Merge remote-tracking branch 'origin/develop' into trunk

* Proposed fix for comparing floating points (needed on macOS)

* Work around weird behaviour in testing floats using catch2

* Better from_chars implementation

CMakeLists.txt

@@ -22,16 +22,17 @@
 # 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(
 	libcifpp
-	VERSION 7.0.0
+	VERSION 7.0.5
 	LANGUAGES CXX)
 
 list(PREPEND CMAKE_MODULE_PATH "${CMAKE_CURRENT_SOURCE_DIR}/cmake")
 
+include(FindAtomic)
 include(CheckFunctionExists)
 include(CheckIncludeFiles)
 include(CheckLibraryExists)
@@ -40,11 +41,6 @@ include(CheckCXXSourceCompiles)
 include(GenerateExportHeader)
 include(CTest)
 include(FetchContent)
-include(ExternalProject)
-
-set(CXX_EXTENSIONS OFF)
-set(CMAKE_CXX_STANDARD 20)
-set(CMAKE_CXX_STANDARD_REQUIRED ON)
 
 # When building with ninja-multiconfig, build both debug and release by default
 if(CMAKE_GENERATOR STREQUAL "Ninja Multi-Config")
@@ -71,10 +67,7 @@ if(NOT(BUILD_FOR_CCP4 AND WIN32))
 	option(BUILD_SHARED_LIBS "Build a shared library instead of a static one" OFF)
 endif()
 
-if(BUILD_FOR_CCP4)
-	unset(CIFPP_DOWNLOAD_CCD)
-	unset(CIFPP_INSTALL_UPDATE_SCRIPT)
-else()
+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)
@@ -84,6 +77,9 @@ else()
 		option(CIFPP_INSTALL_UPDATE_SCRIPT
 			"Install the script to update CCD and dictionary files" ON)
 	endif()
+else()
+	unset(CIFPP_DOWNLOAD_CCD)
+	unset(CIFPP_INSTALL_UPDATE_SCRIPT)
 endif()
 
 # When CCP4 is sourced in the environment, we can recreate the symmetry
@@ -126,6 +122,7 @@ if(WIN32)
 		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...
@@ -137,6 +134,7 @@ if(MSVC)
 	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()
@@ -164,7 +162,7 @@ if(GXX_LIBSTDCPP)
 
 	try_run(STD_REGEX_RUNNING STD_REGEX_COMPILING
 		${CMAKE_CURRENT_BINARY_DIR}/test
-		${PROJECT_SOURCE_DIR}/cmake/test-rx.cpp)
+		${CMAKE_CURRENT_SOURCE_DIR}/cmake/test-rx.cpp)
 
 	if(STD_REGEX_RUNNING STREQUAL FAILED_TO_RUN)
 		message(
@@ -189,10 +187,6 @@ if(GXX_LIBSTDCPP)
 	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.
@@ -221,7 +215,12 @@ if(MSVC)
 	endforeach()
 endif()
 
-find_package(ZLIB REQUIRED)
+find_package(ZLIB QUIET)
+find_package(Threads)
+
+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
@@ -247,96 +246,109 @@ else()
 	set(EIGEN_INCLUDE_DIR ${my-eigen3_SOURCE_DIR})
 endif()
 
-include(FindFilesystem)
-list(APPEND CIFPP_REQUIRED_LIBRARIES ${STDCPPFS_LIBRARY})
-
-include(FindAtomic)
-list(APPEND CIFPP_REQUIRED_LIBRARIES ${STDCPPATOMIC_LIBRARY})
-
 # Create a revision file, containing the current git version info
 include(VersionString)
-write_version_header(${PROJECT_SOURCE_DIR}/src/ LIB_NAME "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
-		"${PROJECT_SOURCE_DIR}/src/symop-map-generator.cpp")
+		"${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
+		OUTPUT ${CMAKE_CURRENT_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)
+		$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
+		${CMAKE_CURRENT_SOURCE_DIR}/src/symop_table_data.hpp
 		DEPENDS symop-map-generator "$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/pdb_record.hpp
-	${PROJECT_SOURCE_DIR}/src/pdb/pdb2cif_remark_3.hpp
-	${PROJECT_SOURCE_DIR}/src/pdb/pdb2cif_remark_3.cpp
-	${PROJECT_SOURCE_DIR}/src/pdb/reconstruct.cpp
-	${PROJECT_SOURCE_DIR}/src/pdb/validate-pdbx.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.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/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)
 
+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
-	${PROJECT_SOURCE_DIR}/include/cif++/exports.hpp)
+	${CMAKE_CURRENT_SOURCE_DIR}/include/cif++/exports.hpp)
 
 if(BOOST_REGEX)
 	target_compile_definitions(cifpp PRIVATE USE_BOOST_REGEX=1
@@ -353,12 +365,11 @@ set_target_properties(cifpp PROPERTIES POSITION_INDEPENDENT_CODE ON)
 
 target_include_directories(
 	cifpp
-	PUBLIC "$<BUILD_INTERFACE:${PROJECT_SOURCE_DIR}/include>"
+	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
-	${CIFPP_REQUIRED_LIBRARIES})
+target_link_libraries(cifpp PUBLIC Threads::Threads ZLIB::ZLIB std::atomic)
 
 if(CMAKE_CXX_COMPILER_ID STREQUAL "AppleClang")
 	target_link_options(cifpp PRIVATE -undefined dynamic_lookup)
@@ -366,7 +377,7 @@ endif(CMAKE_CXX_COMPILER_ID STREQUAL "AppleClang")
 
 if(CIFPP_DOWNLOAD_CCD)
 	# download the components.cif file from CCD
-	set(COMPONENTS_CIF ${PROJECT_SOURCE_DIR}/rsrc/components.cif)
+	set(COMPONENTS_CIF ${CMAKE_CURRENT_SOURCE_DIR}/rsrc/components.cif)
 
 	if(EXISTS ${COMPONENTS_CIF})
 		file(SIZE ${COMPONENTS_CIF} CCD_FILE_SIZE)
@@ -401,7 +412,7 @@ if(CIFPP_DOWNLOAD_CCD)
 			add_custom_command(
 				OUTPUT ${COMPONENTS_CIF}
 				COMMAND "${GUNZIP}" ${COMPONENTS_CIF}.gz
-				WORKING_DIRECTORY ${PROJECT_SOURCE_DIR}/rsrc/)
+				WORKING_DIRECTORY ${CMAKE_CURRENT_SOURCE_DIR}/rsrc/)
 
 			add_custom_target(COMPONENTS ALL DEPENDS ${COMPONENTS_CIF})
 		endif()
@@ -431,6 +442,10 @@ if(CIFPP_DATA_DIR)
 	target_compile_definitions(cifpp PUBLIC DATA_DIR="${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
@@ -452,14 +467,9 @@ else()
 endif()
 
 # Install rules
-install(
-	TARGETS cifpp
-	EXPORT cifpp-targets
-	ARCHIVE DESTINATION ${CMAKE_INSTALL_LIBDIR}
-	LIBRARY DESTINATION ${CMAKE_INSTALL_LIBDIR}
-	RUNTIME DESTINATION ${CMAKE_INSTALL_BINDIR}
-	INCLUDES
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
+install(TARGETS cifpp
+	EXPORT cifpp
+	FILE_SET cifpp_headers DESTINATION ${CMAKE_INSTALL_INCLUDEDIR})
 
 if(MSVC AND BUILD_SHARED_LIBS)
 	install(
@@ -479,50 +489,33 @@ if(OLD_CONFIG_FILES)
 	install(CODE "file(REMOVE ${OLD_CONFIG_FILES})")
 endif()
 
-install(
-	EXPORT cifpp-targets
-	FILE "cifpp-targets.cmake"
+install(EXPORT cifpp
 	NAMESPACE cifpp::
-	DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp)
+	FILE "cifpp-targets.cmake"
+	DESTINATION lib/cmake/cifpp)
 
 install(
-	DIRECTORY include/cif++
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
-	COMPONENT Devel)
+	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)
 
-install(
-	FILES include/cif++.hpp
-	DESTINATION ${CMAKE_INSTALL_INCLUDEDIR}
-	COMPONENT Devel)
-
-if(CIFPP_DATA_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_DATA_DIR})
+if(CIFPP_DATA_DIR AND CIFPP_DOWNLOAD_CCD)
+	install(FILES ${COMPONENTS_CIF}
+		DESTINATION ${CMAKE_INSTALL_DATADIR}/libcifpp)
 endif()
 
-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})
-endif()
-
-set(CONFIG_TEMPLATE_FILE ${PROJECT_SOURCE_DIR}/cmake/cifpp-config.cmake.in)
+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/cifpp-config.cmake
-	INSTALL_DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp
+	INSTALL_DESTINATION lib/cmake/cifpp
 	PATH_VARS CIFPP_DATA_DIR)
 
 install(
 	FILES "${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config.cmake"
 	"${CMAKE_CURRENT_BINARY_DIR}/cifpp/cifpp-config-version.cmake"
-	DESTINATION ${CMAKE_INSTALL_LIBDIR}/cmake/cifpp
-	COMPONENT Devel)
+	DESTINATION lib/cmake/cifpp)
 
 set_target_properties(
 	cifpp
@@ -540,109 +533,42 @@ write_basic_package_version_file(
 	VERSION ${PROJECT_VERSION}
 	COMPATIBILITY AnyNewerVersion)
 
-# In case we're included as sub_directory:
-if(NOT PROJECT_IS_TOP_LEVEL)
-	set(CIFPP_SHARE_DIR ${PROJECT_SOURCE_DIR}/rsrc PARENT_SCOPE)
-endif()
-
 if(BUILD_TESTING)
-	# We're using the older version 2 of Catch2
-	FetchContent_Declare(
-		Catch2
-		GIT_REPOSITORY https://github.com/catchorg/Catch2.git
-		GIT_TAG v2.13.9)
-
-	FetchContent_MakeAvailable(Catch2)
-
-	list(
-		APPEND
-		CIFPP_tests
-		unit-v2
-		unit-3d
-		format
-		model
-		rename-compound
-		sugar
-		spinner
-		reconstruction
-		validate-pdbx)
-
-	foreach(CIFPP_TEST IN LISTS CIFPP_tests)
-		set(CIFPP_TEST "${CIFPP_TEST}-test")
-		set(CIFPP_TEST_SOURCE "${CMAKE_CURRENT_SOURCE_DIR}/test/${CIFPP_TEST}.cpp")
-
-		add_executable(
-			${CIFPP_TEST} ${CIFPP_TEST_SOURCE}
-			"${CMAKE_CURRENT_SOURCE_DIR}/test/test-main.cpp")
-
-		target_link_libraries(${CIFPP_TEST} PRIVATE Threads::Threads cifpp::cifpp
-			Catch2::Catch2)
-		target_include_directories(${CIFPP_TEST} PRIVATE "${EIGEN_INCLUDE_DIR}")
-
-		if(MSVC)
-			# Specify unwind semantics so that MSVC knowns how to handle exceptions
-			target_compile_options(${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}> --data-dir
-			${CMAKE_CURRENT_SOURCE_DIR}/test)
-
-		add_test(NAME ${CIFPP_TEST} COMMAND $<TARGET_FILE:${CIFPP_TEST}> --data-dir
-			${CMAKE_CURRENT_SOURCE_DIR}/test)
-	endforeach()
+	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 ${CMAKE_SYSTEM_NAME} STREQUAL
-		"GNU")
-		if("${CMAKE_INSTALL_PREFIX}" STREQUAL "/usr/local")
-			set(CIFPP_CRON_DIR
-				"/etc/cron.weekly"
-				CACHE PATH "The cron directory, for the update script")
-		else()
-			set(CIFPP_CRON_DIR
-				"${CIFPP_ETC_DIR}/cron.weekly"
-				CACHE PATH "The cron directory, for the update script")
-		endif()
-	elseif(${CMAKE_SYSTEM_NAME} STREQUAL "FreeBSD")
-		set(CIFPP_CRON_DIR
-			"${CIFPP_ETC_DIR}/periodic/weekly"
-			CACHE PATH "The cron directory, for the update script")
+	configure_file(${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 ${CIFPP_CACHE_DIR})
-
 	# a config file, to make it complete
-	if(NOT EXISTS "${CIFPP_ETC_DIR}/libcifpp.conf")
+	# 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}")
+			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 "${CIFPP_ETC_DIR}/libcifpp/cache-update.d")
+		install(DIRECTORY DESTINATION ${CMAKE_INSTALL_SYSCONFDIR}/libcifpp/cache-update.d)
 	endif()
 
 	target_compile_definitions(cifpp PUBLIC CACHE_DIR="${CIFPP_CACHE_DIR}")
@@ -651,13 +577,3 @@ endif()
 if(BUILD_DOCUMENTATION)
 	add_subdirectory(docs)
 endif()
-
-set(CPACK_RESOURCE_FILE_LICENSE "${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 "/rsrc/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)

README.md

@@ -52,7 +52,7 @@ int main(int argc, char *argv[])
 
     if (file.empty())
     {
-        std::cerr << "Empty file" << std::endl;
+        std::cerr << "Empty file\n";
         exit(1);
     }
 
@@ -66,8 +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++
@@ -76,7 +76,7 @@ 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;

changelog

@@ -1,3 +1,21 @@
+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.

cmake/FindFilesystem.cmake

@@ -1,75 +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)
-mark_as_advanced(Filesystem_FOUND)
-
-if(Filesystem_FIND_REQUIRED AND NOT Filesystem_FOUND)
-    message(FATAL_ERROR "Cannot run simple program using std::filesystem")
-endif()
-

examples/example.cpp

@@ -13,7 +13,7 @@ int main(int argc, char *argv[])
 		exit(1);
 	}
 
-	cif::file file = cif::pdb::read(argv[1]);
+	cif::file file(argv[1]);
 
 	if (file.empty())
 	{

include/cif++/atom_type.hpp

@@ -200,7 +200,7 @@ enum class radius_type
 };
 
 /// @brief The number of radii per element which can be requested from atom_type_info
-constexpr size_t kRadiusTypeCount = static_cast<size_t>(radius_type::type_count);
+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
@@ -276,7 +276,7 @@ class atom_type_traits
 	{
 		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

include/cif++/category.hpp

@@ -141,31 +141,39 @@ class category
 
 	/// \endcond
 
-	category() = default;                     ///< Default constructor
-	category(std::string_view name);          ///< Constructor taking a \a name
-	category(const category &rhs);            ///< Copy constructor
-	category(category &&rhs);                 ///< Move constructor
-	category &operator=(const category &rhs); ///< Copy assignement operator
-	category &operator=(category &&rhs);      ///< Move assignement operator
+	category() = default;            ///< Default constructor
+	category(std::string_view name); ///< Constructor taking a \a name
+	category(const category &rhs);   ///< Copy constructor
+
+	category(category &&rhs) noexcept ///< Move constructor
+	{
+		swap(*this, rhs);
+	}
+
+	category &operator=(category rhs) ///< assignement operator
+	{
+		swap(*this, rhs);
+		return *this;
+	}
 
 	/// @brief Destructor
 	/// @note Please note that the destructor is not virtual. It is assumed that
 	/// you will not derive from this class.
 	~category();
 
+	friend void swap(category &a, category &b) noexcept;
+
 	// --------------------------------------------------------------------
 
 	const std::string &name() const { return m_name; } ///< Returns the name of the category
 
-	[[deprecated("use key_items instead")]]
-	iset key_fields() const;                           ///< Returns the cif::iset of key item names. Retrieved from the @ref category_validator for this category
+	[[deprecated("use key_items instead")]] iset key_fields() const; ///< Returns the cif::iset of key item names. Retrieved from the @ref category_validator for this category
 
-	iset key_items() const;                           ///< Returns the cif::iset of key item names. Retrieved from the @ref category_validator for this category
+	iset key_items() const; ///< Returns the cif::iset of key item names. Retrieved from the @ref category_validator for this category
 
-	[[deprecated("use key_item_indices instead")]]
-	std::set<uint16_t> key_field_indices() const;      ///< Returns a set of indices for the key items.
+	[[deprecated("use key_item_indices instead")]] std::set<uint16_t> key_field_indices() const; ///< Returns a set of indices for the key items.
 
