_single_line = match_extra << 1,             /* treat text as single line and ignore any \n's when matching ^ and $. */
-   match_unused1 = match_single_line << 1,           /* unused */
-   match_unused2 = match_unused1 << 1,               /* unused */
-   match_unused3 = match_unused2 << 1,               /* unused */
-   match_max = match_unused3,
-
-   format_perl = 0,                                  /* perl style replacement */
-   format_default = 0,                               /* ditto. */
-   format_sed = match_max << 1,                      /* sed style replacement. */
-   format_all = format_sed << 1,                     /* enable all extensions to syntax. */
-   format_no_copy = format_all << 1,                 /* don't copy non-matching segments. */
-   format_first_only = format_no_copy << 1,          /* Only replace first occurrence. */
-   format_is_if = format_first_only << 1,            /* internal use only. */
-   format_literal = format_is_if << 1,               /* treat string as a literal */
-
-   match_not_any = match_not_bol | match_not_eol | match_not_bob 
-      | match_not_eob | match_not_bow | match_not_eow | match_not_dot_newline 
-      | match_not_dot_null | match_prev_avail | match_init | match_not_null
-      | match_continuous | match_partial | match_stop | match_not_initial_null 
-      | match_stop | match_all | match_perl | match_posix | match_nosubs
-      | match_extra | match_single_line | match_unused1 | match_unused2 
-      | match_unused3 | match_max | format_perl | format_default | format_sed
-      | format_all | format_no_copy | format_first_only | format_is_if
-      | format_literal
-
-
-} match_flags;
-
-typedef match_flags match_flag_type;
-
-#ifdef __cplusplus
-inline match_flags operator&(match_flags m1, match_flags m2)
-{ return static_cast<match_flags>(static_cast<std::int32_t>(m1) & static_cast<std::int32_t>(m2)); }
-inline match_flags operator|(match_flags m1, match_flags m2)
-{ return static_cast<match_flags>(static_cast<std::int32_t>(m1) | static_cast<std::int32_t>(m2)); }
-inline match_flags operator^(match_flags m1, match_flags m2)
-{ return static_cast<match_flags>(static_cast<std::int32_t>(m1) ^ static_cast<std::int32_t>(m2)); }
-inline match_flags operator~(match_flags m1)
-{ return static_cast<match_flags>(~static_cast<std::int32_t>(m1)); }
-inline match_flags& operator&=(match_flags& m1, match_flags m2)
-{ m1 = m1&m2; return m1; }
-inline match_flags& operator|=(match_flags& m1, match_flags m2)
-{ m1 = m1|m2; return m1; }
-inline match_flags& operator^=(match_flags& m1, match_flags m2)
-{ m1 = m1^m2; return m1; }
-#endif
-
-#ifdef __cplusplus
-} /* namespace regex_constants */
-/*
- * import names into boost for backwards compatibility:
- */
-using regex_constants::match_flag_type;
-using regex_constants::match_default;
-using regex_constants::match_not_bol;
-using regex_constants::match_not_eol;
-using regex_constants::match_not_bob;
-using regex_constants::match_not_eob;
-using regex_constants::match_not_bow;
-using regex_constants::match_not_eow;
-using regex_constants::match_not_dot_newline;
-using regex_constants::match_not_dot_null;
-using regex_constants::match_prev_avail;
-/* using regex_constants::match_init; */
-using regex_constants::match_any;
-using regex_constants::match_not_null;
-using regex_constants::match_continuous;
-using regex_constants::match_partial;
-/*using regex_constants::match_stop; */
-using regex_constants::match_all;
-using regex_constants::match_perl;
-using regex_constants::match_posix;
-using regex_constants::match_nosubs;
-using regex_constants::match_extra;
-using regex_constants::match_single_line;
-/*using regex_constants::match_max; */
-using regex_constants::format_all;
-using regex_constants::format_sed;
-using regex_constants::format_perl;
-using regex_constants::format_default;
-using regex_constants::format_no_copy;
-using regex_constants::format_first_only;
-/*using regex_constants::format_is_if;*/
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-
-
-} /* namespace boost */
-#endif /* __cplusplus */
-#endif /* include guard */
-

regex/include/boost/regex/v5/match_results.hpp

@@ -1,667 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2009
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         match_results.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares template class match_results.
-  */
-
-#ifndef BOOST_REGEX_V5_MATCH_RESULTS_HPP
-#define BOOST_REGEX_V5_MATCH_RESULTS_HPP
-
-namespace boost{
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable : 4251 4459)
-#if BOOST_REGEX_MSVC < 1700
-#     pragma warning(disable : 4231)
-#endif
-#  if BOOST_REGEX_MSVC < 1600
-#     pragma warning(disable : 4660)
-#  endif
-#endif
-
-namespace BOOST_REGEX_DETAIL_NS{
-
-class named_subexpressions;
-
-}
-
-template <class BidiIterator, class Allocator>
-class match_results
-{ 
-private:
-   typedef          std::vector<sub_match<BidiIterator>, Allocator> vector_type;
-public: 
-   typedef          sub_match<BidiIterator>                         value_type;
-   typedef typename std::allocator_traits<Allocator>::value_type const &    const_reference;
-   typedef          const_reference                                         reference;
-   typedef typename vector_type::const_iterator                             const_iterator;
-   typedef          const_iterator                                          iterator;
-   typedef typename std::iterator_traits<
-                                    BidiIterator>::difference_type          difference_type;
-   typedef typename std::allocator_traits<Allocator>::size_type             size_type;
-   typedef          Allocator                                               allocator_type;
-   typedef typename std::iterator_traits<
-                                    BidiIterator>::value_type               char_type;
-   typedef          std::basic_string<char_type>                            string_type;
-   typedef          BOOST_REGEX_DETAIL_NS::named_subexpressions             named_sub_type;
-
-   // construct/copy/destroy:
-   explicit match_results(const Allocator& a = Allocator())
-      : m_subs(a), m_base(), m_null(), m_last_closed_paren(0), m_is_singular(true) {}
-   //
-   // IMPORTANT: in the code below, the crazy looking checks around m_is_singular are
-   // all required because it is illegal to copy a singular iterator.
-   // See https://svn.boost.org/trac/boost/ticket/3632.
-   //
-   match_results(const match_results& m)
-      : m_subs(m.m_subs), m_base(), m_null(), m_named_subs(m.m_named_subs), m_last_closed_paren(m.m_last_closed_paren), m_is_singular(m.m_is_singular)
-   {
-      if(!m_is_singular)
-      {
-         m_base = m.m_base;
-         m_null = m.m_null;
-      }
-   }
-   match_results& operator=(const match_results& m)
-   {
-      m_subs = m.m_subs;
-      m_named_subs = m.m_named_subs;
-      m_last_closed_paren = m.m_last_closed_paren;
-      m_is_singular = m.m_is_singular;
-      if(!m_is_singular)
-      {
-         m_base = m.m_base;
-         m_null = m.m_null;
-      }
-      return *this;
-   }
-   ~match_results(){}
-
-   // size:
-   size_type size() const
-   { return empty() ? 0 : m_subs.size() - 2; }
-   size_type max_size() const
-   { return m_subs.max_size(); }
-   bool empty() const
-   { return m_subs.size() < 2; }
-   // element access:
-   difference_type length(int sub = 0) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      sub += 2;
-      if((sub < (int)m_subs.size()) && (sub > 0))
-         return m_subs[sub].length();
-      return 0;
-   }
-   difference_type length(const char_type* sub) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      const char_type* sub_end = sub;
-      while(*sub_end) ++sub_end;
-      return length(named_subexpression_index(sub, sub_end));
-   }
-   template <class charT>
-   difference_type length(const charT* sub) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      const charT* sub_end = sub;
-      while(*sub_end) ++sub_end;
-      return length(named_subexpression_index(sub, sub_end));
-   }
-   template <class charT, class Traits, class A>
-   difference_type length(const std::basic_string<charT, Traits, A>& sub) const
-   {
-      return length(sub.c_str());
-   }
-   difference_type position(size_type sub = 0) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      sub += 2;
-      if(sub < m_subs.size())
-      {
-         const sub_match<BidiIterator>& s = m_subs[sub];
-         if(s.matched || (sub == 2))
-         {
-            return std::distance((BidiIterator)(m_base), (BidiIterator)(s.first));
-         }
-      }
-      return ~static_cast<difference_type>(0);
-   }
-   difference_type position(const char_type* sub) const
-   {
-      const char_type* sub_end = sub;
-      while(*sub_end) ++sub_end;
-      return position(named_subexpression_index(sub, sub_end));
-   }
-   template <class charT>
-   difference_type position(const charT* sub) const
-   {
-      const charT* sub_end = sub;
-      while(*sub_end) ++sub_end;
-      return position(named_subexpression_index(sub, sub_end));
-   }
-   template <class charT, class Traits, class A>
-   difference_type position(const std::basic_string<charT, Traits, A>& sub) const
-   {
-      return position(sub.c_str());
-   }
-   string_type str(int sub = 0) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      sub += 2;
-      string_type result;
-      if(sub < (int)m_subs.size() && (sub > 0))
-      {
-         const sub_match<BidiIterator>& s = m_subs[sub];
-         if(s.matched)
-         {
-            result = s.str();
-         }
-      }
-      return result;
-   }
-   string_type str(const char_type* sub) const
-   {
-      return (*this)[sub].str();
-   }
-   template <class Traits, class A>
-   string_type str(const std::basic_string<char_type, Traits, A>& sub) const
-   {
-      return (*this)[sub].str();
-   }
-   template <class charT>
-   string_type str(const charT* sub) const
-   {
-      return (*this)[sub].str();
-   }
-   template <class charT, class Traits, class A>
-   string_type str(const std::basic_string<charT, Traits, A>& sub) const
-   {
-      return (*this)[sub].str();
-   }
-   const_reference operator[](int sub) const
-   {
-      if(m_is_singular && m_subs.empty())
-         raise_logic_error();
-      sub += 2;
-      if(sub < (int)m_subs.size() && (sub >= 0))
-      {
-         return m_subs[sub];
-      }
-      return m_null;
-   }
-   //
-   // Named sub-expressions:
-   //
-   const_reference named_subexpression(const char_type* i, const char_type* j) const
-   {
-      //
-      // Scan for the leftmost *matched* subexpression with the specified named:
-      //
-      if(m_is_singular)
-         raise_logic_error();
-      BOOST_REGEX_DETAIL_NS::named_subexpressions::range_type r = m_named_subs->equal_range(i, j);
-      while((r.first != r.second) && ((*this)[r.first->index].matched == false))
-         ++r.first;
-      return r.first != r.second ? (*this)[r.first->index] : m_null;
-   }
-   template <class charT>
-   const_reference named_subexpression(const charT* i, const charT* j) const
-   {
-      static_assert(sizeof(charT) <= sizeof(char_type), "Failed internal logic");
-      if(i == j)
-         return m_null;
-      std::vector<char_type> s;
-      while(i != j)
-         s.insert(s.end(), *i++);
-      return named_subexpression(&*s.begin(), &*s.begin() + s.size());
-   }
-   int named_subexpression_index(const char_type* i, const char_type* j) const
-   {
-      //
-      // Scan for the leftmost *matched* subexpression with the specified named.
-      // If none found then return the leftmost expression with that name,
-      // otherwise an invalid index:
-      //
-      if(m_is_singular)
-         raise_logic_error();
-      BOOST_REGEX_DETAIL_NS::named_subexpressions::range_type s, r;
-      s = r = m_named_subs->equal_range(i, j);
-      while((r.first != r.second) && ((*this)[r.first->index].matched == false))
-         ++r.first;
-      if(r.first == r.second)
-         r = s;
-      return r.first != r.second ? r.first->index : -20;
-   }
-   template <class charT>
-   int named_subexpression_index(const charT* i, const charT* j) const
-   {
-      static_assert(sizeof(charT) <= sizeof(char_type), "Failed internal logic");
-      if(i == j)
-         return -20;
-      std::vector<char_type> s;
-      while(i != j)
-         s.insert(s.end(), *i++);
-      return named_subexpression_index(&*s.begin(), &*s.begin() + s.size());
-   }
-   template <class Traits, class A>
-   const_reference operator[](const std::basic_string<char_type, Traits, A>& s) const
-   {
-      return named_subexpression(s.c_str(), s.c_str() + s.size());
-   }
-   const_reference operator[](const char_type* p) const
-   {
-      const char_type* e = p;
-      while(*e) ++e;
-      return named_subexpression(p, e);
-   }
-
-   template <class charT>
-   const_reference operator[](const charT* p) const
-   {
-      static_assert(sizeof(charT) <= sizeof(char_type), "Failed internal logic");
-      if(*p == 0)
-         return m_null;
-      std::vector<char_type> s;
-      while(*p)
-         s.insert(s.end(), *p++);
-      return named_subexpression(&*s.begin(), &*s.begin() + s.size());
-   }
-   template <class charT, class Traits, class A>
-   const_reference operator[](const std::basic_string<charT, Traits, A>& ns) const
-   {
-      static_assert(sizeof(charT) <= sizeof(char_type), "Failed internal logic");
-      if(ns.empty())
-         return m_null;
-      std::vector<char_type> s;
-      for(unsigned i = 0; i < ns.size(); ++i)
-         s.insert(s.end(), ns[i]);
-      return named_subexpression(&*s.begin(), &*s.begin() + s.size());
-   }
-
-   const_reference prefix() const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      return (*this)[-1];
-   }
-
-   const_reference suffix() const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      return (*this)[-2];
-   }
-   const_iterator begin() const
-   {
-      return (m_subs.size() > 2) ? (m_subs.begin() + 2) : m_subs.end();
-   }
-   const_iterator end() const
-   {
-      return m_subs.end();
-   }
-   // format:
-   template <class OutputIterator, class Functor>
-   OutputIterator format(OutputIterator out,
-                         Functor fmt,
-                         match_flag_type flags = format_default) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      typedef typename BOOST_REGEX_DETAIL_NS::compute_functor_type<Functor, match_results<BidiIterator, Allocator>, OutputIterator>::type F;
-      F func(fmt);
-      return func(*this, out, flags);
-   }
-   template <class Functor>
-   string_type format(Functor fmt, match_flag_type flags = format_default) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      std::basic_string<char_type> result;
-      BOOST_REGEX_DETAIL_NS::string_out_iterator<std::basic_string<char_type> > i(result);
-
-      typedef typename BOOST_REGEX_DETAIL_NS::compute_functor_type<Functor, match_results<BidiIterator, Allocator>, BOOST_REGEX_DETAIL_NS::string_out_iterator<std::basic_string<char_type> > >::type F;
-      F func(fmt);
-
-      func(*this, i, flags);
-      return result;
-   }
-   // format with locale:
-   template <class OutputIterator, class Functor, class RegexT>
-   OutputIterator format(OutputIterator out,
-                         Functor fmt,
-                         match_flag_type flags,
-                         const RegexT& re) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      typedef ::boost::regex_traits_wrapper<typename RegexT::traits_type> traits_type;
-      typedef typename BOOST_REGEX_DETAIL_NS::compute_functor_type<Functor, match_results<BidiIterator, Allocator>, OutputIterator, traits_type>::type F;
-      F func(fmt);
-      return func(*this, out, flags, re.get_traits());
-   }
-   template <class RegexT, class Functor>
-   string_type format(Functor fmt,
-                      match_flag_type flags,
-                      const RegexT& re) const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      typedef ::boost::regex_traits_wrapper<typename RegexT::traits_type> traits_type;
-      std::basic_string<char_type> result;
-      BOOST_REGEX_DETAIL_NS::string_out_iterator<std::basic_string<char_type> > i(result);
-
-      typedef typename BOOST_REGEX_DETAIL_NS::compute_functor_type<Functor, match_results<BidiIterator, Allocator>, BOOST_REGEX_DETAIL_NS::string_out_iterator<std::basic_string<char_type> >, traits_type >::type F;
-      F func(fmt);
-
-      func(*this, i, flags, re.get_traits());
-      return result;
-   }
-
-   const_reference get_last_closed_paren()const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      return m_last_closed_paren == 0 ? m_null : (*this)[m_last_closed_paren];
-   }
-
-   allocator_type get_allocator() const
-   {
-      return m_subs.get_allocator();
-   }
-   void swap(match_results& that)
-   {
-      std::swap(m_subs, that.m_subs);
-      std::swap(m_named_subs, that.m_named_subs);
-      std::swap(m_last_closed_paren, that.m_last_closed_paren);
-      if(m_is_singular)
-      {
-         if(!that.m_is_singular)
-         {
-            m_base = that.m_base;
-            m_null = that.m_null;
-         }
-      }
-      else if(that.m_is_singular)
-      {
-         that.m_base = m_base;
-         that.m_null = m_null;
-      }
-      else
-      {
-         std::swap(m_base, that.m_base);
-         std::swap(m_null, that.m_null);
-      }
-      std::swap(m_is_singular, that.m_is_singular);
-   }
-   bool operator==(const match_results& that)const
-   {
-      if(m_is_singular)
-      {
-         return that.m_is_singular;
-      }
-      else if(that.m_is_singular)
-      {
-         return false;
-      }
-      return (m_subs == that.m_subs) && (m_base == that.m_base) && (m_last_closed_paren == that.m_last_closed_paren);
-   }
-   bool operator!=(const match_results& that)const
-   { return !(*this == that); }
-
-#ifdef BOOST_REGEX_MATCH_EXTRA
-   typedef typename sub_match<BidiIterator>::capture_sequence_type capture_sequence_type;
-
-   const capture_sequence_type& captures(int i)const
-   {
-      if(m_is_singular)
-         raise_logic_error();
-      return (*this)[i].captures();
-   }
-#endif
-
-   //
-   // private access functions:
-   void  set_second(BidiIterator i)
-   {
-      BOOST_REGEX_ASSERT(m_subs.size() > 2);
-      m_subs[2].second = i;
-      m_subs[2].matched = true;
-      m_subs[0].first = i;
-      m_subs[0].matched = (m_subs[0].first != m_subs[0].second);
-      m_null.first = i;
-      m_null.second = i;
-      m_null.matched = false;
-      m_is_singular = false;
-   }
-
-   void  set_second(BidiIterator i, size_type pos, bool m = true, bool escape_k = false)
-   {
-      if(pos)
-         m_last_closed_paren = static_cast<int>(pos);
-      pos += 2;
-      BOOST_REGEX_ASSERT(m_subs.size() > pos);
-      m_subs[pos].second = i;
-      m_subs[pos].matched = m;
-      if((pos == 2) && !escape_k)
-      {
-         m_subs[0].first = i;
-         m_subs[0].matched = (m_subs[0].first != m_subs[0].second);
-         m_null.first = i;
-         m_null.second = i;
-         m_null.matched = false;
-         m_is_singular = false;
-      }
-   }
-   void  set_size(size_type n, BidiIterator i, BidiIterator j)
-   {
-      value_type v(j);
-      size_type len = m_subs.size();
-      if(len > n + 2)
-      {
-         m_subs.erase(m_subs.begin()+n+2, m_subs.end());
-         std::fill(m_subs.begin(), m_subs.end(), v);
-      }
-      else
-      {
-         std::fill(m_subs.begin(), m_subs.end(), v);
-         if(n+2 != len)
-            m_subs.insert(m_subs.end(), n+2-len, v);
-      }
-      m_subs[1].first = i;
-      m_last_closed_paren = 0;
-   }
-   void  set_base(BidiIterator pos)
-   {
-      m_base = pos;
-   }
-   BidiIterator base()const
-   {
-      return m_base;
-   }
-   void  set_first(BidiIterator i)
-   {
-      BOOST_REGEX_ASSERT(m_subs.size() > 2);
-      // set up prefix:
-      m_subs[1].second = i;
-      m_subs[1].matched = (m_subs[1].first != i);
-      // set up $0:
-      m_subs[2].first = i;
-      // zero out everything else:
-      for(size_type n = 3; n < m_subs.size(); ++n)
-      {
-         m_subs[n].first = m_subs[n].second = m_subs[0].second;
-         m_subs[n].matched = false;
-      }
-   }
-   void  set_first(BidiIterator i, size_type pos, bool escape_k = false)
-   {
-      BOOST_REGEX_ASSERT(pos+2 < m_subs.size());
-      if(pos || escape_k)
-      {
-         m_subs[pos+2].first = i;
-         if(escape_k)
-         {
-            m_subs[1].second = i;
-            m_subs[1].matched = (m_subs[1].first != m_subs[1].second);
-         }
-      }
-      else
-         set_first(i);
-   }
-   void  maybe_assign(const match_results<BidiIterator, Allocator>& m);
-
-   void  set_named_subs(std::shared_ptr<named_sub_type> subs)
-   {
-      m_named_subs = subs;
-   }
-
-private:
-   //
-   // Error handler called when an uninitialized match_results is accessed:
-   //
-   static void raise_logic_error()
-   {
-      std::logic_error e("Attempt to access an uninitialized boost::match_results<> class.");
-#ifndef BOOST_REGEX_STANDALONE
-      boost::throw_exception(e);
-#else
-      throw e;
-#endif
-   }
-
-
-   vector_type            m_subs;                      // subexpressions
-   BidiIterator   m_base;                              // where the search started from
-   sub_match<BidiIterator> m_null;                     // a null match
-   std::shared_ptr<named_sub_type> m_named_subs;     // Shared copy of named subs in the regex object
-   int m_last_closed_paren;                            // Last ) to be seen - used for formatting
-   bool m_is_singular;                                 // True if our stored iterators are singular
-};
-
-template <class BidiIterator, class Allocator>
-void  match_results<BidiIterator, Allocator>::maybe_assign(const match_results<BidiIterator, Allocator>& m)
-{
-   if(m_is_singular)
-   {
-      *this = m;
-      return;
-   }
-   const_iterator p1, p2;
-   p1 = begin();
-   p2 = m.begin();
-   //
-   // Distances are measured from the start of *this* match, unless this isn't
-   // a valid match in which case we use the start of the whole sequence.  Note that
-   // no subsequent match-candidate can ever be to the left of the first match found.
-   // This ensures that when we are using bidirectional iterators, that distances 
-   // measured are as short as possible, and therefore as efficient as possible
-   // to compute.  Finally note that we don't use the "matched" data member to test
-   // whether a sub-expression is a valid match, because partial matches set this
-   // to false for sub-expression 0.
-   //
-   BidiIterator l_end = this->suffix().second;
-   BidiIterator l_base = (p1->first == l_end) ? this->prefix().first : (*this)[0].first;
-   difference_type len1 = 0;
-   difference_type len2 = 0;
-   difference_type base1 = 0;
-   difference_type base2 = 0;
-   std::size_t i;
-   for(i = 0; i < size(); ++i, ++p1, ++p2)
-   {
-      //
-      // Leftmost takes priority over longest; handle special cases
-      // where distances need not be computed first (an optimisation
-      // for bidirectional iterators: ensure that we don't accidently
-      // compute the length of the whole sequence, as this can be really
-      // expensive).
-      //
-      if(p1->first == l_end)
-      {
-         if(p2->first != l_end)
-         {
-            // p2 must be better than p1, and no need to calculate
-            // actual distances:
-            base1 = 1;
-            base2 = 0;
-            break;
-         }
-         else
-         {
-            // *p1 and *p2 are either unmatched or match end-of sequence,
-            // either way no need to calculate distances:
-            if((p1->matched == false) && (p2->matched == true))
-               break;
-            if((p1->matched == true) && (p2->matched == false))
-               return;
-            continue;
-         }
-      }
-      else if(p2->first == l_end)
-      {
-         // p1 better than p2, and no need to calculate distances:
-         return;
-      }
-      base1 = std::distance(l_base, p1->first);
-      base2 = std::distance(l_base, p2->first);
-      BOOST_REGEX_ASSERT(base1 >= 0);
-      BOOST_REGEX_ASSERT(base2 >= 0);
-      if(base1 < base2) return;
-      if(base2 < base1) break;
-
-      len1 = std::distance((BidiIterator)p1->first, (BidiIterator)p1->second);
-      len2 = std::distance((BidiIterator)p2->first, (BidiIterator)p2->second);
-      BOOST_REGEX_ASSERT(len1 >= 0);
-      BOOST_REGEX_ASSERT(len2 >= 0);
-      if((len1 != len2) || ((p1->matched == false) && (p2->matched == true)))
-         break;
-      if((p1->matched == true) && (p2->matched == false))
-         return;
-   }
-   if(i == size())
-      return;
-   if(base2 < base1)
-      *this = m;
-   else if((len2 > len1) || ((p1->matched == false) && (p2->matched == true)) )
-      *this = m;
-}
-
-template <class BidiIterator, class Allocator>
-void swap(match_results<BidiIterator, Allocator>& a, match_results<BidiIterator, Allocator>& b)
-{
-   a.swap(b);
-}
-
-template <class charT, class traits, class BidiIterator, class Allocator>
-std::basic_ostream<charT, traits>&
-   operator << (std::basic_ostream<charT, traits>& os,
-                const match_results<BidiIterator, Allocator>& s)
-{
-   return (os << s.str());
-}
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-} // namespace boost
-
-#endif
-
-