-	std::set<uint16_t> key_item_indices() const;      ///< Returns a set of indices for the key items.
+	std::set<uint16_t> key_item_indices() const; ///< Returns a set of indices for the key items.
 
 	/// @brief Set the validator for this category to @a v
 	/// @param v The category_validator to assign. A nullptr value is allowed.
@@ -174,7 +182,7 @@ class category
 
 	/// @brief Update the links in this category
 	/// @param db The enclosing @ref datablock
-	void update_links(datablock &db);
+	void update_links(const datablock &db);
 
 	/// @brief Return the global @ref validator for the data
 	/// @return The @ref validator or nullptr if not assigned
@@ -285,15 +293,15 @@ class category
 	}
 
 	/// Return a count of the rows in this container
-	size_t size() const
+	std::size_t size() const
 	{
 		return std::distance(cbegin(), cend());
 	}
 
 	/// Return the theoretical maximum number or rows that can be stored
-	size_t max_size() const
+	std::size_t max_size() const
 	{
-		return std::numeric_limits<size_t>::max(); // this is a bit optimistic, I guess
+		return std::numeric_limits<std::size_t>::max(); // this is a bit optimistic, I guess
 	}
 
 	/// Return true if the category is empty
@@ -823,9 +831,9 @@ class category
 	/// @brief Return the total number of rows that match condition @a cond
 	/// @param cond The condition to match
 	/// @return The count
-	size_t count(condition &&cond) const
+	std::size_t count(condition &&cond) const
 	{
-		size_t result = 0;
+		std::size_t result = 0;
 
 		if (cond)
 		{
@@ -895,14 +903,14 @@ class category
 	/// @brief Erase all rows that match condition @a cond
 	/// @param cond The condition
 	/// @return The number of rows that have been erased
-	size_t erase(condition &&cond);
+	std::size_t erase(condition &&cond);
 
 	/// @brief Erase all rows that match condition @a cond calling
 	/// the visitor function @a visit for each before actually erasing it.
 	/// @param cond The condition
 	/// @param visit The visitor function
 	/// @return The number of rows that have been erased
-	size_t erase(condition &&cond, std::function<void(row_handle)> &&visit);
+	std::size_t erase(condition &&cond, std::function<void(row_handle)> &&visit);
 
 	/// @brief Emplace the values in @a ri in a new row
 	/// @param ri An object containing the values to insert
@@ -964,6 +972,32 @@ class category
 
 	// --------------------------------------------------------------------
 
+	using value_provider_type = std::function<std::string_view(std::string_view)>;
+
+	/// \brief Update a single item named @a item_name in the rows that match
+	/// \a cond to values provided by a callback function \a value_provider
+	/// making sure the linked categories are updated according to the link.
+	/// That means, child categories are updated if the links are absolute
+	/// and unique. If they are not, the child category rows are split.
+
+	void update_value(condition &&cond, std::string_view item_name,
+		value_provider_type &&value_provider)
+	{
+		auto rs = find(std::move(cond));
+		std::vector<row_handle> rows;
+		std::copy(rs.begin(), rs.end(), std::back_inserter(rows));
+		update_value(rows, item_name, std::move(value_provider));
+	}
+
+	/// \brief Update a single item named @a item_name in the rows \a rows
+	/// to values provided by a callback function \a value_provider
+	/// making sure the linked categories are updated according to the link.
+	/// That means, child categories are updated if the links are absolute
+	/// and unique. If they are not, the child category rows are split.
+
+	void update_value(const std::vector<row_handle> &rows, std::string_view item_name,
+		value_provider_type &&value_provider);
+
 	/// \brief Update a single item named @a item_name in the rows that match \a cond to value \a value
 	/// making sure the linked categories are updated according to the link.
 	/// That means, child categories are updated if the links are absolute
@@ -982,15 +1016,18 @@ class category
 	/// That means, child categories are updated if the links are absolute
 	/// and unique. If they are not, the child category rows are split.
 
-	void update_value(const std::vector<row_handle> &rows, std::string_view item_name, std::string_view value);
+	void update_value(const std::vector<row_handle> &rows, std::string_view item_name, std::string_view value)
+	{
+		update_value(rows, item_name, [value](std::string_view)
+			{ return value; });
+	}
 
 	// --------------------------------------------------------------------
 	// Naming used to be very inconsistent. For backward compatibility,
 	// the old function names are here as deprecated variants.
 
 	/// \brief Return the index number for \a column_name
-	[[deprecated("Use get_item_ix instead")]]
-	uint16_t get_column_ix(std::string_view column_name) const
+	[[deprecated("Use get_item_ix instead")]] uint16_t get_column_ix(std::string_view column_name) const
 	{
 		return get_item_ix(column_name);
 	}
@@ -998,8 +1035,7 @@ class category
 	/// @brief Return the name for column with index @a ix
 	/// @param ix The index number
 	/// @return The name of the column
-	[[deprecated("use get_item_name instead")]]
-	std::string_view get_column_name(uint16_t ix) const
+	[[deprecated("use get_item_name instead")]] std::string_view get_column_name(uint16_t ix) const
 	{
 		return get_item_name(ix);
 	}
@@ -1007,8 +1043,7 @@ class category
 	/// @brief Make sure a item with name @a item_name is known and return its index number
 	/// @param item_name The name of the item
 	/// @return The index number of the item
-	[[deprecated("use add_item instead")]]
-	uint16_t add_column(std::string_view item_name)
+	[[deprecated("use add_item instead")]] uint16_t add_column(std::string_view item_name)
 	{
 		return add_item(item_name);
 	}
@@ -1016,15 +1051,13 @@ class category
 	/** @brief Remove column name @a colum_name
 	 * @param column_name The column to be removed
 	 */
-	[[deprecated("use remove_item instead")]]
-	void remove_column(std::string_view column_name)
+	[[deprecated("use remove_item instead")]] void remove_column(std::string_view column_name)
 	{
 		remove_item(column_name);
 	}
 
 	/** @brief Rename column @a from_name to @a to_name */
-	[[deprecated("use rename_item instead")]]
-	void rename_column(std::string_view from_name, std::string_view to_name)
+	[[deprecated("use rename_item instead")]] void rename_column(std::string_view from_name, std::string_view to_name)
 	{
 		rename_item(from_name, to_name);
 	}
@@ -1032,15 +1065,13 @@ class category
 	/// @brief Return whether a column with name @a name exists in this category
 	/// @param name The name of the column
 	/// @return True if the column exists
-	[[deprecated("use has_item instead")]]
-	bool has_column(std::string_view name) const
+	[[deprecated("use has_item instead")]] bool has_column(std::string_view name) const
 	{
 		return has_item(name);
 	}
 
 	/// @brief Return the cif::iset of columns in this category
-	[[deprecated("use get_items instead")]]
-	iset get_columns() const
+	[[deprecated("use get_items instead")]] iset get_columns() const
 	{
 		return get_items();
 	}
@@ -1096,7 +1127,7 @@ class category
 			{
 				item_validator = m_cat_validator->get_validator_for_item(item_name);
 				if (item_validator == nullptr)
-					m_validator->report_error( validation_error::item_not_allowed_in_category, m_name, item_name, false);
+					m_validator->report_error(validation_error::item_not_allowed_in_category, m_name, item_name, false);
 			}
 
 			m_items.emplace_back(item_name, item_validator);
@@ -1140,8 +1171,7 @@ class category
 
 	/// This function returns effectively the list of fully qualified item
 	/// names, that is category_name + '.' + item_name for each item
-	[[deprecated("use get_item_order instead")]]
-	std::vector<std::string> get_tag_order() const
+	[[deprecated("use get_item_order instead")]] std::vector<std::string> get_tag_order() const
 	{
 		return get_item_order();
 	}
@@ -1233,6 +1263,7 @@ class category
 		{
 		}
 
+// TODO: NEED TO FIX THIS!
 		category *linked;
 		const link_validator *v;
 	};

include/cif++/compound.hpp

@@ -138,7 +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

include/cif++/condition.hpp

@@ -29,6 +29,7 @@
 #include "cif++/row.hpp"
 
 #include <cassert>
+#include <concepts>
 #include <functional>
 #include <iostream>
 #include <regex>
@@ -370,7 +371,7 @@ namespace detail
 	{
 		key_equals_condition_impl(item &&i)
 			: m_item_name(i.name())
-			, m_value(i.value())
+			, m_value(std::forward<item>(i).value())
 		{
 		}
 
@@ -470,6 +471,106 @@ namespace detail
 		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>
@@ -876,7 +977,8 @@ inline condition operator or(condition &&a, condition &&b)
 			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);
@@ -886,6 +988,26 @@ inline condition operator or(condition &&a, condition &&b)
 				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)));
 	}
 
@@ -959,13 +1081,16 @@ struct key
 	std::string m_item_name; ///< The item name
 };
 
-/**
- * @brief Operator to create an equals condition based on a key @a key and a value @a v
- */
 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_name, v }));
+	return condition(new detail::key_equals_number_condition_impl(key.m_item_name, v));
 }
 
 /**
@@ -979,6 +1104,16 @@ inline condition operator==(const key &key, std::string_view value)
 		return condition(new detail::key_is_empty_condition_impl(key.m_item_name));
 }
 
+/**
+ * @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
  */
@@ -999,7 +1134,7 @@ inline condition operator!=(const key &key, std::string_view value)
 /**
  * @brief Operator to create a greater than condition based on a key @a key and a value @a v
  */
-template <typename T>
+template <Numeric T>
 condition operator>(const key &key, const T &v)
 {
 	std::ostringstream s;
@@ -1007,14 +1142,14 @@ condition operator>(const key &key, const T &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].template compare<T>(v, icase) > 0; },
+		{ return r[item_name].compare(v) > 0; },
 		s.str()));
 }
 
 /**
  * @brief Operator to create a greater than or equals condition based on a key @a key and a value @a v
  */
-template <typename T>
+template <Numeric T>
 condition operator>=(const key &key, const T &v)
 {
 	std::ostringstream s;
@@ -1022,14 +1157,14 @@ condition operator>=(const key &key, const T &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].template compare<T>(v, icase) >= 0; },
+		{ return r[item_name].compare(v) >= 0; },
 		s.str()));
 }
 
 /**
  * @brief Operator to create a less than condition based on a key @a key and a value @a v
  */
-template <typename T>
+template <Numeric T>
 condition operator<(const key &key, const T &v)
 {
 	std::ostringstream s;
@@ -1037,14 +1172,14 @@ condition operator<(const key &key, const T &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].template compare<T>(v, icase) < 0; },
+		{ return r[item_name].compare(v) < 0; },
 		s.str()));
 }
 
 /**
  * @brief Operator to create a less than or equals condition based on a key @a key and a value @a v
  */
-template <typename T>
+template <Numeric T>
 condition operator<=(const key &key, const T &v)
 {
 	std::ostringstream s;
@@ -1052,7 +1187,63 @@ condition operator<=(const key &key, const T &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].template compare<T>(v, icase) <= 0; },
+		{ 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()));
 }
 
@@ -1093,6 +1284,19 @@ condition operator==(const key &key, const std::optional<T> &v)
 		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
  */
@@ -1146,7 +1350,7 @@ namespace literals
 	 * @param length The length of @a text
 	 * @return key The cif::key created
 	 */
-	inline key operator""_key(const char *text, size_t length)
+	inline key operator""_key(const char *text, std::size_t length)
 	{
 		return key(std::string(text, length));
 	}

include/cif++/datablock.hpp

@@ -61,12 +61,26 @@ class datablock : public std::list<category>
 
 	/** @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));
+	}
+
 	// --------------------------------------------------------------------
 
 	/**

include/cif++/file.hpp

@@ -85,11 +85,11 @@ class file : public std::list<datablock>
 	 * @param data The pointer to the character string with data to load
 	 * @param length The length of the data
 	 */
-	explicit file(const char *data, size_t length)
+	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);
 			}
@@ -100,10 +100,22 @@ class file : public std::list<datablock>
 	}
 
 	/** @cond */
-	file(const file &) = default;
-	file(file &&) = default;
-	file &operator=(const file &) = default;
-	file &operator=(file &&) = default;
+	file(const file &rhs)
+		: std::list<datablock>(rhs)
+	{
+	}
+
+	file(file &&rhs)
+	{
+		this->swap(rhs);
+	}
+
+	file &operator=(file f)
+	{
+		this->swap(f);
+		return *this;
+	}
+
 	/** @endcond */
 
 	/**

include/cif++/format.hpp

@@ -125,7 +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)), ...);

include/cif++/gzio.hpp

@@ -36,7 +36,7 @@ namespace cif::gzio
 {
 
 /** The default buffer size to use */
-const size_t kDefaultBufferSize = 256;
+const std::size_t kDefaultBufferSize = 256;
 
 // --------------------------------------------------------------------
 
@@ -112,7 +112,7 @@ 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:
@@ -317,7 +317,7 @@ 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:

include/cif++/item.hpp

@@ -117,11 +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);
 	}
 
@@ -138,11 +136,9 @@ 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);
 	}
 
@@ -155,11 +151,9 @@ class item
 		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);
 	}
 
@@ -174,12 +168,21 @@ 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>
@@ -219,7 +222,8 @@ class item
 	/** @endcond */
 
 	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_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; }
@@ -234,10 +238,10 @@ class item
 	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)
@@ -284,19 +288,16 @@ struct item_value
 	}
 
 	/** @cond */
-	item_value(item_value &&rhs)
+	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;
 	}
 
@@ -318,7 +319,7 @@ struct item_value
 		return m_length != 0;
 	}
 
-	size_t m_length = 0; ///< Length of the data
+	std::size_t m_length = 0; ///< Length of the data
 	union
 	{
 		char m_local_data[8]; ///< Storage area for small strings (strings smaller than kBufferSize)
@@ -327,7 +328,7 @@ struct item_value
 	};
 
 	/** The maximum length of locally stored strings */
-	static constexpr size_t kBufferSize = sizeof(m_local_data);
+	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.
@@ -363,8 +364,35 @@ struct 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;
 	}
 
@@ -508,7 +536,7 @@ struct item_handle
 	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...
@@ -532,7 +560,7 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> an
 
 			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 (r.ec != std::errc() or r.ptr != e)
+			if ((bool)r.ec or r.ptr != e)
 			{
 				result = {};
 				if (cif::VERBOSE)
@@ -567,7 +595,7 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> an
 
 			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 (r.ec != std::errc() or r.ptr != e)
+			if ((bool)r.ec or r.ptr != e)
 			{
 				if (cif::VERBOSE)
 				{
@@ -580,6 +608,8 @@ struct item_handle::item_value_as<T, std::enable_if_t<std::is_arithmetic_v<T> an
 				}
 				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)
@@ -634,7 +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)

include/cif++/iterator.hpp

@@ -49,7 +49,7 @@ namespace cif
 /**
  * @brief Implementation of an iterator that can return
  * multiple values in a tuple. Of course, that tuple can
- * then used in structured binding to receive the values
+ * then be used in structured binding to receive the values
  * in a for loop e.g.
  *
  * @tparam Category The category for this iterator
@@ -67,7 +67,7 @@ class iterator_impl
 	/** @endcond */
 
 	/** variable that contains the number of elements in the tuple */
-	static constexpr size_t N = sizeof...(Ts);
+	static constexpr std::size_t N = sizeof...(Ts);
 
 	/** @cond */
 	using category_type = std::remove_cv_t<Category>;
@@ -84,11 +84,11 @@ 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_item_ix(rhs.m_item_ix)
 	{
@@ -96,8 +96,7 @@ class iterator_impl
 
 	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_item_ix(rhs.m_item_ix)
 	{
@@ -106,19 +105,17 @@ class iterator_impl
 
 	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_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_item_ix = i.m_item_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;
 	}
 
@@ -136,18 +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>());
 
@@ -179,20 +176,13 @@ class iterator_impl
 	/** @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_item_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_item_ix;
 };
@@ -219,37 +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;
 	}
 
@@ -257,7 +244,7 @@ class iterator_impl<Category>
 
 	reference operator*()
 	{
-		return { *m_category, *m_current };
+		return m_current;
 	}
 
 	pointer operator->()
@@ -267,18 +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;
 	}
@@ -308,8 +295,7 @@ class iterator_impl<Category>
 	/** @endcond */
 
   private:
-	category_type *m_category = nullptr;
-	row_type *m_current = nullptr;
+	row_handle m_current;
 };
 
 /**
@@ -342,11 +328,11 @@ 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_item_ix(rhs.m_item_ix)
 	{
@@ -354,29 +340,26 @@ class iterator_impl<Category, T>
 
 	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_item_ix(rhs.m_item_ix)
 	{
-		m_value = get(m_current);
+		m_value = get();
 	}
 
 	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_current(const_cast<row_handle&>(rhs.m_current))
 		, m_item_ix(cix[0])
 	{
 		m_value = get();
 	}
 
-	iterator_impl &operator=(const iterator_impl &i)
+	iterator_impl &operator=(iterator_impl i)
 	{
-		m_category = i.m_category;
-		m_current = i.m_current;
-		m_item_ix = i.m_item_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;
 	}
 
@@ -394,18 +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();
 
@@ -439,17 +422,10 @@ class iterator_impl<Category, T>
   private:
 	value_type get() const
 	{
-		if (m_current != nullptr)
-		{
-			row_handle rh{ *m_category, *m_current };
-			return rh[m_item_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_item_ix;
 };
@@ -474,7 +450,7 @@ class iterator_proxy
 {
   public:
 	/** @cond */
-	static constexpr const size_t N = sizeof...(Ts);
+	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>;
@@ -497,7 +473,7 @@ class iterator_proxy
 
 	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
-	size_t size() const { return std::distance(begin(), end()); } ///< Return size of the range
+	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())); }
@@ -536,7 +512,7 @@ class conditional_iterator_proxy
 {
   public:
 	/** @cond */
-	static constexpr const size_t N = sizeof...(Ts);
+	static constexpr const std::size_t N = sizeof...(Ts);
 
 	using category_type = std::remove_cv_t<CategoryType>;
 
@@ -630,7 +606,7 @@ class conditional_iterator_proxy
 
 	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
-	size_t size() const { return std::distance(begin(), end()); } ///< Return size of the range
+	std::size_t size() const { return std::distance(begin(), end()); } ///< Return size of the range
 
 	row_handle front() { return *begin(); } ///< Return reference to the first row
 	// row_handle back() { return *begin(); }

include/cif++/matrix.hpp

@@ -59,27 +59,27 @@ template <typename M>
 class matrix_expression
 {
   public:
-	constexpr size_t dim_m() const { return static_cast<const M &>(*this).dim_m(); } ///< Return the size (dimension) in direction m
-	constexpr size_t dim_n() const { return static_cast<const M &>(*this).dim_n(); } ///< Return the size (dimension) in direction 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; } ///< Convenient way to test for empty matrices
 
 	/** Return a reference to element [ @a i, @a j ] */
-	constexpr auto &operator()(size_t i, size_t j)
+	constexpr auto &operator()(std::size_t i, std::size_t j)
 	{
 		return static_cast<M &>(*this).operator()(i, j);
 	}
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr auto operator()(size_t i, size_t j) const
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return static_cast<const M &>(*this).operator()(i, j);
 	}
 
 	/** Swap the contents of rows @a r1 and @a r2 */
-	void swap_row(size_t r1, size_t 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);
@@ -88,9 +88,9 @@ class matrix_expression
 	}
 
 	/** Swap the contents of columns @a c1 and @a c2 */
-	void swap_col(size_t c1, size_t 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);
@@ -103,11 +103,11 @@ class matrix_expression
 	{
 		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())
@@ -156,9 +156,9 @@ class matrix : public matrix_expression<matrix<F>>
 		, 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);
 		}
 	}
@@ -171,7 +171,7 @@ class matrix : public matrix_expression<matrix<F>>
 	 * @param n Requested dimension N
 	 * @param v Value to store in each element
 	 */
-	matrix(size_t m, size_t n, value_type v = 0)
+	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)
@@ -187,11 +187,11 @@ class matrix : public matrix_expression<matrix<F>>
 	matrix &operator=(const matrix &m) = default;
 	/** @endcond */
 
-	constexpr size_t dim_m() const { return m_m; } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_n; } ///< Return dimension 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
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr value_type operator()(size_t i, size_t j) const
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		assert(i < m_m);
 		assert(j < m_n);
@@ -199,7 +199,7 @@ class matrix : public matrix_expression<matrix<F>>
 	}
 
 	/** Return a reference to element [ @a i, @a j ] */
-	constexpr value_type &operator()(size_t i, size_t j)
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		assert(i < m_m);
 		assert(j < m_n);
@@ -207,7 +207,7 @@ 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;
 };
 
@@ -224,7 +224,7 @@ class matrix : public matrix_expression<matrix<F>>
  * element m i,j is mapped to [i * n + j] and thus storage is row major
  */
 
-template <typename F, size_t M, size_t N>
+template <typename F, std::size_t M, std::size_t N>
 class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 {
   public:
@@ -232,16 +232,16 @@ class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 	using value_type = F;
 
 	/** The storage size */
-	static constexpr size_t kSize = M * N;
+	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);
 		}
 	}
@@ -266,18 +266,18 @@ class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 	/** @endcond */
 
 	/** Store the values in @a a in the matrix */
-	template<size_t... Ixs>
+	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; } ///< Return dimension m
-	constexpr size_t dim_n() const { return N; } ///< Return dimension 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
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr value_type operator()(size_t i, size_t j) const
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		assert(i < M);
 		assert(j < N);
@@ -285,7 +285,7 @@ class matrix_fixed : public matrix_expression<matrix_fixed<F, M, N>>
 	}
 
 	/** Return a reference to element [ @a i, @a j ] */
-	constexpr value_type &operator()(size_t i, size_t j)
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		assert(i < M);
 		assert(j < N);
@@ -322,7 +322,7 @@ class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
 	using value_type = F;
 
 	/** constructor for a matrix of size @a n x @a n elements with value @a v */
-	symmetric_matrix(size_t n, value_type v = 0)
+	symmetric_matrix(std::size_t n, value_type v = 0)
 		: m_n(n)
 		, m_data((m_n * (m_n + 1)) / 2)
 	{
@@ -337,11 +337,11 @@ class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
 	symmetric_matrix &operator=(const symmetric_matrix &m) = default;
 	/** @endcond */
 
-	constexpr size_t dim_m() const { return m_n; } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_n; } ///< Return dimension 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
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr value_type operator()(size_t i, size_t j) const
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		return i < j
 		           ? m_data[(j * (j + 1)) / 2 + i]
@@ -349,7 +349,7 @@ class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
 	}
 
 	/** Return a reference to element [ @a i, @a j ] */
-	constexpr value_type &operator()(size_t i, size_t j)
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		if (i > j)
 			std::swap(i, j);
@@ -358,7 +358,7 @@ 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;
 };
 
@@ -373,7 +373,7 @@ class symmetric_matrix : public matrix_expression<symmetric_matrix<F>>
  * 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, size_t M>
+template <typename F, std::size_t M>
 class symmetric_matrix_fixed : public matrix_expression<symmetric_matrix_fixed<F, M>>
 {
   public:
@@ -393,11 +393,11 @@ class symmetric_matrix_fixed : public matrix_expression<symmetric_matrix_fixed<F
 	symmetric_matrix_fixed &operator=(const symmetric_matrix_fixed &m) = default;
 	/** @endcond */
 
-	constexpr size_t dim_m() const { return M; } ///< Return dimension m
-	constexpr size_t dim_n() const { return M; } ///< Return dimension n
+	constexpr std::size_t dim_m() const { return M; } ///< Return dimension m
+	constexpr std::size_t dim_n() const { return M; } ///< Return dimension n
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr value_type operator()(size_t i, size_t j) const
+	constexpr value_type operator()(std::size_t i, std::size_t j) const
 	{
 		return i < j
 		           ? m_data[(j * (j + 1)) / 2 + i]
@@ -405,7 +405,7 @@ class symmetric_matrix_fixed : public matrix_expression<symmetric_matrix_fixed<F
 	}
 
 	/** Return a reference to element [ @a i, @a j ] */
-	constexpr value_type &operator()(size_t i, size_t j)
+	constexpr value_type &operator()(std::size_t i, std::size_t j)
 	{
 		if (i > j)
 			std::swap(i, j);
@@ -444,22 +444,22 @@ class identity_matrix : public matrix_expression<identity_matrix<F>>
 	using value_type = F;
 
 	/** constructor taking a dimension @a n */
-	identity_matrix(size_t n)
+	identity_matrix(std::size_t n)
 		: m_n(n)
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_n; } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_n; } ///< Return dimension 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
 
 	/** Return the value of element [ @a i, @a j ] */
-	constexpr value_type operator()(size_t i, size_t j) const
+	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;
 };
 
 // --------------------------------------------------------------------
@@ -484,11 +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(); } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_m1.dim_n(); } ///< Return dimension 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
 
 	/** Access to the value of element [ @a i, @a j ] */
-	constexpr auto operator()(size_t i, size_t j) const
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return m_m1(i, j) - m_m2(i, j);
 	}
@@ -523,17 +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(); } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_m1.dim_n(); } ///< Return dimension 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
 
 	/** Access to the value of element [ @a i, @a j ] */
-	constexpr auto operator()(size_t i, size_t j) const
+	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;
@@ -564,11 +564,11 @@ class matrix_scalar_multiplication : public matrix_expression<matrix_scalar_mult
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_m.dim_m(); } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_m.dim_n(); } ///< Return dimension 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
 
 	/** Access to the value of element [ @a i, @a j ] */
-	constexpr auto operator()(size_t i, size_t j) const
+	constexpr auto operator()(std::size_t i, std::size_t j) const
 	{
 		return m_m(i, j) * m_v;
 	}
@@ -655,21 +655,21 @@ class matrix_cofactors : public matrix_expression<matrix_cofactors<M>>
 	{
 	}
 
-	constexpr size_t dim_m() const { return m_m.dim_m(); } ///< Return dimension m
-	constexpr size_t dim_n() const { return m_m.dim_n(); } ///< Return dimension 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
 
 	/** Access to the value of element [ @a i, @a j ] */
-	constexpr auto operator()(size_t i, size_t j) const
+	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++/model.hpp

@@ -350,7 +350,12 @@ class atom
 	std::string get_pdb_ins_code() const { return get_property("pdbx_PDB_ins_code"); } ///< Return the pdb_ins_code property
 
 	/// Return true if this atom is an alternate
-	bool is_alternate() const { return not get_label_alt_id().empty(); }
+	bool is_alternate() const
+	{
+		if (auto alt_id = get_label_alt_id(); alt_id.empty() or alt_id == ".")
+			return false;
+		return true;
+	}
 
 	/// Convenience method to return a string that might be ID in PDB space
 	std::string pdb_id() const
@@ -550,6 +555,9 @@ class residue
 	/// \brief Return true if this residue has alternate atoms
 	bool has_alternate_atoms() const;
 
+	/// \brief Return true if this residue has alternate atoms for the atom \a atomID
+	bool has_alternate_atoms_for(const std::string &atomID) const;
+
 	/// \brief Return the list of unique alt ID's present in this residue
 	std::set<std::string> get_alternate_ids() const;
 
@@ -572,6 +580,10 @@ class residue
 								   m_auth_seq_id == rhs.m_auth_seq_id);
 	}
 
+	/// @brief Create a new atom and add it to the list
+	/// @return newly created atom
+	virtual atom create_new_atom(atom_type inType, const std::string &inAtomID, point inLocation);
+
   protected:
 	/** @cond */
 	residue() {}
@@ -604,7 +616,7 @@ class monomer : public residue
 	monomer &operator=(monomer &&rhs);
 
 	/// \brief constructor with actual values
-	monomer(const polymer &polymer, size_t index, int seqID, const std::string &authSeqID,
+	monomer(const polymer &polymer, std::size_t index, int seqID, const std::string &authSeqID,
 		const std::string &pdbInsCode, const std::string &compoundID);
 
 	bool is_first_in_chain() const; ///< Return if this residue is the first residue in the chain
@@ -624,8 +636,8 @@ class monomer : public residue
 	float omega() const; ///< Return the omega value for this residue
 