regex/include/boost/regex/v5/mem_block_cache.hpp

@@ -1,173 +0,0 @@
- /*
- * Copyright (c) 2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         mem_block_cache.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: memory block cache used by the non-recursive matcher.
-  */
-
-#ifndef BOOST_REGEX_V5_MEM_BLOCK_CACHE_HPP
-#define BOOST_REGEX_V5_MEM_BLOCK_CACHE_HPP
-
-#include <new>
-#ifdef BOOST_HAS_THREADS
-#include <mutex>
-#endif
-
-#ifndef BOOST_NO_CXX11_HDR_ATOMIC
-  #include <atomic>
-  #if ATOMIC_POINTER_LOCK_FREE == 2
-    #define BOOST_REGEX_MEM_BLOCK_CACHE_LOCK_FREE
-    #define BOOST_REGEX_ATOMIC_POINTER std::atomic
-  #endif
-#endif
-
-namespace boost{
-namespace BOOST_REGEX_DETAIL_NS{
-
-#if BOOST_REGEX_MAX_CACHE_BLOCKS != 0
-#ifdef BOOST_REGEX_MEM_BLOCK_CACHE_LOCK_FREE /* lock free implementation */
-struct mem_block_cache
-{
-  std::atomic<void*> cache[BOOST_REGEX_MAX_CACHE_BLOCKS];
-
-   ~mem_block_cache()
-   {
-     for (size_t i = 0;i < BOOST_REGEX_MAX_CACHE_BLOCKS; ++i) {
-       if (cache[i].load()) ::operator delete(cache[i].load());
-     }
-   }
-   void* get()
-   {
-     for (size_t i = 0;i < BOOST_REGEX_MAX_CACHE_BLOCKS; ++i) {
-       void* p = cache[i].load();
-       if (p != NULL) {
-         if (cache[i].compare_exchange_strong(p, NULL)) return p;
-       }
-     }
-     return ::operator new(BOOST_REGEX_BLOCKSIZE);
-   }
-   void put(void* ptr)
-   {
-     for (size_t i = 0;i < BOOST_REGEX_MAX_CACHE_BLOCKS; ++i) {
-       void* p = cache[i].load();
-       if (p == NULL) {
-         if (cache[i].compare_exchange_strong(p, ptr)) return;
-       }
-     }
-     ::operator delete(ptr);
-   }
-
-   static mem_block_cache& instance()
-   {
-      static mem_block_cache block_cache = { { {nullptr} } };
-      return block_cache;
-   }
-};
-
-
-#else /* lock-based implementation */
-
-
-struct mem_block_node
-{
-   mem_block_node* next;
-};
-
-struct mem_block_cache
-{
-   // this member has to be statically initialsed:
-   mem_block_node* next { nullptr };
-   unsigned cached_blocks { 0 };
-#ifdef BOOST_HAS_THREADS
-   std::mutex mut;
-#endif
-
-   ~mem_block_cache()
-   {
-      while(next)
-      {
-         mem_block_node* old = next;
-         next = next->next;
-         ::operator delete(old);
-      }
-   }
-   void* get()
-   {
-#ifdef BOOST_HAS_THREADS
-      std::lock_guard<std::mutex> g(mut);
-#endif
-     if(next)
-      {
-         mem_block_node* result = next;
-         next = next->next;
-         --cached_blocks;
-         return result;
-      }
-      return ::operator new(BOOST_REGEX_BLOCKSIZE);
-   }
-   void put(void* p)
-   {
-#ifdef BOOST_HAS_THREADS
-      std::lock_guard<std::mutex> g(mut);
-#endif
-      if(cached_blocks >= BOOST_REGEX_MAX_CACHE_BLOCKS)
-      {
-         ::operator delete(p);
-      }
-      else
-      {
-         mem_block_node* old = static_cast<mem_block_node*>(p);
-         old->next = next;
-         next = old;
-         ++cached_blocks;
-      }
-   }
-   static mem_block_cache& instance()
-   {
-      static mem_block_cache block_cache;
-      return block_cache;
-   }
-};
-#endif
-#endif
-
-#if BOOST_REGEX_MAX_CACHE_BLOCKS == 0
-
-inline void*  get_mem_block()
-{
-   return ::operator new(BOOST_REGEX_BLOCKSIZE);
-}
-
-inline void  put_mem_block(void* p)
-{
-   ::operator delete(p);
-}
-
-#else
-
-inline void*  get_mem_block()
-{
-   return mem_block_cache::instance().get();
-}
-
-inline void  put_mem_block(void* p)
-{
-   mem_block_cache::instance().put(p);
-}
-
-#endif
-}
-} // namespace boost
-
-#endif
-

regex/include/boost/regex/v5/object_cache.hpp

@@ -1,160 +0,0 @@
-/*
- *
- * Copyright (c) 2004
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         object_cache.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Implements a generic object cache.
-  */
-
-#ifndef BOOST_REGEX_OBJECT_CACHE_HPP
-#define BOOST_REGEX_OBJECT_CACHE_HPP
-
-#include <boost/regex/config.hpp>
-#include <memory>
-#include <map>
-#include <list>
-#include <stdexcept>
-#include <string>
-#ifdef BOOST_HAS_THREADS
-#include <mutex>
-#endif
-
-namespace boost{
-
-template <class Key, class Object>
-class object_cache
-{
-public:
-   typedef std::pair< ::std::shared_ptr<Object const>, Key const*> value_type;
-   typedef std::list<value_type> list_type;
-   typedef typename list_type::iterator list_iterator;
-   typedef std::map<Key, list_iterator> map_type;
-   typedef typename map_type::iterator map_iterator;
-   typedef typename list_type::size_type size_type;
-   static std::shared_ptr<Object const> get(const Key& k, size_type l_max_cache_size);
-
-private:
-   static std::shared_ptr<Object const> do_get(const Key& k, size_type l_max_cache_size);
-
-   struct data
-   {
-      list_type   cont;
-      map_type    index;
-   };
-
-   // Needed by compilers not implementing the resolution to DR45. For reference,
-   // see http://www.open-std.org/JTC1/SC22/WG21/docs/cwg_defects.html#45.
-   friend struct data;
-};
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable: 4702)
-#endif
-template <class Key, class Object>
-std::shared_ptr<Object const> object_cache<Key, Object>::get(const Key& k, size_type l_max_cache_size)
-{
-#ifdef BOOST_HAS_THREADS
-   static std::mutex mut;
-   std::lock_guard<std::mutex> l(mut);
-   return do_get(k, l_max_cache_size);
-#else
-   return do_get(k, l_max_cache_size);
-#endif
-}
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-
-template <class Key, class Object>
-std::shared_ptr<Object const> object_cache<Key, Object>::do_get(const Key& k, size_type l_max_cache_size)
-{
-   typedef typename object_cache<Key, Object>::data object_data;
-   typedef typename map_type::size_type map_size_type;
-   static object_data s_data;
-
-   //
-   // see if the object is already in the cache:
-   //
-   map_iterator mpos = s_data.index.find(k);
-   if(mpos != s_data.index.end())
-   {
-      //
-      // Eureka! 
-      // We have a cached item, bump it up the list and return it:
-      //
-      if(--(s_data.cont.end()) != mpos->second)
-      {
-         // splice out the item we want to move:
-         list_type temp;
-         temp.splice(temp.end(), s_data.cont, mpos->second);
-         // and now place it at the end of the list:
-         s_data.cont.splice(s_data.cont.end(), temp, temp.begin());
-         BOOST_REGEX_ASSERT(*(s_data.cont.back().second) == k);
-         // update index with new position:
-         mpos->second = --(s_data.cont.end());
-         BOOST_REGEX_ASSERT(&(mpos->first) == mpos->second->second);
-         BOOST_REGEX_ASSERT(&(mpos->first) == s_data.cont.back().second);
-      }
-      return s_data.cont.back().first;
-   }
-   //
-   // if we get here then the item is not in the cache,
-   // so create it:
-   //
-   std::shared_ptr<Object const> result(new Object(k));
-   //
-   // Add it to the list, and index it:
-   //
-   s_data.cont.push_back(value_type(result, static_cast<Key const*>(0)));
-   s_data.index.insert(std::make_pair(k, --(s_data.cont.end())));
-   s_data.cont.back().second = &(s_data.index.find(k)->first);
-   map_size_type s = s_data.index.size();
-   BOOST_REGEX_ASSERT(s_data.index[k]->first.get() == result.get());
-   BOOST_REGEX_ASSERT(&(s_data.index.find(k)->first) == s_data.cont.back().second);
-   BOOST_REGEX_ASSERT(s_data.index.find(k)->first == k);
-   if(s > l_max_cache_size)
-   {
-      //
-      // We have too many items in the list, so we need to start
-      // popping them off the back of the list, but only if they're
-      // being held uniquely by us:
-      //
-      list_iterator pos = s_data.cont.begin();
-      list_iterator last = s_data.cont.end();
-      while((pos != last) && (s > l_max_cache_size))
-      {
-         if(pos->first.use_count() == 1)
-         {
-            list_iterator condemmed(pos);
-            ++pos;
-            // now remove the items from our containers, 
-            // then order has to be as follows:
-            BOOST_REGEX_ASSERT(s_data.index.find(*(condemmed->second)) != s_data.index.end());
-            s_data.index.erase(*(condemmed->second));
-            s_data.cont.erase(condemmed); 
-            --s;
-         }
-         else
-            ++pos;
-      }
-      BOOST_REGEX_ASSERT(s_data.index[k]->first.get() == result.get());
-      BOOST_REGEX_ASSERT(&(s_data.index.find(k)->first) == s_data.cont.back().second);
-      BOOST_REGEX_ASSERT(s_data.index.find(k)->first == k);
-   }
-   return result;
-}
-
-}
-
-#endif

regex/include/boost/regex/v5/pattern_except.hpp

@@ -1,106 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         pattern_except.hpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares pattern-matching exception classes.
-  */
-
-#ifndef BOOST_RE_V5_PAT_EXCEPT_HPP
-#define BOOST_RE_V5_PAT_EXCEPT_HPP
-
-#ifndef BOOST_REGEX_CONFIG_HPP
-#include <boost/regex/config.hpp>
-#endif
-
-#include <cstddef>
-#include <stdexcept>
-#include <boost/regex/v5/error_type.hpp>
-#include <boost/regex/v5/regex_traits_defaults.hpp>
-
-namespace boost{
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable : 4275)
-#if BOOST_REGEX_MSVC >= 1800
-#pragma warning(disable : 26812 4459)
-#endif
-#endif
-class regex_error : public std::runtime_error
-{
-public:
-   explicit regex_error(const std::string& s, regex_constants::error_type err = regex_constants::error_unknown, std::ptrdiff_t pos = 0)
-      : std::runtime_error(s)
-      , m_error_code(err)
-      , m_position(pos)
-   {
-   }
-   explicit regex_error(regex_constants::error_type err)
-      : std::runtime_error(::boost::BOOST_REGEX_DETAIL_NS::get_default_error_string(err))
-      , m_error_code(err)
-      , m_position(0)
-   {
-   }
-   ~regex_error() noexcept override {}
-   regex_constants::error_type code()const
-   { return m_error_code; }
-   std::ptrdiff_t position()const
-   { return m_position; }
-   void raise()const 
-   {
-#ifndef BOOST_NO_EXCEPTIONS
-#ifndef BOOST_REGEX_STANDALONE
-      ::boost::throw_exception(*this);
-#else
-      throw* this;
-#endif
-#endif
-   }
-private:
-   regex_constants::error_type m_error_code;
-   std::ptrdiff_t m_position;
-};
-
-typedef regex_error bad_pattern;
-typedef regex_error bad_expression;
-
-namespace BOOST_REGEX_DETAIL_NS{
-
-template <class E>
-inline void raise_runtime_error(const E& ex)
-{
-#ifndef BOOST_REGEX_STANDALONE
-   ::boost::throw_exception(ex);
-#else
-   throw ex;
-#endif
-}
-
-template <class traits>
-void raise_error(const traits& t, regex_constants::error_type code)
-{
-   (void)t;  // warning suppression
-   regex_error e(t.error_string(code), code, 0);
-   ::boost::BOOST_REGEX_DETAIL_NS::raise_runtime_error(e);
-}
-
-}
-
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-
-} // namespace boost
-
-#endif