 	// torsion angles
-	size_t nr_of_chis() const; ///< Return how many torsion angles can be calculated
-	float chi(size_t i) const; ///< Return torsion angle @a i
+	std::size_t nr_of_chis() const; ///< Return how many torsion angles can be calculated
+	float chi(std::size_t i) const; ///< Return torsion angle @a i
 
 	bool is_cis() const; ///< Return true if this residue is in a cis conformation
 
@@ -672,9 +684,11 @@ class monomer : public residue
 		return m_polymer == rhs.m_polymer and m_index == rhs.m_index;
 	}
 
+	atom create_new_atom(atom_type inType, const std::string &inAtomID, point inLocation) override;
+
   private:
 	const polymer *m_polymer;
-	size_t m_index;
+	std::size_t m_index;
 };
 
 // --------------------------------------------------------------------
@@ -741,7 +755,7 @@ class sugar : public residue
 	{
 		int result;
 		auto r = std::from_chars(m_auth_seq_id.data(), m_auth_seq_id.data() + m_auth_seq_id.length(), result);
-		if (r.ec != std::errc())
+		if ((bool)r.ec)
 			throw std::runtime_error("The auth_seq_id should be a number for a sugar");
 		return result;
 	}
@@ -756,9 +770,9 @@ class sugar : public residue
 	void set_link(atom link) { m_link = link; }
 
 	/// \brief Return the sugar number of the sugar linked to C1
-	size_t get_link_nr() const
+	std::size_t get_link_nr() const
 	{
-		size_t result = 0;
+		std::size_t result = 0;
 		if (m_link)
 			result = m_link.get_property_int("auth_seq_id");
 		return result;
@@ -865,10 +879,10 @@ class structure
 {
   public:
 	/// \brief Read the structure from cif::file @a p
-	structure(file &p, size_t modelNr = 1, StructureOpenOptions options = {});
+	structure(file &p, std::size_t modelNr = 1, StructureOpenOptions options = {});
 
 	/// \brief Load the structure from already parsed mmCIF data in @a db
-	structure(datablock &db, size_t modelNr = 1, StructureOpenOptions options = {});
+	structure(datablock &db, std::size_t modelNr = 1, StructureOpenOptions options = {});
 
 	/** @cond */
 	structure(structure &&s) = default;
@@ -881,7 +895,7 @@ class structure
 	~structure() = default;
 
 	/// \brief Return the model number
-	size_t get_model_nr() const { return m_model_nr; }
+	std::size_t get_model_nr() const { return m_model_nr; }
 
 	/// \brief Return a list of all the atoms in this structure
 	const std::vector<atom> &atoms() const { return m_atoms; }
@@ -1091,6 +1105,9 @@ class structure
 	/// \brief emplace the moved atom @a atom
 	atom &emplace_atom(atom &&atom);
 
+	/// \brief Reorder atom_site atoms based on 'natural' ordering
+	void reorder_atoms();
+
   private:
 	friend polymer;
 	friend residue;
@@ -1107,9 +1124,9 @@ class structure
 	void remove_sugar(sugar &sugar);
 
 	datablock &m_db;
-	size_t m_model_nr;
+	std::size_t m_model_nr;
 	std::vector<atom> m_atoms;
-	std::vector<size_t> m_atom_index;
+	std::vector<std::size_t> m_atom_index;
 	std::list<polymer> m_polymers;
 	std::list<branch> m_branches;
 	std::vector<residue> m_non_polymers;

include/cif++/point.hpp

@@ -867,7 +867,7 @@ class spherical_dots
 	}
 
 	/// \brief The number of points
-	size_t size() const { return P; }
+	std::size_t size() const { return P; }
 
 	/// \brief Access a point by index
 	const point operator[](uint32_t inIx) const { return m_points[inIx]; }

include/cif++/row.hpp

@@ -85,7 +85,7 @@ 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> &&items)
 			: m_row(r)
@@ -104,7 +104,7 @@ 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_items[Is]].template as<Ts>()... };
@@ -208,6 +208,7 @@ class row_handle
 	friend class category;
 	friend class category_index;
 	friend class row_initializer;
+	template <typename, typename...> friend class iterator_impl;
 
 	row_handle() = default;
 

include/cif++/symmetry.hpp

@@ -300,7 +300,7 @@ namespace literals
 	 * @endcode
 	 *
 	 */
-	inline sym_op operator""_symop(const char *text, size_t length)
+	inline sym_op operator""_symop(const char *text, std::size_t length)
 	{
 		return sym_op({ text, length });
 	}
@@ -464,7 +464,7 @@ class spacegroup : public std::vector<transformation>
 
   private:
 	int m_nr;
-	size_t m_index;
+	std::size_t m_index;
 };
 
 // --------------------------------------------------------------------

include/cif++/text.hpp

@@ -352,7 +352,7 @@ std::string cif_id_for_number(int number);
  * a dynamic programming approach to get the most efficient filling of
  * the space.
  */
-std::vector<std::string> word_wrap(const std::string &text, size_t width);
+std::vector<std::string> word_wrap(const std::string &text, std::size_t width);
 
 // --------------------------------------------------------------------
 /// \brief std::from_chars for floating point types.
@@ -378,12 +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;
@@ -427,7 +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;
@@ -467,9 +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);
 

include/cif++/utilities.hpp

@@ -113,16 +113,12 @@ namespace colour
 		/**
 		 * @brief Struct for delimited strings.
 		 */
-		template <typename StringType>
 		struct coloured_string_t
 		{
-			static_assert(std::is_reference_v<StringType> or std::is_pointer_v<StringType>,
-				"String type must be pointer or reference");
-
 			/**
 			 * @brief Construct a new coloured string t object
 			 */
-			coloured_string_t(StringType s, colour_type fc, colour_type bc, style_type st)
+			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)
@@ -152,12 +148,14 @@ namespace colour
 					   << cs.m_str
 					   << "\033[0m";
 				}
+				else
+					os << cs.m_str;
 
 				return os;
 			}
 
 			/// @cond
-			StringType m_str;
+			std::string_view m_str;
 			int m_fore_colour, m_back_colour;
 			int m_style;
 			/// @endcond
@@ -191,39 +189,13 @@ namespace colour
  * @param st Text style to use
  */
 
-template <typename char_type>
-inline auto coloured(const char_type *str,
+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 colour::detail::coloured_string_t<const char_type *>(str, fg, bg, st);
-}
-
-/// @brief Manipulator for coloured strings.
-template <typename char_type, typename traits_type, typename allocator_type>
-inline auto coloured(const std::basic_string<char_type, traits_type, allocator_type> &str,
-	colour::colour_type fg, colour::colour_type bg = colour::colour_type::none,
-	colour::style_type st = colour::style_type::regular)
-{
-	return colour::detail::coloured_string_t<const std::basic_string<char_type, traits_type, allocator_type> &>(str, fg, bg, st);
-}
-
-/// @brief Manipulator for coloured strings.
-template <typename char_type, typename traits_type, typename allocator_type>
-inline auto coloured(std::basic_string<char_type, traits_type, allocator_type> &str,
-	colour::colour_type fg, colour::colour_type bg = colour::colour_type::none,
-	colour::style_type st = colour::style_type::regular)
-{
-	return colour::detail::coloured_string_t<std::basic_string<char_type, traits_type, allocator_type> &>(str, fg, bg, st);
-}
-
-/// @brief Manipulator for coloured strings.
-template <typename char_type, typename traits_type>
-inline auto coloured(std::basic_string_view<char_type, traits_type> &str,
-	colour::colour_type fg, colour::colour_type bg = colour::colour_type::none,
-	colour::style_type st = colour::style_type::regular)
-{
-	return colour::detail::coloured_string_t<std::basic_string_view<char_type, traits_type> &>(str, fg, bg, st);
+	return colour::detail::coloured_string_t(str, fg, bg, st);
 }
 
 // --------------------------------------------------------------------

include/cif++/validate.hpp

@@ -120,7 +120,7 @@ class validation_category_impl : public std::error_category
 			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 allowed in category according to dictionary";
+				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:

src/category.cpp

@@ -164,7 +164,7 @@ class category_index
 		return result;
 	}
 
-	size_t size() const;
+	std::size_t size() const;
 	//	bool isValid() const;
 
   private:
@@ -475,12 +475,12 @@ category_index::entry *category_index::erase(category &cat, entry *h, row *k)
 	return fix_up(h);
 }
 
-size_t category_index::size() const
+std::size_t category_index::size() const
 {
 	std::stack<entry *> s;
 	s.push(m_root);
 
-	size_t result = 0;
+	std::size_t result = 0;
 
 	while (not s.empty())
 	{
@@ -521,71 +521,18 @@ category::category(const category &rhs)
 		m_index = new category_index(*this);
 }
 
-category::category(category &&rhs)
-	: m_name(std::move(rhs.m_name))
-	, m_items(std::move(rhs.m_items))
-	, m_validator(rhs.m_validator)
-	, m_cat_validator(rhs.m_cat_validator)
-	, m_parent_links(std::move(rhs.m_parent_links))
-	, m_child_links(std::move(rhs.m_child_links))
-	, m_cascade(rhs.m_cascade)
-	, m_index(rhs.m_index)
-	, m_head(rhs.m_head)
-	, m_tail(rhs.m_tail)
+void swap(category &a, category &b) noexcept
 {
-	rhs.m_head = nullptr;
-	rhs.m_tail = nullptr;
-	rhs.m_index = nullptr;
-}
-
-category &category::operator=(const category &rhs)
-{
-	if (this != &rhs)
-	{
-		if (not empty())
-			clear();
-
-		m_name = rhs.m_name;
-		m_items = rhs.m_items;
-		m_cascade = rhs.m_cascade;
-
-		m_validator = nullptr;
-		m_cat_validator = nullptr;
-
-		delete m_index;
-		m_index = nullptr;
-
-		for (auto r = rhs.m_head; r != nullptr; r = r->m_next)
-			insert_impl(cend(), clone_row(*r));
-
-		m_validator = rhs.m_validator;
-		m_cat_validator = rhs.m_cat_validator;
-
-		if (m_cat_validator != nullptr and m_index == nullptr)
-			m_index = new category_index(*this);
-	}
-
-	return *this;
-}
-
-category &category::operator=(category &&rhs)
-{
-	if (this != &rhs)
-	{
-		m_name = std::move(rhs.m_name);
-		m_items = std::move(rhs.m_items);
-		m_cascade = rhs.m_cascade;
-		m_validator = rhs.m_validator;
-		m_cat_validator = rhs.m_cat_validator;
-		m_parent_links = rhs.m_parent_links;
-		m_child_links = rhs.m_child_links;
-
-		std::swap(m_index, rhs.m_index);
-		std::swap(m_head, rhs.m_head);
-		std::swap(m_tail, rhs.m_tail);
-	}
-
-	return *this;
+	std::swap(a.m_name, b.m_name);
+	std::swap(a.m_items, b.m_items);
+	std::swap(a.m_validator, b.m_validator);
+	std::swap(a.m_cat_validator, b.m_cat_validator);
+	std::swap(a.m_parent_links, b.m_parent_links);
+	std::swap(a.m_child_links, b.m_child_links);
+	std::swap(a.m_cascade, b.m_cascade);
+	std::swap(a.m_index, b.m_index);
+	std::swap(a.m_head, b.m_head);
+	std::swap(a.m_tail, b.m_tail);
 }
 
 category::~category()
@@ -597,7 +544,7 @@ category::~category()
 
 void category::remove_item(std::string_view item_name)
 {
-	for (size_t ix = 0; ix < m_items.size(); ++ix)
+	for (std::size_t ix = 0; ix < m_items.size(); ++ix)
 	{
 		if (not iequals(item_name, m_items[ix].m_name))
 			continue;
@@ -616,7 +563,7 @@ void category::remove_item(std::string_view item_name)
 
 void category::rename_item(std::string_view from_name, std::string_view to_name)
 {
-	for (size_t ix = 0; ix < m_items.size(); ++ix)
+	for (std::size_t ix = 0; ix < m_items.size(); ++ix)
 	{
 		if (not iequals(from_name, m_items[ix].m_name))
 			continue;
@@ -719,7 +666,7 @@ void category::set_validator(const validator *v, datablock &db)
 	update_links(db);
 }
 
-void category::update_links(datablock &db)
+void category::update_links(const datablock &db)
 {
 	m_child_links.clear();
 	m_parent_links.clear();
@@ -728,7 +675,7 @@ void category::update_links(datablock &db)
 	{
 		for (auto link : m_validator->get_links_for_parent(m_name))
 		{
-			auto childCat = db.get(link->m_child_category);
+			auto childCat = const_cast<category *>(db.get(link->m_child_category));
 			if (childCat == nullptr)
 				continue;
 			m_child_links.emplace_back(childCat, link);
@@ -736,7 +683,7 @@ void category::update_links(datablock &db)
 
 		for (auto link : m_validator->get_links_for_child(m_name))
 		{
-			auto parentCat = db.get(link->m_parent_category);
+			auto parentCat = const_cast<category *>(db.get(link->m_parent_category));
 			if (parentCat == nullptr)
 				continue;
 			m_parent_links.emplace_back(parentCat, link);
@@ -844,7 +791,7 @@ bool category::is_valid() const
 
 				iv->validate_value(vi->text(), ec);
 
-				if (ec != std::errc())
+				if ((bool)ec)
 				{
 					m_validator->report_error(ec, m_name, m_items[cix].m_name, false);
 					continue;
@@ -885,7 +832,7 @@ bool category::validate_links() const
 		if (name() == "atom_site" and (parent->name() == "pdbx_poly_seq_scheme" or parent->name() == "entity_poly_seq"))
 			continue;
 
-		size_t missing = 0;
+		std::size_t missing = 0;
 		category first_missing_rows(name());
 
 		for (auto r : *this)
@@ -962,7 +909,7 @@ condition category::get_parents_condition(row_handle rh, const category &parentC
 		{
 			condition cond;
 
-			for (size_t ix = 0; ix < link->m_child_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < link->m_child_keys.size(); ++ix)
 			{
 				auto childValue = rh[link->m_child_keys[ix]];
 
@@ -1004,7 +951,7 @@ condition category::get_children_condition(row_handle rh, const category &childC
 		{
 			condition cond;
 
-			for (size_t ix = 0; ix < link->m_parent_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < link->m_parent_keys.size(); ++ix)
 			{
 				auto childKey = link->m_child_keys[ix];
 				auto parentKey = link->m_parent_keys[ix];
@@ -1180,14 +1127,14 @@ class save_value
 	const T m_sv;
 };
 
-size_t category::erase(condition &&cond)
+std::size_t category::erase(condition &&cond)
 {
 	return erase(std::move(cond), {});
 }
 
-size_t category::erase(condition &&cond, std::function<void(row_handle)> &&visit)
+std::size_t category::erase(condition &&cond, std::function<void(row_handle)> &&visit)
 {
-	size_t result = 0;
+	std::size_t result = 0;
 
 	cond.prepare(*this);
 
@@ -1327,7 +1274,7 @@ std::string category::get_unique_value(std::string_view item_name)
 	if (result.empty())
 	{
 		// brain-dead implementation
-		for (size_t ix = 0; ix < size(); ++ix)
+		for (std::size_t ix = 0; ix < size(); ++ix)
 		{
 			// result = m_name + "-" + std::to_string(ix);
 			result = cif_id_for_number(ix);
@@ -1339,7 +1286,8 @@ std::string category::get_unique_value(std::string_view item_name)
 	return result;
 }
 
-void category::update_value(const std::vector<row_handle> &rows, std::string_view item_name, std::string_view value)
+void category::update_value(const std::vector<row_handle> &rows, std::string_view item_name,
+	value_provider_type &&value_provider)
 {
 	using namespace std::literals;
 
@@ -1352,40 +1300,29 @@ void category::update_value(const std::vector<row_handle> &rows, std::string_vie
 
 	auto &col = m_items[colIx];
 
+	// this is expensive, but better throw early on
 	// check the value
 	if (col.m_validator)
 	{
-		std::error_code ec;
-		col.m_validator->validate_value(value, ec);
-		if (ec)
-			throw validation_exception(ec, m_name, item_name);
-	}
-
-	// first some sanity checks, what was the old value and is it the same for all rows?
-	std::string oldValue{ rows.front()[item_name].text() };
-	for (auto row : rows)
-	{
-		if (oldValue != row[item_name].text())
+		for (auto row : rows)
 		{
-			std::ostringstream os;
+			std::string value{ value_provider(row[item_name].text()) };
 
-			os << "Inconsistent old values in update_value, trying to set " << std::quoted(value)
-			   << " as value for item " << item_name << " in category " << m_name;
-
-			throw std::runtime_error(os.str());
+			std::error_code ec;
+			col.m_validator->validate_value(value, ec);
+			if (ec)
+				throw validation_exception(ec, m_name, item_name);
 		}
 	}
 
-	if (oldValue == value) // no need to do anything
-		return;
-
-	// update rows, but do not cascade
-	for (auto row : rows)
-		row.assign(colIx, value, false);
-
-	// see if we need to update any child categories that depend on this value
+	// update and see if we need to update any child categories that depend on this value
 	for (auto parent : rows)
 	{
+		std::string oldValue{ parent[item_name].text() };
+		std::string value{ value_provider(oldValue) };
+
+		update_value(parent.get_row(), colIx, value, false, false);
+
 		for (auto &&[childCat, linked] : m_child_links)
 		{
 			if (std::find(linked->m_parent_keys.begin(), linked->m_parent_keys.end(), item_name) == linked->m_parent_keys.end())
@@ -1394,7 +1331,7 @@ void category::update_value(const std::vector<row_handle> &rows, std::string_vie
 			condition cond;
 			std::string childItemName;
 
-			for (size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
 			{
 				std::string pk = linked->m_parent_keys[ix];
 				std::string ck = linked->m_child_keys[ix];
@@ -1425,7 +1362,7 @@ void category::update_value(const std::vector<row_handle> &rows, std::string_vie
 			{
 				condition cond_c;
 
-				for (size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
+				for (std::size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
 				{
 					std::string pk = linked->m_parent_keys[ix];
 					std::string ck = linked->m_child_keys[ix];
@@ -1443,7 +1380,7 @@ void category::update_value(const std::vector<row_handle> &rows, std::string_vie
 				// oops, we need to split this child, unless a row already exists for the new value
 				condition check;
 
-				for (size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
+				for (std::size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
 				{
 					std::string pk = linked->m_parent_keys[ix];
 					std::string ck = linked->m_child_keys[ix];
@@ -1507,8 +1444,7 @@ void category::update_value(row *row, uint16_t item, std::string_view value, boo
 	// before updating
 
 	bool reinsert = false;
-	if (updateLinked and // an update of an Item's value
-		m_index != nullptr and key_item_indices().count(item))
+	if (m_index != nullptr and key_item_indices().count(item))
 	{
 		reinsert = m_index->find(*this, row);
 		if (reinsert)
@@ -1539,7 +1475,7 @@ void category::update_value(row *row, uint16_t item, std::string_view value, boo
 			condition cond;
 			std::string childItemName;
 
-			for (size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
 			{
 				std::string pk = linked->m_parent_keys[ix];
 				std::string ck = linked->m_child_keys[ix];
@@ -1576,7 +1512,7 @@ void category::update_value(row *row, uint16_t item, std::string_view value, boo
 
 			condition cond_n;
 
-			for (size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < linked->m_parent_keys.size(); ++ix)
 			{
 				std::string pk = linked->m_parent_keys[ix];
 				std::string ck = linked->m_child_keys[ix];
@@ -1722,7 +1658,7 @@ category::iterator category::insert_impl(const_iterator pos, row *n)
 			m_index->insert(*this, n);
 
 		// insert at end, most often this is the case
-		if (pos.m_current == nullptr)
+		if (pos.m_current.m_row == nullptr)
 		{
 			if (m_head == nullptr)
 				m_tail = m_head = n;
@@ -1733,7 +1669,7 @@ category::iterator category::insert_impl(const_iterator pos, row *n)
 		{
 			assert(m_head != nullptr);
 
-			if (pos.m_current == m_head)
+			if (pos.m_current.m_row == m_head)
 				m_head = n->m_next = m_head;
 			else
 				n = n->m_next = m_head->m_next;
@@ -1761,6 +1697,12 @@ void category::swap_item(uint16_t item_ix, row_handle &a, row_handle &b)
 	auto &ra = *a.m_row;
 	auto &rb = *b.m_row;
 
+	while (ra.size() <= item_ix)
+		ra.emplace_back("");
+
+	while (rb.size() <= item_ix)
+		rb.emplace_back("");
+
 	std::swap(ra.at(item_ix), rb.at(item_ix));
 }
 
@@ -1783,7 +1725,7 @@ void category::sort(std::function<int(row_handle, row_handle)> f)
 	m_tail = rows.back().get_row();
 
 	auto r = m_head;
-	for (size_t i = 1; i < rows.size(); ++i)
+	for (std::size_t i = 1; i < rows.size(); ++i)
 		r = r->m_next = rows[i].get_row();
 	r->m_next = nullptr;
 
@@ -1799,7 +1741,7 @@ void category::reorder_by_index()
 
 namespace detail
 {
-	size_t write_value(std::ostream &os, std::string_view value, size_t offset, size_t width, bool right_aligned)
+	std::size_t write_value(std::ostream &os, std::string_view value, std::size_t offset, std::size_t width, bool right_aligned)
 	{
 		if (value.find('\n') != std::string::npos or width == 0 or value.length() > 132) // write as text item
 		{
@@ -1960,7 +1902,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 	{
 		os << "loop_\n";
 
-		std::vector<size_t> itemWidths(m_items.size());
+		std::vector<std::size_t> itemWidths(m_items.size());
 
 		for (auto cix : order)
 		{
@@ -1982,7 +1924,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 
 				if (v->text().find('\n') == std::string_view::npos)
 				{
-					size_t l = v->text().length();
+					std::size_t l = v->text().length();
 
 					if (not sac_parser::is_unquoted_string(v->text()))
 						l += 2;
@@ -1998,11 +1940,11 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 
 		for (auto r = m_head; r != nullptr; r = r->m_next) // loop over rows
 		{
-			size_t offset = 0;
+			std::size_t offset = 0;
 
 			for (uint16_t cix : order)
 			{
-				size_t w = itemWidths[cix];
+				std::size_t w = itemWidths[cix];
 
 				std::string_view s;
 				auto iv = r->get(cix);
@@ -2012,7 +1954,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 				if (s.empty())
 					s = "?";
 
-				size_t l = s.length();
+				std::size_t l = s.length();
 				if (not sac_parser::is_unquoted_string(s))
 					l += 2;
 				if (l < w)
@@ -2040,7 +1982,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 	else
 	{
 		// first find the indent level
-		size_t l = 0;
+		std::size_t l = 0;
 
 		for (auto &col : m_items)
 		{
@@ -2052,7 +1994,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 
 		l += 3;
 
-		size_t width = 1;
+		std::size_t width = 1;
 
 		for (auto cix : order)
 		{
@@ -2067,7 +2009,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 			if (s.empty())
 				s = "?";
 
-			size_t l2 = s.length();
+			std::size_t l2 = s.length();
 
 			if (not sac_parser::is_unquoted_string(s))
 				l2 += 2;
@@ -2093,7 +2035,7 @@ void category::write(std::ostream &os, const std::vector<uint16_t> &order, bool
 			if (s.empty())
 				s = "?";
 
-			size_t offset = l;
+			std::size_t offset = l;
 			if (s.length() + l >= kMaxLineLength)
 			{
 				os << '\n';
@@ -2133,7 +2075,7 @@ bool category::operator==(const category &rhs) const
 	typedef std::function<int(std::string_view, std::string_view)> compType;
 	std::vector<std::tuple<std::string, compType>> item_names;
 	std::vector<std::string> keys;
-	std::vector<size_t> keyIx;
+	std::vector<std::size_t> keyIx;
 
 	if (catValidator == nullptr)
 	{

src/compound.cpp

@@ -496,7 +496,7 @@ compound *compound_factory_impl::create(const std::string &id)
 		m_index = parser.index_datablocks();
 
 		if (cif::VERBOSE > 1)
-			std::cout << " done" << std::endl;
+			std::cout << " done\n";
 
 		// reload the resource, perhaps this should be improved...
 		if (m_file.empty())
@@ -519,7 +519,7 @@ compound *compound_factory_impl::create(const std::string &id)
 	parser.parse_single_datablock(id, m_index);
 
 	if (cif::VERBOSE > 1)
-		std::cout << " done" << std::endl;
+		std::cout << " done\n";
 
 	if (not file.empty())
 	{
@@ -545,12 +545,29 @@ class local_compound_factory_impl : public compound_factory_impl
 		: compound_factory_impl(next)
 		, m_local_file(file)
 	{
+		// const std::regex peptideRx("(?:[lmp]-)?peptide", std::regex::icase);
+
+		// for (const auto &[id, name, threeLetterCode, group] :
+		// 	file["comp_list"]["chem_comp"].rows<std::string, std::string, std::string, std::string>("id", "name", "three_letter_code", "group"))
+		// {
+		// 	auto &rdb = m_local_file["comp_" + id];
+		// 	if (rdb.empty())
+		// 	{
+		// 		// std::cerr << "Missing data in restraint file for id " + id + '\n';
+		// 		continue;
+		// 	}
+
+		// 	construct_compound(rdb, id, name, threeLetterCode, group);
+		// }
 	}
 
 	compound *create(const std::string &id) override;
 
   private:
-	const cif::file &m_local_file;
+
+	compound *construct_compound(const datablock &db, const std::string &id, const std::string &name, const std::string &three_letter_code, const std::string &group);
+
+	cif::file m_local_file;
 };
 
 compound *local_compound_factory_impl::create(const std::string &id)
@@ -561,12 +578,21 @@ compound *local_compound_factory_impl::create(const std::string &id)
 	{
 		if (db.name() == "comp_" + id)
 		{
-			cif::datablock db_copy(db);
+			auto chem_comp = db.get("chem_comp");
+			if (not chem_comp)
+				break;
 
-			result = new compound(db_copy, 1);
+			try
+			{
+				const auto &[id, name, threeLetterCode, group] =
+					chem_comp->front().get<std::string, std::string, std::string, std::string>("id", "name", "three_letter_code", "group");
 
-			std::shared_lock lock(mMutex);
-			m_compounds.push_back(result);
+				result = construct_compound(db, id, name, threeLetterCode, group);
+			}
+			catch (const std::exception &ex)
+			{
+				std::throw_with_nested(std::runtime_error("Error loading compound " + id));
+			}
 
 			break;
 		}
@@ -575,6 +601,100 @@ compound *local_compound_factory_impl::create(const std::string &id)
 	return result;
 }
 
+compound *local_compound_factory_impl::construct_compound(const datablock &rdb, const std::string &id,
+	const std::string &name, const std::string &three_letter_code, const std::string &group)
+{
+	cif::datablock db(id);
+
+	float formula_weight = 0;
+	int formal_charge = 0;
+	std::map<std::string,std::size_t> formula_data;
+
+	for (std::size_t ord = 1; const auto &[atom_id, type_symbol, type, charge, x, y, z] :
+		rdb["chem_comp_atom"].rows<std::string, std::string, std::string, int, float, float, float>(
+			"atom_id", "type_symbol", "type", "charge", "x", "y", "z"))
+	{
+		auto atom = cif::atom_type_traits(type_symbol);
+		formula_weight += atom.weight();
+
+		formula_data[type_symbol] += 1;
+
+		db["chem_comp_atom"].emplace({
+			{ "comp_id", id },
+			{ "atom_id", atom_id },
+			{ "type_symbol", type_symbol },
+			{ "charge", charge },
+			{ "model_Cartn_x",  x, 3 },
+			{ "model_Cartn_y",  y, 3 },
+			{ "model_Cartn_z",  z, 3 },
+			{ "pdbx_ordinal", ord++ }
+		});
+
+		formal_charge += charge;
+	}
+
+	for (std::size_t ord = 1; const auto &[atom_id_1, atom_id_2, type, aromatic] :
+		rdb["chem_comp_bond"].rows<std::string, std::string, std::string, bool>("atom_id_1", "atom_id_2", "type", "aromatic"))
+	{
+		std::string value_order("SING");
+
+		if (cif::iequals(type, "single") or cif::iequals(type, "sing"))
+			value_order = "SING";
+		else if (cif::iequals(type, "double") or cif::iequals(type, "doub"))
+			value_order = "DOUB";
+		else if (cif::iequals(type, "triple") or cif::iequals(type, "trip"))
+			value_order = "TRIP";
+
+		db["chem_comp_bond"].emplace({
+			{ "comp_id", id },
+			{ "atom_id_1", atom_id_1 },
+			{ "atom_id_2", atom_id_2 },
+			{ "value_order", value_order },
+			{ "pdbx_aromatic_flag", aromatic },
+			// TODO: fetch stereo_config info from chem_comp_chir
+			{ "pdbx_ordinal", ord++ }
+		});
+	}
+
+	db.emplace_back(rdb["pdbx_chem_comp_descriptor"]);
+
+	std::string formula;
+	for (bool first = true; const auto &[symbol, count]: formula_data)
+	{
+		if (std::exchange(first, false))
+			formula += ' ';
+		formula += symbol;
+		if (count > 1)
+			formula += std::to_string(count);
+	}
+
+	std::string type;
+	if (cif::iequals(group, "peptide") or cif::iequals(group, "l-peptide") or cif::iequals(group, "l-peptide linking"))
+		type = "L-PEPTIDE LINKING";
+	else if (cif::iequals(group, "dna"))
+		type = "DNA LINKING";
+	else if (cif::iequals(group, "rna"))
+		type = "RNA LINKING";
+	else
+		type = "NON-POLYMER";
+
+	db["chem_comp"].emplace({
+		{ "id", id },
+		{ "name", name },
+		{ "type", type },
+		{ "formula", formula },
+		{ "pdbx_formal_charge", formal_charge },
+		{ "formula_weight", formula_weight },
+		{ "three_letter_code", three_letter_code }
+	});
+
+	std::shared_lock lock(mMutex);
+
+	auto result = new compound(db);
+	m_compounds.push_back(result);
+	return result;
+}
+
 // --------------------------------------------------------------------
 