regex/include/boost/regex/v5/perl_matcher.hpp

@@ -1,576 +0,0 @@
-/*
- *
- * Copyright (c) 2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
-#ifndef BOOST_REGEX_MATCHER_HPP
-#define BOOST_REGEX_MATCHER_HPP
-
-#include <boost/regex/v5/iterator_category.hpp>
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(push)
-#pragma warning(disable : 4251 4459)
-#if BOOST_REGEX_MSVC < 1700
-#     pragma warning(disable : 4231)
-#endif
-#  if BOOST_REGEX_MSVC < 1600
-#     pragma warning(disable : 4660)
-#  endif
-#if BOOST_REGEX_MSVC < 1910
-#pragma warning(disable:4800)
-#endif
-#endif
-
-namespace boost{
-namespace BOOST_REGEX_DETAIL_NS{
-
-//
-// error checking API:
-//
-inline void  verify_options(boost::regex_constants::syntax_option_type, match_flag_type mf)
-{
-   //
-   // can't mix match_extra with POSIX matching rules:
-   //
-   if ((mf & match_extra) && (mf & match_posix))
-   {
-      std::logic_error msg("Usage Error: Can't mix regular expression captures with POSIX matching rules");
-#ifndef BOOST_REGEX_STANDALONE
-      throw_exception(msg);
-#else
-      throw msg;
-#endif
-   }
-}
-//
-// function can_start:
-//
-template <class charT>
-inline bool can_start(charT c, const unsigned char* map, unsigned char mask)
-{
-   return ((c < static_cast<charT>(0)) ? true : ((c >= static_cast<charT>(1 << CHAR_BIT)) ? true : map[c] & mask));
-}
-inline bool can_start(char c, const unsigned char* map, unsigned char mask)
-{
-   return map[(unsigned char)c] & mask;
-}
-inline bool can_start(signed char c, const unsigned char* map, unsigned char mask)
-{
-   return map[(unsigned char)c] & mask;
-}
-inline bool can_start(unsigned char c, const unsigned char* map, unsigned char mask)
-{
-   return map[c] & mask;
-}
-inline bool can_start(unsigned short c, const unsigned char* map, unsigned char mask)
-{
-   return ((c >= (1 << CHAR_BIT)) ? true : map[c] & mask);
-}
-#if defined(WCHAR_MIN) && (WCHAR_MIN == 0) && !defined(BOOST_NO_INTRINSIC_WCHAR_T)
-inline bool can_start(wchar_t c, const unsigned char* map, unsigned char mask)
-{
-   return ((c >= static_cast<wchar_t>(1u << CHAR_BIT)) ? true : map[c] & mask);
-}
-#endif
-#if !defined(BOOST_NO_INTRINSIC_WCHAR_T)
-inline bool can_start(unsigned int c, const unsigned char* map, unsigned char mask)
-{
-   return (((c >= static_cast<unsigned int>(1u << CHAR_BIT)) ? true : map[c] & mask));
-}
-#endif
-
-template <class C, class T, class A>
-inline int string_compare(const std::basic_string<C,T,A>& s, const C* p)
-{ 
-   if(0 == *p)
-   {
-      if(s.empty() || ((s.size() == 1) && (s[0] == 0)))
-         return 0;
-   }
-   return s.compare(p); 
-}
-template <class Seq, class C>
-inline int string_compare(const Seq& s, const C* p)
-{
-   std::size_t i = 0;
-   while((i < s.size()) && (p[i] == s[i]))
-   {
-      ++i;
-   }
-   return (i == s.size()) ? -(int)p[i] : (int)s[i] - (int)p[i];
-}
-# define STR_COMP(s,p) string_compare(s,p)
-
-template<class charT>
-inline const charT* re_skip_past_null(const charT* p)
-{
-  while (*p != static_cast<charT>(0)) ++p;
-  return ++p;
-}
-
-template <class iterator, class charT, class traits_type, class char_classT>
-iterator  re_is_set_member(iterator next, 
-                          iterator last, 
-                          const re_set_long<char_classT>* set_, 
-                          const regex_data<charT, traits_type>& e, bool icase)
-{   
-   const charT* p = reinterpret_cast<const charT*>(set_+1);
-   iterator ptr;
-   unsigned int i;
-   //bool icase = e.m_flags & regex_constants::icase;
-
-   if(next == last) return next;
-
-   typedef typename traits_type::string_type traits_string_type;
-   const ::boost::regex_traits_wrapper<traits_type>& traits_inst = *(e.m_ptraits);
-   
-   // dwa 9/13/00 suppress incorrect MSVC warning - it claims this is never
-   // referenced
-   (void)traits_inst;
-
-   // try and match a single character, could be a multi-character
-   // collating element...
-   for(i = 0; i < set_->csingles; ++i)
-   {
-      ptr = next;
-      if(*p == static_cast<charT>(0))
-      {
-         // treat null string as special case:
-         if(traits_inst.translate(*ptr, icase))
-         {
-            ++p;
-            continue;
-         }
-         return set_->isnot ? next : (ptr == next) ? ++next : ptr;
-      }
-      else
-      {
-         while(*p && (ptr != last))
-         {
-            if(traits_inst.translate(*ptr, icase) != *p)
-               break;
-            ++p;
-            ++ptr;
-         }
-
-         if(*p == static_cast<charT>(0)) // if null we've matched
-            return set_->isnot ? next : (ptr == next) ? ++next : ptr;
-
-         p = re_skip_past_null(p);     // skip null
-      }
-   }
-
-   charT col = traits_inst.translate(*next, icase);
-
-
-   if(set_->cranges || set_->cequivalents)
-   {
-      traits_string_type s1;
-      //
-      // try and match a range, NB only a single character can match
-      if(set_->cranges)
-      {
-         if((e.m_flags & regex_constants::collate) == 0)
-            s1.assign(1, col);
-         else
-         {
-            charT a[2] = { col, charT(0), };
-            s1 = traits_inst.transform(a, a + 1);
-         }
-         for(i = 0; i < set_->cranges; ++i)
-         {
-            if(STR_COMP(s1, p) >= 0)
-            {
-               do{ ++p; }while(*p);
-               ++p;
-               if(STR_COMP(s1, p) <= 0)
-                  return set_->isnot ? next : ++next;
-            }
-            else
-            {
-               // skip first string
-               do{ ++p; }while(*p);
-               ++p;
-            }
-            // skip second string
-            do{ ++p; }while(*p);
-            ++p;
-         }
-      }
-      //
-      // try and match an equivalence class, NB only a single character can match
-      if(set_->cequivalents)
-      {
-         charT a[2] = { col, charT(0), };
-         s1 = traits_inst.transform_primary(a, a +1);
-         for(i = 0; i < set_->cequivalents; ++i)
-         {
-            if(STR_COMP(s1, p) == 0)
-               return set_->isnot ? next : ++next;
-            // skip string
-            do{ ++p; }while(*p);
-            ++p;
-         }
-      }
-   }
-   if(traits_inst.isctype(col, set_->cclasses) == true)
-      return set_->isnot ? next : ++next;
-   if((set_->cnclasses != 0) && (traits_inst.isctype(col, set_->cnclasses) == false))
-      return set_->isnot ? next : ++next;
-   return set_->isnot ? ++next : next;
-}
-
-template <class BidiIterator>
-class repeater_count
-{
-   repeater_count** stack;
-   repeater_count* next;
-   int state_id;
-   std::size_t count;        // the number of iterations so far
-   BidiIterator start_pos;   // where the last repeat started
-
-   repeater_count* unwind_until(int n, repeater_count* p, int current_recursion_id)
-   { 
-      while(p && (p->state_id != n))
-      {
-         if(-2 - current_recursion_id == p->state_id)
-            return 0;
-         p = p->next;
-         if(p && (p->state_id < 0))
-         {
-            p = unwind_until(p->state_id, p, current_recursion_id);
-            if(!p)
-               return p;
-            p = p->next;
-         }
-      }
-      return p;
-   }
-public:
-   repeater_count(repeater_count** s) : stack(s), next(0), state_id(-1), count(0), start_pos() {}
-   
-   repeater_count(int i, repeater_count** s, BidiIterator start, int current_recursion_id)
-      : start_pos(start)
-   {
-      state_id = i;
-      stack = s;
-      next = *stack;
-      *stack = this;
-      if((state_id > next->state_id) && (next->state_id >= 0))
-         count = 0;
-      else
-      {
-         repeater_count* p = next;
-         p = unwind_until(state_id, p, current_recursion_id);
-         if(p)
-         {
-            count = p->count;
-            start_pos = p->start_pos;
-         }
-         else
-            count = 0;
-      }
-   }
-   ~repeater_count()
-   {
-      if(next)
-         *stack = next;
-   }
-   std::size_t get_count() { return count; }
-   int get_id() { return state_id; }
-   std::size_t operator++() { return ++count; }
-   bool check_null_repeat(const BidiIterator& pos, std::size_t max)
-   {
-      // this is called when we are about to start a new repeat,
-      // if the last one was NULL move our count to max,
-      // otherwise save the current position.
-      bool result = (count == 0) ? false : (pos == start_pos);
-      if(result)
-         count = max;
-      else
-         start_pos = pos;
-      return result;
-   }
-};
-
-struct saved_state;
-
-enum saved_state_type
-{
-   saved_type_end = 0,
-   saved_type_paren = 1,
-   saved_type_recurse = 2,
-   saved_type_assertion = 3,
-   saved_state_alt = 4,
-   saved_state_repeater_count = 5,
-   saved_state_extra_block = 6,
-   saved_state_greedy_single_repeat = 7,
-   saved_state_rep_slow_dot = 8,
-   saved_state_rep_fast_dot = 9,
-   saved_state_rep_char = 10,
-   saved_state_rep_short_set = 11,
-   saved_state_rep_long_set = 12,
-   saved_state_non_greedy_long_repeat = 13, 
-   saved_state_count = 14
-};
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(push)
-#if BOOST_REGEX_MSVC >= 1800
-#pragma warning(disable:26495)
-#endif
-#endif
-template <class Results>
-struct recursion_info
-{
-   typedef typename Results::value_type value_type;
-   typedef typename value_type::iterator iterator;
-   int idx;
-   const re_syntax_base* preturn_address;
-   Results results;
-   repeater_count<iterator>* repeater_stack;
-   iterator location_of_start;
-};
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(pop)
-#endif
-
-template <class BidiIterator, class Allocator, class traits>
-class perl_matcher
-{
-public:
-   typedef typename traits::char_type char_type;
-   typedef perl_matcher<BidiIterator, Allocator, traits> self_type;
-   typedef bool (self_type::*matcher_proc_type)();
-   typedef std::size_t traits_size_type;
-   typedef typename is_byte<char_type>::width_type width_type;
-   typedef typename std::iterator_traits<BidiIterator>::difference_type difference_type;
-   typedef match_results<BidiIterator, Allocator> results_type;
-
-   perl_matcher(BidiIterator first, BidiIterator end, 
-      match_results<BidiIterator, Allocator>& what, 
-      const basic_regex<char_type, traits>& e,
-      match_flag_type f,
-      BidiIterator l_base)
-      :  m_result(what), base(first), last(end), 
-         position(first), backstop(l_base), re(e), traits_inst(e.get_traits()), 
-         m_independent(false), next_count(&rep_obj), rep_obj(&next_count)
-      , m_recursions(0)
-   {
-      construct_init(e, f);
-   }
-
-   bool match();
-   bool find();
-
-   void setf(match_flag_type f)
-   { m_match_flags |= f; }
-   void unsetf(match_flag_type f)
-   { m_match_flags &= ~f; }
-
-private:
-   void construct_init(const basic_regex<char_type, traits>& e, match_flag_type f);
-
-   bool find_imp();
-   bool match_imp();
-   void estimate_max_state_count(std::random_access_iterator_tag*);
-   void estimate_max_state_count(void*);
-   bool match_prefix();
-   bool match_all_states();
-
-   // match procs, stored in s_match_vtable:
-   bool match_startmark();
-   bool match_endmark();
-   bool match_literal();
-   bool match_start_line();
-   bool match_end_line();
-   bool match_wild();
-   bool match_match();
-   bool match_word_boundary();
-   bool match_within_word();
-   bool match_word_start();
-   bool match_word_end();
-   bool match_buffer_start();
-   bool match_buffer_end();
-   bool match_backref();
-   bool match_long_set();
-   bool match_set();
-   bool match_jump();
-   bool match_alt();
-   bool match_rep();
-   bool match_combining();
-   bool match_soft_buffer_end();
-   bool match_restart_continue();
-   bool match_long_set_repeat();
-   bool match_set_repeat();
-   bool match_char_repeat();
-   bool match_dot_repeat_fast();
-   bool match_dot_repeat_slow();
-   bool match_dot_repeat_dispatch()
-   {
-      return ::boost::is_random_access_iterator<BidiIterator>::value ? match_dot_repeat_fast() : match_dot_repeat_slow();
-   }
-   bool match_backstep();
-   bool match_assert_backref();
-   bool match_toggle_case();
-   bool match_recursion();
-   bool match_fail();
-   bool match_accept();
-   bool match_commit();
-   bool match_then();
-   bool skip_until_paren(int index, bool match = true);
-
-   // find procs stored in s_find_vtable:
-   bool find_restart_any();
-   bool find_restart_word();
-   bool find_restart_line();
-   bool find_restart_buf();
-   bool find_restart_lit();
-
-private:
-   // final result structure to be filled in:
-   match_results<BidiIterator, Allocator>& m_result;
-   // temporary result for POSIX matches:
-   std::unique_ptr<match_results<BidiIterator, Allocator> > m_temp_match;
-   // pointer to actual result structure to fill in:
-   match_results<BidiIterator, Allocator>* m_presult;
-   // start of sequence being searched:
-   BidiIterator base;
-   // end of sequence being searched:
-   BidiIterator last; 
-   // current character being examined:
-   BidiIterator position;
-   // where to restart next search after failed match attempt:
-   BidiIterator restart;
-   // where the current search started from, acts as base for $` during grep:
-   BidiIterator search_base;
-   // how far we can go back when matching lookbehind:
-   BidiIterator backstop;
-   // the expression being examined:
-   const basic_regex<char_type, traits>& re;
-   // the expression's traits class:
-   const ::boost::regex_traits_wrapper<traits>& traits_inst;
-   // the next state in the machine being matched:
-   const re_syntax_base* pstate;
-   // matching flags in use:
-   match_flag_type m_match_flags;
-   // how many states we have examined so far:
-   std::ptrdiff_t state_count;
-   // max number of states to examine before giving up:
-   std::ptrdiff_t max_state_count;
-   // whether we should ignore case or not:
-   bool icase;
-   // set to true when (position == last), indicates that we may have a partial match:
-   bool m_has_partial_match;
-   // set to true whenever we get a match:
-   bool m_has_found_match;
-   // set to true whenever we're inside an independent sub-expression:
-   bool m_independent;
-   // the current repeat being examined:
-   repeater_count<BidiIterator>* next_count;
-   // the first repeat being examined (top of linked list):
-   repeater_count<BidiIterator> rep_obj;
-   // the mask to pass when matching word boundaries:
-   typename traits::char_class_type m_word_mask;
-   // the bitmask to use when determining whether a match_any matches a newline or not:
-   unsigned char match_any_mask;
-   // recursion information:
-   std::vector<recursion_info<results_type> > recursion_stack;
-   //
-   // additional members for non-recursive version:
-   //
-   typedef bool (self_type::*unwind_proc_type)(bool);
-
-   void extend_stack();
-   bool unwind(bool);
-   bool unwind_end(bool);
-   bool unwind_paren(bool);
-   bool unwind_recursion_stopper(bool);
-   bool unwind_assertion(bool);
-   bool unwind_alt(bool);
-   bool unwind_repeater_counter(bool);
-   bool unwind_extra_block(bool);
-   bool unwind_greedy_single_repeat(bool);
-   bool unwind_slow_dot_repeat(bool);
-   bool unwind_fast_dot_repeat(bool);
-   bool unwind_char_repeat(bool);
-   bool unwind_short_set_repeat(bool);
-   bool unwind_long_set_repeat(bool);
-   bool unwind_non_greedy_repeat(bool);
-   bool unwind_recursion(bool);
-   bool unwind_recursion_pop(bool);
-   bool unwind_commit(bool);
-   bool unwind_then(bool);
-   bool unwind_case(bool);
-   void destroy_single_repeat();
-   void push_matched_paren(int index, const sub_match<BidiIterator>& sub);
-   void push_recursion_stopper();
-   void push_assertion(const re_syntax_base* ps, bool positive);
-   void push_alt(const re_syntax_base* ps);
-   void push_repeater_count(int i, repeater_count<BidiIterator>** s);
-   void push_single_repeat(std::size_t c, const re_repeat* r, BidiIterator last_position, int state_id);
-   void push_non_greedy_repeat(const re_syntax_base* ps);
-   void push_recursion(int idx, const re_syntax_base* p, results_type* presults, results_type* presults2);
-   void push_recursion_pop();
-   void push_case_change(bool);
-
-   // pointer to base of stack:
-   saved_state* m_stack_base;
-   // pointer to current stack position:
-   saved_state* m_backup_state;
-   // how many memory blocks have we used up?:
-   unsigned used_block_count;
-   // determines what value to return when unwinding from recursion,
-   // allows for mixed recursive/non-recursive algorithm:
-   bool m_recursive_result;
-   // We have unwound to a lookahead/lookbehind, used by COMMIT/PRUNE/SKIP:
-   bool m_unwound_lookahead;
-   // We have unwound to an alternative, used by THEN:
-   bool m_unwound_alt;
-   // We are unwinding a commit - used by independent subs to determine whether to stop there or carry on unwinding:
-   //bool m_unwind_commit;
-   // Recursion limit:
-   unsigned m_recursions;
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(push)
-#if BOOST_REGEX_MSVC >= 1800
-#pragma warning(disable:26495)
-#endif
-#endif
-   // these operations aren't allowed, so are declared private,
-   // bodies are provided to keep explicit-instantiation requests happy:
-   perl_matcher& operator=(const perl_matcher&)
-   {
-      return *this;
-   }
-   perl_matcher(const perl_matcher& that)
-      : m_result(that.m_result), re(that.re), traits_inst(that.traits_inst), rep_obj(0) {}
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(pop)
-#endif
-};
-
-} // namespace BOOST_REGEX_DETAIL_NS
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(pop)
-#endif
-
-} // namespace boost
-
-//
-// include the implementation of perl_matcher:
-//
-#include <boost/regex/v5/perl_matcher_non_recursive.hpp>
-// this one has to be last:
-#include <boost/regex/v5/perl_matcher_common.hpp>
-
-#endif

regex/include/boost/regex/v5/perl_matcher_common.hpp

@@ -1,921 +0,0 @@
-/*
- *
- * Copyright (c) 2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         perl_matcher_common.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Definitions of perl_matcher member functions that are 
-  *                common to both the recursive and non-recursive versions.
-  */
-
-#ifndef BOOST_REGEX_V5_PERL_MATCHER_COMMON_HPP
-#define BOOST_REGEX_V5_PERL_MATCHER_COMMON_HPP
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(push)
-#pragma warning(disable:4459)
-#if BOOST_REGEX_MSVC < 1910
-#pragma warning(disable:4800)
-#endif
-#endif
-
-namespace boost{
-namespace BOOST_REGEX_DETAIL_NS{
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(push)
-#pragma warning(disable:26812)
-#endif
-   template <class BidiIterator, class Allocator, class traits>
-void perl_matcher<BidiIterator, Allocator, traits>::construct_init(const basic_regex<char_type, traits>& e, match_flag_type f)
-{ 
-   typedef typename std::iterator_traits<BidiIterator>::iterator_category category;
-   typedef typename basic_regex<char_type, traits>::flag_type expression_flag_type;
-   
-   if(e.empty())
-   {
-      // precondition failure: e is not a valid regex.
-      std::invalid_argument ex("Invalid regular expression object");
-#ifndef BOOST_REGEX_STANDALONE
-      boost::throw_exception(ex);
-#else
-      throw e;
-#endif
-   }
-   pstate = 0;
-   m_match_flags = f;
-   estimate_max_state_count(static_cast<category*>(0));
-   expression_flag_type re_f = re.flags();
-   icase = re_f & regex_constants::icase;
-   if(!(m_match_flags & (match_perl|match_posix)))
-   {
-      if((re_f & (regbase::main_option_type|regbase::no_perl_ex)) == 0)
-         m_match_flags |= match_perl;
-      else if((re_f & (regbase::main_option_type|regbase::emacs_ex)) == (regbase::basic_syntax_group|regbase::emacs_ex))
-         m_match_flags |= match_perl;
-      else if((re_f & (regbase::main_option_type|regbase::literal)) == (regbase::literal))
-         m_match_flags |= match_perl;
-      else
-         m_match_flags |= match_posix;
-   }
-   if(m_match_flags & match_posix)
-   {
-      m_temp_match.reset(new match_results<BidiIterator, Allocator>());
-      m_presult = m_temp_match.get();
-   }
-   else
-      m_presult = &m_result;
-   m_stack_base = 0;
-   m_backup_state = 0;
-   // find the value to use for matching word boundaries:
-   m_word_mask = re.get_data().m_word_mask; 
-   // find bitmask to use for matching '.':
-   match_any_mask = static_cast<unsigned char>((f & match_not_dot_newline) ? BOOST_REGEX_DETAIL_NS::test_not_newline : BOOST_REGEX_DETAIL_NS::test_newline);
-   // Disable match_any if requested in the state machine:
-   if(e.get_data().m_disable_match_any)
-      m_match_flags &= regex_constants::match_not_any;
-}
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(pop)
-#endif
-
-template <class BidiIterator, class Allocator, class traits>
-void perl_matcher<BidiIterator, Allocator, traits>::estimate_max_state_count(std::random_access_iterator_tag*)
-{
-   //
-   // How many states should we allow our machine to visit before giving up?
-   // This is a heuristic: it takes the greater of O(N^2) and O(NS^2)
-   // where N is the length of the string, and S is the number of states
-   // in the machine.  It's tempting to up this to O(N^2S) or even O(N^2S^2)
-   // but these take unreasonably amounts of time to bale out in pathological
-   // cases.
-   //
-   // Calculate NS^2 first:
-   //
-   static const std::ptrdiff_t k = 100000;
-   std::ptrdiff_t dist = std::distance(base, last);
-   if(dist == 0)
-      dist = 1;
-   std::ptrdiff_t states = re.size();
-   if(states == 0)
-      states = 1;
-   if ((std::numeric_limits<std::ptrdiff_t>::max)() / states < states)
-   {
-      max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits<std::ptrdiff_t>::max)() - 2);
-      return;
-   }
-   states *= states;
-   if((std::numeric_limits<std::ptrdiff_t>::max)() / dist < states)
-   {
-      max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits<std::ptrdiff_t>::max)() - 2);
-      return;
-   }
-   states *= dist;
-   if((std::numeric_limits<std::ptrdiff_t>::max)() - k < states)
-   {
-      max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits<std::ptrdiff_t>::max)() - 2);
-      return;
-   }
-   states += k;
-
-   max_state_count = states;
-
-   //
-   // Now calculate N^2:
-   //
-   states = dist;
-   if((std::numeric_limits<std::ptrdiff_t>::max)() / dist < states)
-   {
-      max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits<std::ptrdiff_t>::max)() - 2);
-      return;
-   }
-   states *= dist;
-   if((std::numeric_limits<std::ptrdiff_t>::max)() - k < states)
-   {
-      max_state_count = (std::min)((std::ptrdiff_t)BOOST_REGEX_MAX_STATE_COUNT, (std::numeric_limits<std::ptrdiff_t>::max)() - 2);
-      return;
-   }
-   states += k;
-   //
-   // N^2 can be a very large number indeed, to prevent things getting out
-   // of control, cap the max states:
-   //
-   if(states > BOOST_REGEX_MAX_STATE_COUNT)
-      states = BOOST_REGEX_MAX_STATE_COUNT;
-   //
-   // If (the possibly capped) N^2 is larger than our first estimate,
-   // use this instead:
-   //
-   if(states > max_state_count)
-      max_state_count = states;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-inline void perl_matcher<BidiIterator, Allocator, traits>::estimate_max_state_count(void*)
-{
-   // we don't know how long the sequence is:
-   max_state_count = BOOST_REGEX_MAX_STATE_COUNT;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-inline bool perl_matcher<BidiIterator, Allocator, traits>::match()
-{
-   return match_imp();
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_imp()
-{
-   // initialise our stack if we are non-recursive:
-   save_state_init init(&m_stack_base, &m_backup_state);
-   used_block_count = BOOST_REGEX_MAX_BLOCKS;
-#if !defined(BOOST_NO_EXCEPTIONS)
-   try{
-#endif
-
-   // reset our state machine:
-   position = base;
-   search_base = base;
-   state_count = 0;
-   m_match_flags |= regex_constants::match_all;
-   m_presult->set_size((m_match_flags & match_nosubs) ? 1u : static_cast<typename results_type::size_type>(1u + re.mark_count()), search_base, last);
-   m_presult->set_base(base);
-   m_presult->set_named_subs(this->re.get_named_subs());
-   if(m_match_flags & match_posix)
-      m_result = *m_presult;
-   verify_options(re.flags(), m_match_flags);
-   if(0 == match_prefix())
-      return false;
-   return (m_result[0].second == last) && (m_result[0].first == base);
-
-#if !defined(BOOST_NO_EXCEPTIONS)
-   }
-   catch(...)
-   {
-      // unwind all pushed states, apart from anything else this
-      // ensures that all the states are correctly destructed
-      // not just the memory freed.
-      while(unwind(true)){}
-      throw;
-   }
-#endif
-}
-
-template <class BidiIterator, class Allocator, class traits>
-inline bool perl_matcher<BidiIterator, Allocator, traits>::find()
-{
-   return find_imp();
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_imp()
-{
-   static matcher_proc_type const s_find_vtable[7] = 
-   {
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_any,
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_word,
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_line,
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_buf,
-      &perl_matcher<BidiIterator, Allocator, traits>::match_prefix,
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_lit,
-      &perl_matcher<BidiIterator, Allocator, traits>::find_restart_lit,
-   };
-
-   // initialise our stack if we are non-recursive:
-   save_state_init init(&m_stack_base, &m_backup_state);
-   used_block_count = BOOST_REGEX_MAX_BLOCKS;
-#if !defined(BOOST_NO_EXCEPTIONS)
-   try{
-#endif
-
-   state_count = 0;
-   if((m_match_flags & regex_constants::match_init) == 0)
-   {
-      // reset our state machine:
-      search_base = position = base;
-      pstate = re.get_first_state();
-      m_presult->set_size((m_match_flags & match_nosubs) ? 1u : static_cast<typename results_type::size_type>(1u + re.mark_count()), base, last);
-      m_presult->set_base(base);
-      m_presult->set_named_subs(this->re.get_named_subs());
-      m_match_flags |= regex_constants::match_init;
-   }
-   else
-   {
-      // start again:
-      search_base = position = m_result[0].second;
-      // If last match was null and match_not_null was not set then increment
-      // our start position, otherwise we go into an infinite loop:
-      if(((m_match_flags & match_not_null) == 0) && (m_result.length() == 0))
-      {
-         if(position == last)
-            return false;
-         else 
-            ++position;
-      }
-      // reset $` start:
-      m_presult->set_size((m_match_flags & match_nosubs) ? 1u : static_cast<typename results_type::size_type>(1u + re.mark_count()), search_base, last);
-      //if((base != search_base) && (base == backstop))
-      //   m_match_flags |= match_prev_avail;
-   }
-   if(m_match_flags & match_posix)
-   {
-      m_result.set_size(static_cast<typename results_type::size_type>(1u + re.mark_count()), base, last);
-      m_result.set_base(base);
-   }
-
-   verify_options(re.flags(), m_match_flags);
-   // find out what kind of expression we have:
-   unsigned type = (m_match_flags & match_continuous) ? 
-      static_cast<unsigned int>(regbase::restart_continue) 
-         : static_cast<unsigned int>(re.get_restart_type());
-
-   // call the appropriate search routine:
-   matcher_proc_type proc = s_find_vtable[type];
-   return (this->*proc)();
-
-#if !defined(BOOST_NO_EXCEPTIONS)
-   }
-   catch(...)
-   {
-      // unwind all pushed states, apart from anything else this
-      // ensures that all the states are correctly destructed
-      // not just the memory freed.
-      while(unwind(true)){}
-      throw;
-   }
-#endif
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_prefix()
-{
-   m_has_partial_match = false;
-   m_has_found_match = false;
-   pstate = re.get_first_state();
-   m_presult->set_first(position);
-   restart = position;
-   match_all_states();
-   if(!m_has_found_match && m_has_partial_match && (m_match_flags & match_partial))
-   {
-      m_has_found_match = true;
-      m_presult->set_second(last, 0, false);
-      position = last;
-      if((m_match_flags & match_posix) == match_posix)
-      {
-         m_result.maybe_assign(*m_presult);
-      }
-   }
-#ifdef BOOST_REGEX_MATCH_EXTRA
-   if(m_has_found_match && (match_extra & m_match_flags))
-   {
-      //
-      // we have a match, reverse the capture information:
-      //
-      for(unsigned i = 0; i < m_presult->size(); ++i)
-      {
-         typename sub_match<BidiIterator>::capture_sequence_type & seq = ((*m_presult)[i]).get_captures();
-         std::reverse(seq.begin(), seq.end());
-      }
-   }
-#endif
-   if(!m_has_found_match)
-      position = restart; // reset search postion
-   return m_has_found_match;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_literal()
-{
-   unsigned int len = static_cast<const re_literal*>(pstate)->length;
-   const char_type* what = reinterpret_cast<const char_type*>(static_cast<const re_literal*>(pstate) + 1);
-   //
-   // compare string with what we stored in
-   // our records:
-   for(unsigned int i = 0; i < len; ++i, ++position)
-   {
-      if((position == last) || (traits_inst.translate(*position, icase) != what[i]))
-         return false;
-   }
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_start_line()
-{
-   if(position == backstop)
-   {
-      if((m_match_flags & match_prev_avail) == 0)
-      {
-         if((m_match_flags & match_not_bol) == 0)
-         {
-            pstate = pstate->next.p;
-            return true;
-         }
-         return false;
-      }
-   }
-   else if(m_match_flags & match_single_line)
-      return false;
-
-   // check the previous value character:
-   BidiIterator t(position);
-   --t;
-   if(position != last)
-   {
-      if(is_separator(*t) && !((*t == static_cast<char_type>('\r')) && (*position == static_cast<char_type>('\n'))) )
-      {
-         pstate = pstate->next.p;
-         return true;
-      }
-   }
-   else if(is_separator(*t))
-   {
-      pstate = pstate->next.p;
-      return true;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_end_line()
-{
-   if(position != last)
-   {
-      if(m_match_flags & match_single_line)
-         return false;
-      // we're not yet at the end so *first is always valid:
-      if(is_separator(*position))
-      {
-         if((position != backstop) || (m_match_flags & match_prev_avail))
-         {