 std::unique_ptr<compound_factory> compound_factory::s_instance;
@@ -757,8 +877,8 @@ void compound_factory::report_missing_compound(std::string_view compound_id)
 				  << "update=true\n\n"
 				  << "If you do not have a working cron script, you can manually update the files\n"
 				  << "in /var/cache/libcifpp using the following commands:\n\n"
-				  << "curl -o " << CACHE_DIR << "/components.cif https://files.wwpdb.org/pub/pdb/data/monomers/components.cif.gz\n"
-				  << "curl -o " << CACHE_DIR << "/mmcif_pdbx.dic https://mmcif.wwpdb.org/dictionaries/ascii/mmcif_pdbx_v50.dic.gz\n"
+				  << "curl -o " << CACHE_DIR << "/components.cif https://files.wwpdb.org/pub/pdb/data/monomers/components.cif\n"
+				  << "curl -o " << CACHE_DIR << "/mmcif_pdbx.dic https://mmcif.wwpdb.org/dictionaries/ascii/mmcif_pdbx_v50.dic\n"
 				  << "curl -o " << CACHE_DIR << "/mmcif_ma.dic https://github.com/ihmwg/ModelCIF/raw/master/dist/mmcif_ma.dic\n\n";
 #endif
 

src/condition.cpp

@@ -76,6 +76,20 @@ namespace detail
 		return this;
 	}
 
+	condition_impl *key_equals_number_condition_impl::prepare(const category &c)
+	{
+		m_item_ix = c.get_item_ix(m_item_name);
+
+		if (c.get_cat_validator() != nullptr and
+			c.key_item_indices().contains(m_item_ix) and
+			c.key_item_indices().size() == 1)
+		{
+			m_single_hit = c[{ { m_item_name, m_value } }];
+		}
+
+		return this;
+	}
+
 	bool found_in_range(condition_impl *c, std::vector<and_condition_impl *>::iterator b, std::vector<and_condition_impl *>::iterator e)
 	{
 		bool result = true;

src/datablock.cpp

@@ -38,21 +38,6 @@ datablock::datablock(const datablock &db)
 		cat.update_links(*this);
 }
 
-datablock &datablock::operator=(const datablock &db)
-{
-	if (this != &db)
-	{
-		std::list<category>::operator=(db);
-		m_name = db.m_name;
-		m_validator = db.m_validator;
-
-		for (auto &cat : *this)
-			cat.update_links(*this);
-	}
-
-	return *this;
-}
-
 void datablock::set_validator(const validator *v)
 {
 	m_validator = v;
@@ -64,6 +49,7 @@ void datablock::set_validator(const validator *v)
 	}
 	catch (const std::exception &)
 	{
+		m_validator = nullptr;
 		throw_with_nested(std::runtime_error("Error while setting validator in datablock " + m_name));
 	}
 }
@@ -123,6 +109,9 @@ bool datablock::validate_links() const
 {
 	bool result = true;
 
+	for (auto &cat : *this)
+		const_cast<category &>(cat).update_links(*this);
+
 	for (auto &cat : *this)
 		result = cat.validate_links() and result;
 
@@ -190,6 +179,11 @@ std::tuple<datablock::iterator, bool> datablock::emplace(std::string_view name)
 	}
 
 	assert(i != end());
+
+	// links may have changed...
+	for (auto &cat : *this)
+		cat.update_links(*this);
+
 	return std::make_tuple(i, is_new);
 }
 
@@ -282,7 +276,11 @@ void datablock::write(std::ostream &os) const
 		cat_order_t cat_order;
 
 		for (auto &cat : *this)
+		{
+			if (cat.name() == "entry" or cat.name() == "audit_conform")
+				continue;
 			cat_order.emplace_back(cat.name(), -1, false);
+		}
 
 		for (auto i = cat_order.begin(); i != cat_order.end(); ++i)
 			calculate_cat_order(cat_order, i, *m_validator);
@@ -292,25 +290,18 @@ void datablock::write(std::ostream &os) const
 			const auto &[cat_a, count_a, on_stack_a] = a;
 			const auto &[cat_b, count_b, on_stack_b] = b;
 
-			int d = 0;
-
-			if (cat_a == "audit_conform")
-				d = -1;
-			else if (cat_b == "audit_conform")
-				d = 1;
-			else if (cat_a == "entry")
-				d = -1;
-			else if (cat_b == "entry")
-				d = 1;
-			else
-			{
-				d = std::get<1>(a) - std::get<1>(b);
-				if (d == 0)
-					d = cat_b.compare(cat_a);
-			}
+			int d = std::get<1>(a) - std::get<1>(b);
+			if (d == 0)
+				d = cat_b.compare(cat_a);
 
 			return d < 0; });
 
+		if (auto entry = get("entry"); entry != nullptr)
+			entry->write(os);
+
+		if (auto audit_conform = get("audit_conform"); audit_conform != nullptr)
+			audit_conform->write(os);
+
 		for (auto &&[cat, count, on_stack] : cat_order)
 			get(cat)->write(os);
 	}
@@ -320,20 +311,13 @@ void datablock::write(std::ostream &os) const
 		// and if it exists, _AND_ we have a Validator, write out the
 		// audit_conform record.
 
-		for (auto &cat : *this)
-		{
-			if (cat.name() != "entry")
-				continue;
-
-			cat.write(os);
-
-			break;
-		}
+		if (auto entry = get("entry"); entry != nullptr)
+			entry->write(os);
 
 		// If the dictionary declares an audit_conform category, put it in,
 		// but only if it does not exist already!
-		if (get("audit_conform"))
-			get("audit_conform")->write(os);
+		if (auto audit_conform = get("audit_conform"); audit_conform != nullptr)
+			audit_conform->write(os);
 
 		for (auto &cat : *this)
 		{
@@ -348,7 +332,7 @@ void datablock::write(std::ostream &os, const std::vector<std::string> &item_nam
 	os << "data_" << m_name << '\n'
 	   << "# \n";
 
-	std::vector<std::string> cat_order;
+	std::vector<std::string> cat_order{ "entry", "audit_conform" };
 	for (auto &o : item_name_order)
 	{
 		std::string cat_name, item_name;

src/dictionary_parser.cpp

@@ -308,17 +308,17 @@ class dictionary_parser : public parser
 
 		using key_type = std::tuple<std::string, std::string, int>;
 
-		std::map<key_type, size_t> linkIndex;
+		std::map<key_type, std::size_t> linkIndex;
 
 		// Each link group consists of a set of keys
 		std::vector<std::tuple<std::vector<std::string>, std::vector<std::string>>> linkKeys;
 
-		auto addLink = [&](size_t ix, const std::string &pk, const std::string &ck)
+		auto addLink = [&](std::size_t ix, const std::string &pk, const std::string &ck)
 		{
 			auto &&[pkeys, ckeys] = linkKeys.at(ix);
 
 			bool found = false;
-			for (size_t i = 0; i < pkeys.size(); ++i)
+			for (std::size_t i = 0; i < pkeys.size(); ++i)
 			{
 				if (pkeys[i] == pk and ckeys[i] == ck)
 				{
@@ -357,7 +357,7 @@ class dictionary_parser : public parser
 				linkKeys.push_back({});
 			}
 
-			size_t ix = linkIndex.at(key);
+			std::size_t ix = linkIndex.at(key);
 			addLink(ix, piv->m_item_name, civ->m_item_name);
 		}
 
@@ -385,7 +385,7 @@ class dictionary_parser : public parser
 					linkKeys.push_back({});
 				}
 
-				size_t ix = linkIndex.at(key);
+				std::size_t ix = linkIndex.at(key);
 				addLink(ix, piv->m_item_name, civ->m_item_name);
 			}
 		}

src/file.cpp

@@ -31,11 +31,32 @@ namespace cif
 {
 
 // --------------------------------------------------------------------
+// TODO: This is wrong. A validator should be assigned to datablocks,
+// not to a file. Since audit_conform is a category specifying the
+// content of a datablock. Not the entire file.
+
 void file::set_validator(const validator *v)
 {
 	m_validator = v;
-	for (auto &db : *this)
-		db.set_validator(v);
+	for (bool first = true; auto &db : *this)
+	{
+		try
+		{
+			db.set_validator(v);
+		}
+		catch (const std::exception &e)
+		{
+			if (first)
+				throw;
+
+			// Accept failure on secondary datablocks
+			// now that many mmCIF files have invalid
+			// restraint data concatenated.
+			std::cerr << e.what() << '\n';
+		}
+
+		first = false;
+	}
 }
 
 bool file::is_valid() const
@@ -78,12 +99,12 @@ bool file::validate_links() const
 {
 	if (m_validator == nullptr)
 		std::runtime_error("No validator loaded explicitly, cannot continue");
-	
+
 	bool result = true;
 
 	for (auto &db : *this)
 		result = db.validate_links() and result;
-	
+
 	return result;
 }
 
@@ -97,7 +118,7 @@ void file::load_dictionary()
 			std::string name = audit_conform->front().get<std::string>("dict_name");
 
 			if (name == "mmcif_pdbx_v50")
-				name = "mmcif_pdbx.dic";	// we had a bug here in libcifpp... 
+				name = "mmcif_pdbx.dic"; // we had a bug here in libcifpp...
 
 			if (not name.empty())
 			{
@@ -125,7 +146,8 @@ void file::load_dictionary(std::string_view name)
 
 bool file::contains(std::string_view name) const
 {
-	return std::find_if(begin(), end(), [name](const datablock &db) { return iequals(db.name(), name); }) != end();
+	return std::find_if(begin(), end(), [name](const datablock &db)
+			   { return iequals(db.name(), name); }) != end();
 }
 
 datablock &file::operator[](std::string_view name)

src/item.cpp

@@ -52,10 +52,10 @@ std::string_view item_handle::text() const
 	return {};
 }
 
-void item_handle::assign_value(const item &v)
+void item_handle::assign_value(std::string_view value)
 {
 	assert(not m_row_handle.empty());
-	m_row_handle.assign(m_item_ix, v.value(), true);
+	m_row_handle.assign(m_item_ix, value, true);
 }
 
 void item_handle::swap(item_handle &b)

src/pdb/cif2pdb.cpp

@@ -182,7 +182,7 @@ std::vector<std::string> MapAsymIDs2ChainIDs(const std::vector<std::string> &asy
 }
 
 // support for wrapping text using a 'continuation marker'
-size_t WriteContinuedLine(std::ostream &pdbFile, std::string header, int &count, int cLen, std::string text, std::string::size_type lStart = 0)
+std::size_t WriteContinuedLine(std::ostream &pdbFile, std::string header, int &count, int cLen, std::string text, std::string::size_type lStart = 0)
 {
 	if (lStart == 0)
 	{
@@ -217,15 +217,15 @@ size_t WriteContinuedLine(std::ostream &pdbFile, std::string header, int &count,
 	return lines.size();
 }
 
-size_t WriteOneContinuedLine(std::ostream &pdbFile, std::string header, int cLen, std::string line, int lStart = 0)
+std::size_t WriteOneContinuedLine(std::ostream &pdbFile, std::string header, int cLen, std::string line, int lStart = 0)
 {
 	int count = 0;
 	return WriteContinuedLine(pdbFile, header, count, cLen, line, lStart);
 }
 
-size_t WriteCitation(std::ostream &pdbFile, const datablock &db, row_handle r, int reference)
+std::size_t WriteCitation(std::ostream &pdbFile, const datablock &db, row_handle r, int reference)
 {
-	size_t result = 0;
+	std::size_t result = 0;
 
 	std::string s1;
 
@@ -560,7 +560,7 @@ void WriteTitle(std::ostream &pdbFile, const datablock &db)
 			std::string cs = ++continuation > 1 ? std::to_string(continuation) : std::string();
 
 			pdbFile << cif::format(kRevDatFmt, revNum, cs, date, db.name(), modType);
-			for (size_t i = 0; i < 4; ++i)
+			for (std::size_t i = 0; i < 4; ++i)
 				pdbFile << cif::format(" %-6.6s", (i < types.size() ? types[i] : std::string()));
 			pdbFile << '\n';
 
@@ -681,7 +681,7 @@ class Fi : public FBase
 		{
 			long l = 0;
 			auto r = std::from_chars(s.data(), s.data() + s.length(), l);
-			if (r.ec != std::errc())
+			if ((bool)r.ec)
 			{
 				if (VERBOSE > 0)
 					std::cerr << "Failed to write '" << s << "' as a long from field " << mField << ", this indicates an error in the code for writing PDB files\n";
@@ -719,7 +719,7 @@ class Ff : public FBase
 
 			double d = 0;
 			auto r = cif::from_chars(s.data(), s.data() + s.length(), d);
-			if (r.ec != std::errc())
+			if ((bool)r.ec)
 			{
 				if (VERBOSE > 0)
 					std::cerr << "Failed to write '" << s << "' as a double from field " << mField << ", this indicates an error in the code for writing PDB files\n";
@@ -748,7 +748,7 @@ class Fs : public FBase
 	virtual void out(std::ostream &os)
 	{
 		std::string s{ text() };
-		size_t width = os.width();
+		std::size_t width = os.width();
 
 		if (s.empty())
 		{
@@ -2368,7 +2368,7 @@ void WriteRemark280(std::ostream &pdbFile, const datablock &db)
 
 			const char *keys[] = { "pdbx_details", "ph", "method", "temp" };
 
-			for (size_t i = 0; i < (sizeof(keys) / sizeof(const char *)); ++i)
+			for (std::size_t i = 0; i < (sizeof(keys) / sizeof(const char *)); ++i)
 			{
 				const char *c = keys[i];
 
@@ -2634,7 +2634,7 @@ void WriteRemark470(std::ostream &pdbFile, const datablock &db)
 			{
 				pdbFile << cif::format("REMARK 470 %3.3s %3.3s %1.1s%4d%1.1s  ", modelNr, resName, chainID, seqNr, iCode) << "  ";
 
-				for (size_t i = 0; i < 6 and not a.second.empty(); ++i)
+				for (std::size_t i = 0; i < 6 and not a.second.empty(); ++i)
 				{
 					pdbFile << cif2pdbAtomName(a.second.front(), resName, db) << ' ';
 					a.second.pop_front();
@@ -3284,7 +3284,7 @@ int WriteMiscellaneousFeatures(std::ostream &pdbFile, const datablock &db)
 		std::string siteID = std::get<0>(s);
 		std::deque<std::string> &res = std::get<1>(s);
 