-            // check that we're not in the middle of \r\n sequence
-            BidiIterator t(position);
-            --t;
-            if((*t == static_cast<char_type>('\r')) && (*position == static_cast<char_type>('\n')))
-            {
-               return false;
-            }
-         }
-         pstate = pstate->next.p;
-         return true;
-      }
-   }
-   else if((m_match_flags & match_not_eol) == 0)
-   {
-      pstate = pstate->next.p;
-      return true;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_wild()
-{
-   if(position == last) 
-      return false;
-   if(is_separator(*position) && ((match_any_mask & static_cast<const re_dot*>(pstate)->mask) == 0))
-      return false;
-   if((*position == char_type(0)) && (m_match_flags & match_not_dot_null))
-      return false;
-   pstate = pstate->next.p;
-   ++position;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_word_boundary()
-{
-   bool b; // indcates whether next character is a word character
-   if(position != last)
-   {
-      // prev and this character must be opposites:
-      b = traits_inst.isctype(*position, m_word_mask);
-   }
-   else
-   {
-      if (m_match_flags & match_not_eow)
-         return false;
-      b = false;
-   }
-   if((position == backstop) && ((m_match_flags & match_prev_avail) == 0))
-   {
-      if(m_match_flags & match_not_bow)
-         return false;
-      else
-         b ^= false;
-   }
-   else
-   {
-      --position;
-      b ^= traits_inst.isctype(*position, m_word_mask);
-      ++position;
-   }
-   if(b)
-   {
-      pstate = pstate->next.p;
-      return true;
-   }
-   return false; // no match if we get to here...
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_within_word()
-{
-   bool b = !match_word_boundary();
-   if(b)
-      pstate = pstate->next.p;
-   return b;
-   /*
-   if(position == last)
-      return false;
-   // both prev and this character must be m_word_mask:
-   bool prev = traits_inst.isctype(*position, m_word_mask);
-   {
-      bool b;
-      if((position == backstop) && ((m_match_flags & match_prev_avail) == 0)) 
-         return false;
-      else
-      {
-         --position;
-         b = traits_inst.isctype(*position, m_word_mask);
-         ++position;
-      }
-      if(b == prev)
-      {
-         pstate = pstate->next.p;
-         return true;
-      }
-   }
-   return false;
-   */
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_word_start()
-{
-   if(position == last)
-      return false; // can't be starting a word if we're already at the end of input
-   if(!traits_inst.isctype(*position, m_word_mask))
-      return false; // next character isn't a word character
-   if((position == backstop) && ((m_match_flags & match_prev_avail) == 0))
-   {
-      if(m_match_flags & match_not_bow)
-         return false; // no previous input
-   }
-   else
-   {
-      // otherwise inside buffer:
-      BidiIterator t(position);
-      --t;
-      if(traits_inst.isctype(*t, m_word_mask))
-         return false; // previous character not non-word
-   }
-   // OK we have a match:
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_word_end()
-{
-   if((position == backstop) && ((m_match_flags & match_prev_avail) == 0))
-      return false;  // start of buffer can't be end of word
-   BidiIterator t(position);
-   --t;
-   if(traits_inst.isctype(*t, m_word_mask) == false)
-      return false;  // previous character wasn't a word character
-
-   if(position == last)
-   {
-      if(m_match_flags & match_not_eow)
-         return false; // end of buffer but not end of word
-   }
-   else
-   {
-      // otherwise inside buffer:
-      if(traits_inst.isctype(*position, m_word_mask))
-         return false; // next character is a word character
-   }
-   pstate = pstate->next.p;
-   return true;      // if we fall through to here then we've succeeded
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_buffer_start()
-{
-   if((position != backstop) || (m_match_flags & match_not_bob))
-      return false;
-   // OK match:
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_buffer_end()
-{
-   if((position != last) || (m_match_flags & match_not_eob))
-      return false;
-   // OK match:
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_backref()
-{
-   //
-   // Compare with what we previously matched.
-   // Note that this succeeds if the backref did not partisipate
-   // in the match, this is in line with ECMAScript, but not Perl
-   // or PCRE.
-   //
-   int index = static_cast<const re_brace*>(pstate)->index;
-   if(index >= hash_value_mask)
-   {
-      named_subexpressions::range_type r = re.get_data().equal_range(index);
-      BOOST_REGEX_ASSERT(r.first != r.second);
-      do
-      {
-         index = r.first->index;
-         ++r.first;
-      }while((r.first != r.second) && ((*m_presult)[index].matched != true));
-   }
-
-   if((m_match_flags & match_perl) && !(*m_presult)[index].matched)
-      return false;
-
-   BidiIterator i = (*m_presult)[index].first;
-   BidiIterator j = (*m_presult)[index].second;
-   while(i != j)
-   {
-      if((position == last) || (traits_inst.translate(*position, icase) != traits_inst.translate(*i, icase)))
-         return false;
-      ++i;
-      ++position;
-   }
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_long_set()
-{
-   typedef typename traits::char_class_type char_class_type;
-   // let the traits class do the work:
-   if(position == last)
-      return false;
-   BidiIterator t = re_is_set_member(position, last, static_cast<const re_set_long<char_class_type>*>(pstate), re.get_data(), icase);
-   if(t != position)
-   {
-      pstate = pstate->next.p;
-      position = t;
-      return true;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_set()
-{
-   if(position == last)
-      return false;
-   if(static_cast<const re_set*>(pstate)->_map[static_cast<unsigned char>(traits_inst.translate(*position, icase))])
-   {
-      pstate = pstate->next.p;
-      ++position;
-      return true;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_jump()
-{
-   pstate = static_cast<const re_jump*>(pstate)->alt.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_combining()
-{
-   if(position == last)
-      return false;
-   if(is_combining(traits_inst.translate(*position, icase)))
-      return false;
-   ++position;
-   while((position != last) && is_combining(traits_inst.translate(*position, icase)))
-      ++position;
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_soft_buffer_end()
-{
-   if(m_match_flags & match_not_eob)
-      return false;
-   BidiIterator p(position);
-   while((p != last) && is_separator(traits_inst.translate(*p, icase)))++p;
-   if(p != last)
-      return false;
-   pstate = pstate->next.p;
-   return true;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_restart_continue()
-{
-   if(position == search_base)
-   {
-      pstate = pstate->next.p;
-      return true;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_backstep()
-{
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable:4127)
-#endif
-   if( ::boost::is_random_access_iterator<BidiIterator>::value)
-   {
-      std::ptrdiff_t maxlen = std::distance(backstop, position);
-      if(maxlen < static_cast<const re_brace*>(pstate)->index)
-         return false;
-      std::advance(position, -static_cast<const re_brace*>(pstate)->index);
-   }
-   else
-   {
-      int c = static_cast<const re_brace*>(pstate)->index;
-      while(c--)
-      {
-         if(position == backstop)
-            return false;
-         --position;
-      }
-   }
-   pstate = pstate->next.p;
-   return true;
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-}
-
-template <class BidiIterator, class Allocator, class traits>
-inline bool perl_matcher<BidiIterator, Allocator, traits>::match_assert_backref()
-{
-   // return true if marked sub-expression N has been matched:
-   int index = static_cast<const re_brace*>(pstate)->index;
-   bool result = false;
-   if(index == 9999)
-   {
-      // Magic value for a (DEFINE) block:
-      return false;
-   }
-   else if(index > 0)
-   {
-      // Have we matched subexpression "index"?
-      // Check if index is a hash value:
-      if(index >= hash_value_mask)
-      {
-         named_subexpressions::range_type r = re.get_data().equal_range(index);
-         while(r.first != r.second)
-         {
-            if((*m_presult)[r.first->index].matched)
-            {
-               result = true;
-               break;
-            }
-            ++r.first;
-         }
-      }
-      else
-      {
-         result = (*m_presult)[index].matched;
-      }
-      pstate = pstate->next.p;
-   }
-   else
-   {
-      // Have we recursed into subexpression "index"?
-      // If index == 0 then check for any recursion at all, otherwise for recursion to -index-1.
-      int idx = -(index+1);
-      if(idx >= hash_value_mask)
-      {
-         named_subexpressions::range_type r = re.get_data().equal_range(idx);
-         int stack_index = recursion_stack.empty() ? -1 : recursion_stack.back().idx;
-         while(r.first != r.second)
-         {
-            result |= (stack_index == r.first->index);
-            if(result)break;
-            ++r.first;
-         }
-      }
-      else
-      {
-         result = !recursion_stack.empty() && ((recursion_stack.back().idx == idx) || (index == 0));
-      }
-      pstate = pstate->next.p;
-   }
-   return result;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_fail()
-{
-   // Just force a backtrack:
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::match_accept()
-{
-   if(!recursion_stack.empty())
-   {
-      return skip_until_paren(recursion_stack.back().idx);
-   }
-   else
-   {
-      return skip_until_paren(INT_MAX);
-   }
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_restart_any()
-{
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable:4127)
-#endif
-   const unsigned char* _map = re.get_map();
-   while(true)
-   {
-      // skip everything we can't match:
-      while((position != last) && !can_start(*position, _map, (unsigned char)mask_any) )
-         ++position;
-      if(position == last)
-      {
-         // run out of characters, try a null match if possible:
-         if(re.can_be_null())
-            return match_prefix();
-         break;
-      }
-      // now try and obtain a match:
-      if(match_prefix())
-         return true;
-      if(position == last)
-         return false;
-      ++position;
-   }
-   return false;
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_restart_word()
-{
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(push)
-#pragma warning(disable:4127)
-#endif
-   // do search optimised for word starts:
-   const unsigned char* _map = re.get_map();
-   if((m_match_flags & match_prev_avail) || (position != base))
-      --position;
-   else if(match_prefix())
-      return true;
-   do
-   {
-      while((position != last) && traits_inst.isctype(*position, m_word_mask))
-         ++position;
-      while((position != last) && !traits_inst.isctype(*position, m_word_mask))
-         ++position;
-      if(position == last)
-         break;
-
-      if(can_start(*position, _map, (unsigned char)mask_any) )
-      {
-         if(match_prefix())
-            return true;
-      }
-      if(position == last)
-         break;
-   } while(true);
-   return false;
-#ifdef BOOST_REGEX_MSVC
-#pragma warning(pop)
-#endif
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_restart_line()
-{
-   // do search optimised for line starts:
-   const unsigned char* _map = re.get_map();
-   if(match_prefix())
-      return true;
-   while(position != last)
-   {
-      while((position != last) && !is_separator(*position))
-         ++position;
-      if(position == last)
-         return false;
-      ++position;
-      if(position == last)
-      {
-         if(re.can_be_null() && match_prefix())
-            return true;
-         return false;
-      }
-
-      if( can_start(*position, _map, (unsigned char)mask_any) )
-      {
-         if(match_prefix())
-            return true;
-      }
-      if(position == last)
-         return false;
-      //++position;
-   }
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_restart_buf()
-{
-   if((position == base) && ((m_match_flags & match_not_bob) == 0))
-      return match_prefix();
-   return false;
-}
-
-template <class BidiIterator, class Allocator, class traits>
-bool perl_matcher<BidiIterator, Allocator, traits>::find_restart_lit()
-{
-   return false;
-}
-
-} // namespace BOOST_REGEX_DETAIL_NS
-
-} // namespace boost
-
-#ifdef BOOST_REGEX_MSVC
-#  pragma warning(pop)
-#endif
-
-#endif
-