-		size_t numRes = res.size();
+		std::size_t numRes = res.size();
 
 		int nr = 1;
 		while (res.empty() == false)
@@ -3393,7 +3393,7 @@ std::tuple<int, int> WriteCoordinatesForModel(std::ostream &pdbFile, const datab
 		{
 			int nr = 0;
 			auto r = std::from_chars(modelNum.data(), modelNum.data() + modelNum.length(), nr);
-			if (r.ec != std::errc())
+			if ((bool)r.ec)
 			{
 				if (VERBOSE > 0)
 					std::cerr << "Model number '" << modelNum << "' is not a valid integer\n";

src/pdb/pdb2cif.cpp

@@ -160,7 +160,7 @@ PDBRecord::~PDBRecord()
 {
 }
 
-void *PDBRecord::operator new(size_t size, size_t vLen)
+void *PDBRecord::operator new(std::size_t size, std::size_t vLen)
 {
 	return malloc(size + vLen + 1);
 }
@@ -170,7 +170,7 @@ void PDBRecord::operator delete(void *p)
 	free(p);
 }
 
-void PDBRecord::operator delete(void *p, size_t vLen)
+void PDBRecord::operator delete(void *p, std::size_t vLen)
 {
 	free(p);
 }
@@ -180,7 +180,7 @@ bool PDBRecord::is(const char *name) const
 	return iequals(mName, name);
 }
 
-char PDBRecord::vC(size_t column)
+char PDBRecord::vC(std::size_t column)
 {
 	char result = ' ';
 	if (column - 7 < mVlen)
@@ -188,7 +188,7 @@ char PDBRecord::vC(size_t column)
 	return result;
 }
 
-std::string PDBRecord::vS(size_t columnFirst, size_t columnLast)
+std::string PDBRecord::vS(std::size_t columnFirst, std::size_t columnLast)
 {
 	std::string result;
 
@@ -272,7 +272,7 @@ int PDBRecord::vI(int columnFirst, int columnLast)
 	return result;
 }
 
-std::string PDBRecord::vF(size_t columnFirst, size_t columnLast)
+std::string PDBRecord::vF(std::size_t columnFirst, std::size_t columnLast)
 {
 	// for now... TODO: check format?
 	return vS(columnFirst, columnLast);
@@ -780,17 +780,17 @@ class PDBFileParser
 
 	// ----------------------------------------------------------------
 
-	char vC(size_t column) const
+	char vC(std::size_t column) const
 	{
 		return mRec->vC(column);
 	}
 
-	std::string vS(size_t columnFirst, size_t columnLast = std::numeric_limits<size_t>::max()) const
+	std::string vS(std::size_t columnFirst, std::size_t columnLast = std::numeric_limits<std::size_t>::max()) const
 	{
 		return mRec->vS(columnFirst, columnLast);
 	}
 
-	std::string vF(size_t columnFirst, size_t columnLast) const
+	std::string vF(std::size_t columnFirst, std::size_t columnLast) const
 	{
 		return mRec->vF(columnFirst, columnLast);
 	}
@@ -847,7 +847,7 @@ class PDBFileParser
 	void ParseRemarks();
 
 	//	void ParseRemark3();
-	//	size_t ParseRemark3(const std::string& program, const Remark3Template templ[], size_t N);
+	//	std::size_t ParseRemark3(const std::string& program, const Remark3Template templ[], std::size_t N);
 	//	std::string NextRemark3Line();
 
 	void ParseRemark200();
@@ -1132,7 +1132,7 @@ void PDBFileParser::PreParseInput(std::istream &is)
 		if (not cs.empty())
 		{
 			auto r = std::from_chars(cs.data(), cs.data() + cs.length(), result);
-			if (r.ec != std::errc())
+			if ((bool)r.ec)
 				throw std::runtime_error("Continuation std::string '" + cs + "' is not valid");
 		}
 
@@ -1320,7 +1320,7 @@ void PDBFileParser::PreParseInput(std::istream &is)
 		{
 			std::string siteName = value.substr(5, 3);
 			cif::trim_right(value);
-			size_t n = value.length() - 12;
+			std::size_t n = value.length() - 12;
 			value += std::string(11 - (n % 11), ' ');
 
 			while (lookahead.substr(0, 6) == type and lookahead.substr(11, 3) == siteName)
@@ -1397,7 +1397,7 @@ void PDBFileParser::PreParseInput(std::istream &is)
 			{
 				auto f = cur->vF(74, 78);
 				auto r = cif::from_chars(f.data(), f.data() + f.length(), link.distance);
-				if (r.ec != std::errc() and cif::VERBOSE > 0)
+				if ((bool)r.ec and cif::VERBOSE > 0)
 					std::cerr << "Error parsing link distance at line " << cur->mLineNr << '\n';
 			}
 			//	74 – 78         Real(5.2)      Length          Link distance
@@ -4202,7 +4202,7 @@ void PDBFileParser::ConstructEntities()
 
 			chains.push_back(std::string{ chain.mDbref.chainID });
 
-			size_t seqLen = 0, seqCanLen = 0;
+			std::size_t seqLen = 0, seqCanLen = 0;
 
 			for (auto &res : chain.mSeqres)
 			{
@@ -4265,7 +4265,7 @@ void PDBFileParser::ConstructEntities()
 			}
 
 			auto cat_ps = getCategory("entity_poly_seq");
-			for (size_t i = 0; i < chain.mSeqres.size(); ++i)
+			for (std::size_t i = 0; i < chain.mSeqres.size(); ++i)
 			{
 				auto &rs = chain.mSeqres[i];
 
@@ -4335,7 +4335,7 @@ void PDBFileParser::ConstructEntities()
 	std::map<std::tuple<std::string, std::string>, int> ndbSeqNum; // for nonpoly scheme
 	std::map<std::string, int> entityAuthSeqNum;                   // for nonpoly scheme too
 
-	for (size_t i = 0; i < mHets.size(); ++i)
+	for (std::size_t i = 0; i < mHets.size(); ++i)
 	{
 		auto &heti = mHets[i];
 
@@ -4669,47 +4669,6 @@ void PDBFileParser::ConstructEntities()
 			}
 		}
 	}
-
-	// Finish by calculating the formula_weight for each entity
-	for (auto entity : *getCategory("entity"))
-	{
-		auto entity_id = entity["id"].as<std::string>();
-		float formula_weight = 0;
-
-		if (entity["type"] == "polymer")
-		{
-			int n = 0;
-
-			for (std::string comp_id : getCategory("pdbx_poly_seq_scheme")->find<std::string>(cif::key("entity_id") == entity_id, "mon_id"))
-			{
-				auto compound = cif::compound_factory::instance().create(comp_id);
-				assert(compound);
-				if (not compound)
-					throw std::runtime_error("missing information for compound " + comp_id);
-				formula_weight += compound->formula_weight();
-				++n;
-			}
-
-			formula_weight -= (n - 1) * 18.015;
-		}
-		else if (entity["type"] == "water")
-			formula_weight = 18.015;
-		else
-		{
-			auto comp_id = getCategory("pdbx_nonpoly_scheme")->find_first<std::optional<std::string>>(cif::key("entity_id") == entity_id, "mon_id");
-			if (comp_id.has_value())
-			{
-				auto compound = cif::compound_factory::instance().create(*comp_id);
-				assert(compound);
-				if (not compound)
-					throw std::runtime_error("missing information for compound " + *comp_id);
-				formula_weight = compound->formula_weight();
-			}
-		}
-
-		if (formula_weight > 0)
-			entity.assign({ { "formula_weight", formula_weight, 3 } });
-	}
 }
 
 void PDBFileParser::ConstructSugarTrees(int &asymNr)
@@ -5347,7 +5306,7 @@ void PDBFileParser::ParseConnectivtyAnnotation()
 
 				double d;
 				auto r = cif::from_chars(distance.data(), distance.data() + distance.length(), d);
-				if (r.ec != std::errc())
+				if ((bool)r.ec)
 				{
 					if (cif::VERBOSE > 0)
 						std::cerr << "Distance value '" << distance << "' is not a valid float in LINK record\n";
@@ -5753,7 +5712,7 @@ void PDBFileParser::ParseCoordinate(int modelNr)
 	stable_sort(atoms.begin(), atoms.end(), rLess);
 
 	// now reiterate the atoms to reorder alternates
-	for (size_t i = 0; i + 1 < atoms.size(); ++i)
+	for (std::size_t i = 0; i + 1 < atoms.size(); ++i)
 	{
 		char altLoc = std::get<3>(atoms[i])->vC(17);
 
@@ -6402,7 +6361,7 @@ bool PDBFileParser::PDBChain::SameSequence(const PDBChain &rhs) const
 {
 	bool result = mSeqres.size() == rhs.mSeqres.size();
 
-	for (size_t i = 0; result and i < mSeqres.size(); ++i)
+	for (std::size_t i = 0; result and i < mSeqres.size(); ++i)
 		result = mSeqres[i].mMonID == rhs.mSeqres[i].mMonID;
 
 	return result;
@@ -6454,11 +6413,11 @@ file read(std::istream &is)
 			{
 				std::throw_with_nested(std::runtime_error("Since the file did not start with a valid PDB HEADER line mmCIF was assumed, but that failed."));
 			}
-
-			// Since we're using the cif::pdb way of reading the file, the data may need
-			// reconstruction
-			reconstruct_pdbx(result);
 		}
+
+		// Since we're using the cif::pdb way of reading the file, the data may need
+		// reconstruction
+		reconstruct_pdbx(result);
 	}
 
 	// Must be a PDB like file, right?

src/pdb/pdb2cif_remark_3.cpp

@@ -980,8 +980,8 @@ std::string Remark3Parser::nextLine()
 
 	while (mRec != nullptr and mRec->is("REMARK   3"))
 	{
-		size_t valueIndent = 0;
-		for (size_t i = 4; i < mRec->mVlen; ++i)
+		std::size_t valueIndent = 0;
+		for (std::size_t i = 4; i < mRec->mVlen; ++i)
 		{
 			if (mRec->mValue[i] == ' ')
 				continue;
@@ -1232,7 +1232,7 @@ void Remark3Parser::storeCapture(const char *category, std::initializer_list<con
 			}
 			//			else if (iequals(category, "struct_ncs_dom"))
 			//			{
-			//				size_t id = cat.size() + 1;
+			//				std::size_t id = cat.size() + 1;
 			//
 			//				cat.emplace({
 			//					{ "id", id }

src/pdb/pdb_record.hpp

@@ -40,24 +40,24 @@ struct PDBRecord
 	PDBRecord *mNext;
 	uint32_t mLineNr;
 	char mName[11];
-	size_t mVlen;
+	std::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 new(std::size_t);
+	void *operator new(std::size_t size, std::size_t vLen);
 
 	void operator delete(void *p);
-	void operator delete(void *p, size_t vLen);
+	void operator delete(void *p, std::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());
+	char vC(std::size_t column);
+	std::string vS(std::size_t columnFirst, std::size_t columnLast = std::numeric_limits<std::size_t>::max());
 	int vI(int columnFirst, int columnLast);
-	std::string vF(size_t columnFirst, size_t columnLast);
+	std::string vF(std::size_t columnFirst, std::size_t columnLast);
 };
 
 } // namespace pdbx

src/pdb/reconstruct.cpp

@@ -92,6 +92,72 @@ condition get_condition(residue_key_type &k)
 
 // --------------------------------------------------------------------
 
+void checkEntities(datablock &db)
+{
+	using namespace cif::literals;
+
+	auto &cf = cif::compound_factory::instance();
+
+	for (auto entity : db["entity"].find("formula_weight"_key == null or "formula_weight"_key == 0))
+	{
+		const auto &[entity_id, type] = entity.get<std::string, std::string>("id", "type");
+
+		float formula_weight = 0;
+
+		if (type == "polymer")
+		{
+			int n = 0;
+
+			for (std::string comp_id : db["pdbx_poly_seq_scheme"].find<std::string>("entity_id"_key == entity_id, "mon_id"))
+			{
+				auto compound = cf.create(comp_id);
+				assert(compound);
+				if (not compound)
+					throw std::runtime_error("missing information for compound " + comp_id);
+				formula_weight += compound->formula_weight();
+				++n;
+			}
+
+			formula_weight -= (n - 1) * 18.015;
+		}
+		else if (type == "water")
+			formula_weight = 18.015;
+		else if (type == "branched")
+		{
+			int n = 0;
+
+			for (std::string comp_id : db["pdbx_entity_branch_list"].find<std::string>("entity_id"_key == entity_id, "comp_id"))
+			{
+				auto compound = cf.create(comp_id);
+				assert(compound);
+				if (not compound)
+					throw std::runtime_error("missing information for compound " + comp_id);
+				formula_weight += compound->formula_weight();
+				++n;
+			}
+
+			formula_weight -= (n - 1) * 18.015;
+		}
+		else if (type == "non-polymer")
+		{
+			auto comp_id = db["pdbx_nonpoly_scheme"].find_first<std::optional<std::string>>("entity_id"_key == entity_id, "mon_id");
+			if (comp_id.has_value())
+			{
+				auto compound = cf.create(*comp_id);
+				if (not compound)
+				{
+					std::cerr << "missing information for compound " << *comp_id << "\n";
+					continue;
+				}
+				formula_weight = compound->formula_weight();
+			}
+		}
+
+		if (formula_weight > 0)
+			entity.assign({ { "formula_weight", formula_weight, 3 } });
+	}
+}
+
 void createEntityIDs(datablock &db)
 {
 	// Suppose the file does not have entity ID's. We have to make up some
@@ -147,11 +213,11 @@ void createEntityIDs(datablock &db)
 		lastSeqID = seq_id;
 	}
 
-	std::map<size_t, std::string> entity_ids;
+	std::map<std::size_t, std::string> entity_ids;
 
 	atom_site.add_item("label_entity_id");
 
-	for (size_t i = 0; i < entities.size(); ++i)
+	for (std::size_t i = 0; i < entities.size(); ++i)
 	{
 		if (entity_ids.contains(i))
 			continue;
@@ -159,14 +225,14 @@ void createEntityIDs(datablock &db)
 		auto entity_id = std::to_string(i + 1);
 		entity_ids[i] = entity_id;
 
-		for (size_t j = i + 1; j < entities.size(); ++j)
+		for (std::size_t j = i + 1; j < entities.size(); ++j)
 		{
 			if (entities[i] == entities[j])
 				entity_ids[j] = entity_id;
 		}
 	}
 
-	for (size_t ix = 0; auto &e : entities)
+	for (std::size_t ix = 0; auto &e : entities)
 	{
 		auto k = e.front();
 		const auto &entity_id = entity_ids[ix++];
@@ -325,7 +391,7 @@ void checkChemCompRecords(datablock &db)
 				items.emplace_back(item{ "formula_weight", compound->formula_weight() });
 
 			if (not items.empty())
-				chem_comp_entry.assign(std::move(items));
+				chem_comp_entry.assign(items);
 		}
 	}
 }
@@ -348,6 +414,12 @@ void checkAtomRecords(datablock &db)
 	if (atom_site.contains(key("label_seq_id") < 0))
 		fixNegativeSeqID(atom_site);
 