regex/include/boost/regex/v5/primary_transform.hpp

@@ -1,120 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
- 
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE:        primary_transform.hpp
-  *   VERSION:     see <boost/version.hpp>
-  *   DESCRIPTION: Heuristically determines the sort string format in use
-  *                by the current locale.
-  */
-
-#ifndef BOOST_REGEX_PRIMARY_TRANSFORM
-#define BOOST_REGEX_PRIMARY_TRANSFORM
-
-namespace boost{
-   namespace BOOST_REGEX_DETAIL_NS{
-
-
-enum{
-   sort_C,
-   sort_fixed,
-   sort_delim,
-   sort_unknown
-};
-
-template <class S, class charT>
-unsigned count_chars(const S& s, charT c)
-{
-   //
-   // Count how many occurrences of character c occur
-   // in string s: if c is a delimeter between collation
-   // fields, then this should be the same value for all
-   // sort keys:
-   //
-   unsigned int count = 0;
-   for(unsigned pos = 0; pos < s.size(); ++pos)
-   {
-      if(s[pos] == c) ++count;
-   }
-   return count;
-}
-
-
-template <class traits, class charT>
-unsigned find_sort_syntax(const traits* pt, charT* delim)
-{
-   //
-   // compare 'a' with 'A' to see how similar they are,
-   // should really use a-accute but we can't portably do that,
-   //
-   typedef typename traits::string_type string_type;
-   typedef typename traits::char_type char_type;
-
-   // Suppress incorrect warning for MSVC
-   (void)pt;
-
-   char_type a[2] = {'a', '\0', };
-   string_type sa(pt->transform(a, a+1));
-   if(sa == a)
-   {
-      *delim = 0;
-      return sort_C;
-   }
-   char_type A[2] = { 'A', '\0', };
-   string_type sA(pt->transform(A, A+1));
-   char_type c[2] = { ';', '\0', };
-   string_type sc(pt->transform(c, c+1));
-
-   int pos = 0;
-   while((pos <= static_cast<int>(sa.size())) && (pos <= static_cast<int>(sA.size())) && (sa[pos] == sA[pos])) ++pos;
-   --pos;
-   if(pos < 0)
-   {
-      *delim = 0;
-      return sort_unknown;
-   }
-   //
-   // at this point sa[pos] is either the end of a fixed width field
-   // or the character that acts as a delimiter:
-   //
-   charT maybe_delim = sa[pos];
-   if((pos != 0) && (count_chars(sa, maybe_delim) == count_chars(sA, maybe_delim)) && (count_chars(sa, maybe_delim) == count_chars(sc, maybe_delim)))
-   {
-      *delim = maybe_delim;
-      return sort_delim;
-   }
-   //
-   // OK doen't look like a delimiter, try for fixed width field:
-   //
-   if((sa.size() == sA.size()) && (sa.size() == sc.size()))
-   {
-      // note assumes that the fixed width field is less than
-      // (numeric_limits<charT>::max)(), should be true for all types
-      // I can't imagine 127 character fields...
-      *delim = static_cast<charT>(++pos);
-      return sort_fixed;
-   }
-   //
-   // don't know what it is:
-   //
-   *delim = 0;
-   return sort_unknown;
-}
-
-
-   } // namespace BOOST_REGEX_DETAIL_NS
-} // namespace boost
-
-#endif
-
-
-

regex/include/boost/regex/v5/regbase.hpp

@@ -1,158 +0,0 @@
-/*
- *
- * Copyright (c) 1998-2002
- * John Maddock
- *
- * Use, modification and distribution are subject to the 
- * Boost Software License, Version 1.0. (See accompanying file 
- * LICENSE_1_0.txt or copy at http://www.boost.org/LICENSE_1_0.txt)
- *
- */
-
- /*
-  *   LOCATION:    see http://www.boost.org for most recent version.
-  *   FILE         regbase.cpp
-  *   VERSION      see <boost/version.hpp>
-  *   DESCRIPTION: Declares class regbase.
-  */
-
-#ifndef BOOST_REGEX_V5_REGBASE_HPP
-#define BOOST_REGEX_V5_REGBASE_HPP
-
-namespace boost{
-//
-// class regbase
-// handles error codes and flags
-//
-class regbase
-{
-public:
-   enum flag_type_
-   {
-      //
-      // Divide the flags up into logical groups:
-      // bits 0-7 indicate main synatx type.
-      // bits 8-15 indicate syntax subtype.
-      // bits 16-31 indicate options that are common to all
-      // regex syntaxes.
-      // In all cases the default is 0.
-      //
-      // Main synatx group:
-      //
-      perl_syntax_group = 0,                      // default
-      basic_syntax_group = 1,                     // POSIX basic
-      literal = 2,                                // all characters are literals
-      main_option_type = literal | basic_syntax_group | perl_syntax_group, // everything!
-      //
-      // options specific to perl group:
-      //
-      no_bk_refs = 1 << 8,                        // \d not allowed
-      no_perl_ex = 1 << 9,                        // disable perl extensions
-      no_mod_m = 1 << 10,                         // disable Perl m modifier
-      mod_x = 1 << 11,                            // Perl x modifier
-      mod_s = 1 << 12,                            // force s modifier on (overrides match_not_dot_newline)
-      no_mod_s = 1 << 13,                         // force s modifier off (overrides match_not_dot_newline)
-
-      //
-      // options specific to basic group:
-      //
-      no_char_classes = 1 << 8,                   // [[:CLASS:]] not allowed
-      no_intervals = 1 << 9,                      // {x,y} not allowed
-      bk_plus_qm = 1 << 10,                       // uses \+ and \?
-      bk_vbar = 1 << 11,                          // use \| for alternatives
-      emacs_ex = 1 << 12,                         // enables emacs extensions
-
-      //
-      // options common to all groups:
-      //
-      no_escape_in_lists = 1 << 16,                     // '\' not special inside [...]
-      newline_alt = 1 << 17,                            // \n is the same as |
-      no_except = 1 << 18,                              // no exception on error
-      failbit = 1 << 19,                                // error flag
-      icase = 1 << 20,                                  // characters are matched regardless of case
-      nocollate = 0,                                    // don't use locale specific collation (deprecated)
-      collate = 1 << 21,                                // use locale specific collation
-      nosubs = 1 << 22,                                 // don't mark sub-expressions
-      save_subexpression_location = 1 << 23,            // save subexpression locations
-      no_empty_expressions = 1 << 24,                   // no empty expressions allowed
-      optimize = 0,                                     // not really supported
-      
-
-
-      basic = basic_syntax_group | collate | no_escape_in_lists,
-      extended = no_bk_refs | collate | no_perl_ex | no_escape_in_lists,
-      normal = 0,
-      emacs = basic_syntax_group | collate | emacs_ex | bk_vbar,
-      awk = no_bk_refs | collate | no_perl_ex,
-      grep = basic | newline_alt,
-      egrep = extended | newline_alt,
-      sed = basic,
-      perl = normal,
-      ECMAScript = normal,
-      JavaScript = normal,
-      JScript = normal
-   };
-   typedef unsigned int flag_type;
-
-   enum restart_info
-   {
-      restart_any = 0,
-      restart_word = 1,
-      restart_line = 2,
-      restart_buf = 3,
-      restart_continue = 4,
-      restart_lit = 5,
-      restart_fixed_lit = 6, 
-      restart_count = 7
-   };
-};
-
-//
-// provide std lib proposal compatible constants:
-//
-namespace regex_constants{
-
-   enum flag_type_
-   {
-
-      no_except = ::boost::regbase::no_except,
-      failbit = ::boost::regbase::failbit,
-      literal = ::boost::regbase::literal,
-      icase = ::boost::regbase::icase,
-      nocollate = ::boost::regbase::nocollate,
-      collate = ::boost::regbase::collate,
-      nosubs = ::boost::regbase::nosubs,
-      optimize = ::boost::regbase::optimize,
-      bk_plus_qm = ::boost::regbase::bk_plus_qm,
-      bk_vbar = ::boost::regbase::bk_vbar,
-      no_intervals = ::boost::regbase::no_intervals,
-      no_char_classes = ::boost::regbase::no_char_classes,
-      no_escape_in_lists = ::boost::regbase::no_escape_in_lists,
-      no_mod_m = ::boost::regbase::no_mod_m,
-      mod_x = ::boost::regbase::mod_x,
-      mod_s = ::boost::regbase::mod_s,
-      no_mod_s = ::boost::regbase::no_mod_s,
-      save_subexpression_location = ::boost::regbase::save_subexpression_location,
-      no_empty_expressions = ::boost::regbase::no_empty_expressions,
-
-      basic = ::boost::regbase::basic,
-      extended = ::boost::regbase::extended,
-      normal = ::boost::regbase::normal,
-      emacs = ::boost::regbase::emacs,
-      awk = ::boost::regbase::awk,
-      grep = ::boost::regbase::grep,
-      egrep = ::boost::regbase::egrep,
-      sed = basic,
-      perl = normal,
-      ECMAScript = normal,
-      JavaScript = normal,
-      JScript = normal
-   };
-   typedef ::boost::regbase::flag_type syntax_option_type;
-
-} // namespace regex_constants
-
-} // namespace boost
-
-#endif
-