+	std::set<int> polymer_entities;
+	for (int id : db["entity"].find<int>("type"_key == "polymer", "id"))
+		polymer_entities.insert(id);
+
+	std::set<std::string> missingCompounds;
+
 	for (auto row : atom_site)
 	{
 		residue_key_type k = row.get<std::optional<std::string>,
@@ -378,20 +450,31 @@ void checkAtomRecords(datablock &db)
 		std::string asym_id = get_asym_id(k);
 		std::string comp_id = get_comp_id(k);
 
-		bool is_peptide = cf.is_peptide(comp_id);
+		if (missingCompounds.contains(comp_id))
+			continue;
+
+		bool is_polymer = polymer_entities.contains(row["label_entity_id"].as<int>());
 		auto compound = cf.create(comp_id);
 
 		if (not compound)
-			throw std::runtime_error("Missing compound information for " + comp_id);
+		{
+			missingCompounds.insert(comp_id);
+			std::cerr << "Missing compound information for " << comp_id << "\n";
+			continue;
+		}
 
 		auto chem_comp_entry = chem_comp.find_first("id"_key == comp_id);
 
+		std::optional<bool> non_std;
+		if  (cf.is_monomer(comp_id))
+			non_std = cf.is_std_monomer(comp_id);
+
 		if (not chem_comp_entry)
 		{
 			chem_comp.emplace({ //
 				{ "id", comp_id },
 				{ "type", compound->type() },
-				{ "mon_nstd_flag", cf.is_std_monomer(comp_id) ? "y" : "n" },
+				{ "mon_nstd_flag", non_std },
 				{ "name", compound->name() },
 				{ "formula", compound->formula() },
 				{ "formula_weight", compound->formula_weight() } });
@@ -402,8 +485,8 @@ void checkAtomRecords(datablock &db)
 
 			if (not chem_comp_entry["type"])
 				items.emplace_back(item{ "type", compound->type() });
-			if (not chem_comp_entry["mon_nstd_flag"])
-				items.emplace_back(item{ "mon_nstd_flag", cf.is_std_monomer(comp_id) ? "y" : "n" });
+			if (not chem_comp_entry["mon_nstd_flag"] and non_std.has_value())
+				items.emplace_back(item{ "mon_nstd_flag", non_std });
 			if (not chem_comp_entry["name"])
 				items.emplace_back(item{ "name", compound->name() });
 			if (not chem_comp_entry["formula"])
@@ -412,23 +495,21 @@ void checkAtomRecords(datablock &db)
 				items.emplace_back(item{ "formula_weight", compound->formula_weight() });
 
 			if (not items.empty())
-				chem_comp_entry.assign(std::move(items));
+				chem_comp_entry.assign(items);
 		}
 
-		if (is_peptide and not has_seq_id(k))
+		if (is_polymer and not has_seq_id(k))
 			throw std::runtime_error("atom_site record has peptide comp_id but no sequence number, cannot continue");
 
 		int seq_id = get_seq_id(k);
 
-		if (row["label_seq_id"].empty())
+		if (is_polymer and row["label_seq_id"].empty() and cf.is_monomer(comp_id))
 			row["label_seq_id"] = std::to_string(seq_id);
 
 		if (row["label_atom_id"].empty())
 			row["label_atom_id"] = row["auth_atom_id"].text();
 		if (row["label_asym_id"].empty())
 			row["label_asym_id"] = row["auth_asym_id"].text();
-		if (row["label_seq_id"].empty())
-			row["label_seq_id"] = row["auth_seq_id"].text();
 		if (row["label_comp_id"].empty())
 			row["label_comp_id"] = row["auth_comp_id"].text();
 		if (row["label_atom_id"].empty())
@@ -448,14 +529,14 @@ void checkAtomRecords(datablock &db)
 
 			float v;
 			auto s = row.get<std::string>(item_name);
-			if (auto [ptr, ec] = cif::from_chars(s.data(), s.data() + s.length(), v); ec != std::errc())
+			if (auto [ptr, ec] = cif::from_chars(s.data(), s.data() + s.length(), v); (bool)ec)
 				continue;
 
 			if (s.length() < prec + 1 or s[s.length() - prec - 1] != '.')
 			{
 				char b[12];
 
-				if (auto [ptr, ec] = cif::to_chars(b, b + sizeof(b), v, cif::chars_format::fixed, prec); ec == std::errc())
+				if (auto [ptr, ec] = cif::to_chars(b, b + sizeof(b), v, cif::chars_format::fixed, prec); (bool)ec)
 					row.assign(item_name, { b, static_cast<std::string::size_type>(ptr - b) }, false, false);
 			}
 		}
@@ -517,21 +598,21 @@ void checkAtomAnisotropRecords(datablock &db)
 		{
 			if (cif::VERBOSE and std::exchange(warnReplaceTypeSymbol, false))
 				std::clog << "Replacing type_symbol in atom_site_anisotrop record(s)\n";
-			row["type_symbol"] != parent["type_symbol"].text();
+			row["type_symbol"] = parent["type_symbol"].text();
 		}
 
-		if (row["pdbx_auth_alt_id"].empty())
+		if (row["pdbx_auth_alt_id"].empty() and not parent["pdbx_auth_alt_id"].empty())
 			row["pdbx_auth_alt_id"] = parent["pdbx_auth_alt_id"].text();
-		if (row["pdbx_label_seq_id"].empty())
+		if (row["pdbx_label_seq_id"].empty() and not parent["pdbx_label_seq_id"].empty())
 			row["pdbx_label_seq_id"] = parent["label_seq_id"].text();
-		if (row["pdbx_label_asym_id"].empty())
+		if (row["pdbx_label_asym_id"].empty() and not parent["pdbx_label_asym_id"].empty())
 			row["pdbx_label_asym_id"] = parent["label_asym_id"].text();
-		if (row["pdbx_label_atom_id"].empty())
+		if (row["pdbx_label_atom_id"].empty() and not parent["pdbx_label_atom_id"].empty())
 			row["pdbx_label_atom_id"] = parent["label_atom_id"].text();
-		if (row["pdbx_label_comp_id"].empty())
+		if (row["pdbx_label_comp_id"].empty() and not parent["pdbx_label_comp_id"].empty())
 			row["pdbx_label_comp_id"] = parent["label_comp_id"].text();
-		if (row["pdbx_PDB_model_num"].empty())
-			row["pdbx_PDB_model_num"] = parent["pdbx_PDB_model_num"].text();
+		// if (row["pdbx_PDB_model_num"].empty() and not parent["pdbx_PDB_model_num"].empty())
+		// 	row["pdbx_PDB_model_num"] = parent["pdbx_PDB_model_num"].text();
 	}
 
 	if (not to_be_deleted.empty())
@@ -705,7 +786,11 @@ void createEntityPoly(datablock &db)
 				if (cf.is_base(comp_id))
 				{
 					c_type = "polydeoxyribonucleotide";
-					letter = letter_can = compound_factory::kBaseMap.at(comp_id);
+					letter_can = compound_factory::kBaseMap.at(comp_id);
+					if (comp_id.length() == 1)
+						letter = letter_can;
+					else
+						letter = '(' + letter_can + ')';
 				}
 				else if (cf.is_peptide(comp_id))
 				{
@@ -737,6 +822,18 @@ void createEntityPoly(datablock &db)
 
 					non_std_monomer = true;
 				}
+				else
+				{
+					// c_type = "other";
+
+					letter_can = c->one_letter_code();
+					if (letter_can == 0)
+						letter_can = 'X';
+
+					letter = '(' + comp_id + ')';
+
+					non_std_monomer = true;
+				}
 
 				if (type.empty())
 					type = c_type;
@@ -803,7 +900,7 @@ void createEntityPoly(datablock &db)
 
 void createEntityPolySeq(datablock &db)
 {
-	if (db.get("entity_poly") == nullptr)
+	if (auto cat = db.get("entity_poly"); cat == nullptr or cat->empty())
 		createEntityPoly(db);
 
 	using namespace literals;
@@ -854,7 +951,10 @@ void createEntityPolySeq(datablock &db)
 
 void createPdbxPolySeqScheme(datablock &db)
 {
-	if (db.get("entity_poly_seq") == nullptr)
+	if (auto cat = db.get("entity_poly"); cat == nullptr or cat->empty())
+		createEntityPoly(db);
+
+	if (auto cat = db.get("entity_poly_seq"); cat == nullptr or cat->empty())
 		createEntityPolySeq(db);
 
 	using namespace literals;
@@ -991,7 +1091,7 @@ bool reconstruct_pdbx(file &file, std::string_view dictionary)
 	// ... and any additional datablock will contain compound information
 	cif::compound_source cs(file);
 
-	if (db.get("atom_site") == nullptr)
+	if (auto cat = db.get("atom_site"); cat == nullptr or cat->empty())
 		throw std::runtime_error("Cannot reconstruct PDBx file, atom data missing");
 
 	auto &validator = validator_factory::instance()[dictionary];
@@ -999,7 +1099,7 @@ bool reconstruct_pdbx(file &file, std::string_view dictionary)
 	std::string entry_id;
 
 	// Phenix files do not have an entry record
-	if (db.get("entry") == nullptr)
+	if (auto cat = db.get("entry"); cat == nullptr or cat->empty())
 	{
 		entry_id = db.name();
 		category entry("entry");
@@ -1064,7 +1164,7 @@ bool reconstruct_pdbx(file &file, std::string_view dictionary)
 				              iv->m_type != nullptr and
 				              iv->m_type->m_primitive_type == cif::DDL_PrimitiveType::Numb;
 
-				for (size_t ix = 0; auto row : cat)
+				for (std::size_t ix = 0; auto row : cat)
 				{
 					if (number)
 						row.assign(key, std::to_string(++ix), false, false);
@@ -1253,16 +1353,19 @@ bool reconstruct_pdbx(file &file, std::string_view dictionary)
 
 	// Now create any missing categories
 	// Next make sure we have struct_asym records
-	if (db.get("struct_asym") == nullptr)
+	if (auto cat = db.get("struct_asym"); cat == nullptr or cat->empty())
 		createStructAsym(db);
 
-	if (db.get("entity") == nullptr)
+	if (auto cat = db.get("entity"); cat == nullptr or cat->empty())
 		createEntity(db);
 
-	if (db.get("pdbx_poly_seq_scheme") == nullptr)
+	// fill in missing formula_weight, e.g.
+	checkEntities(db);
+
+	if (auto cat = db.get("pdbx_poly_seq_scheme"); cat == nullptr or cat->empty())
 		createPdbxPolySeqScheme(db);
 
-	if (db.get("ndb_poly_seq_scheme") != nullptr)
+	if (auto cat = db.get("ndb_poly_seq_scheme"); cat == nullptr or cat->empty())
 		comparePolySeqSchemes(db);
 
 	// skip unknown categories for now

src/pdb/validate-pdbx.cpp

@@ -48,7 +48,7 @@ condition get_parents_condition(const validator &validator, row_handle rh, const
 		{
 			condition cond;
 
-			for (size_t ix = 0; ix < link->m_child_keys.size(); ++ix)
+			for (std::size_t ix = 0; ix < link->m_child_keys.size(); ++ix)
 			{
 				auto childValue = rh[link->m_child_keys[ix]];
 
@@ -71,7 +71,7 @@ bool is_valid_pdbx_file(const file &file, std::string_view dictionary)
 {
 	std::error_code ec;
 	bool result = is_valid_pdbx_file(file, dictionary, ec);
-	return result and ec == std::errc();
+	return result and not (bool)ec;
 }
 
 bool is_valid_pdbx_file(const file &file, std::error_code &ec)
@@ -189,6 +189,7 @@ bool is_valid_pdbx_file(const file &file, std::string_view dictionary, std::erro
 				for (auto asym_id : struct_asym.find<std::string>("entity_id"_key == entity_id, "id"))
 				{
 					if (pdbx_poly_seq_scheme.count(
+							"entity_id"_key == entity_id and
 							"asym_id"_key == asym_id and
 							"mon_id"_key == mon_id and
 							"seq_id"_key == num and
@@ -202,6 +203,7 @@ bool is_valid_pdbx_file(const file &file, std::string_view dictionary, std::erro
 			for (const auto &[seq_id, mon_id, hetero] : pdbx_poly_seq_scheme.find<int, std::string, bool>("entity_id"_key == entity_id, "seq_id", "mon_id", "hetero"))
 			{
 				if (entity_poly_seq.count(
+						"entity_id"_key == entity_id and
 						"mon_id"_key == mon_id and
 						"num"_key == seq_id and
 						"hetero"_key == hetero) != 1)
@@ -326,7 +328,7 @@ bool is_valid_pdbx_file(const file &file, std::string_view dictionary, std::erro
 		ec = make_error_code(validation_error::not_valid_pdbx);
 	}
 
-	if (not result and ec == std::errc())
+	if (not result and (bool)ec)
 		ec = make_error_code(validation_error::not_valid_pdbx);
 
 	return result;

src/symmetry.cpp

@@ -111,7 +111,7 @@ sym_op::sym_op(std::string_view s)
 	m_tb = r.ptr[2] - '0';
 	m_tc = r.ptr[3] - '0';
 
-	if (r.ec != std::errc() or rnri > 192 or r.ptr[0] != '_' or m_ta > 9 or m_tb > 9 or m_tc > 9)
+	if ((bool)r.ec or rnri > 192 or r.ptr[0] != '_' or m_ta > 9 or m_tb > 9 or m_tc > 9)
 		throw std::invalid_argument("Could not convert string into sym_op");
 }
 
@@ -119,7 +119,7 @@ std::string sym_op::string() const
 {
 	char b[9];
 	auto r = std::to_chars(b, b + sizeof(b), m_nr);
-	if (r.ec != std::errc() or r.ptr > b + 4)
+	if ((bool)r.ec or r.ptr > b + 4)
 		throw std::runtime_error("Could not write out symmetry operation to string");
 	
 	*r.ptr++ = '_';
@@ -128,7 +128,7 @@ std::string sym_op::string() const
 	*r.ptr++ = '0' + m_tc;
 	*r.ptr = 0;
 
-	return { b, static_cast<size_t>(r.ptr - b) };
+	return { b, static_cast<std::size_t>(r.ptr - b) };
 }
 
 // --------------------------------------------------------------------
@@ -184,7 +184,7 @@ void transformation::try_create_quaternion()
 
 	auto ev = es.eigenvalues();
 
-	for (size_t j = 0; j < 4; ++j)
+	for (std::size_t j = 0; j < 4; ++j)
 	{
 		if (std::abs(ev[j].real() - 1) > 0.01)
 			continue;
@@ -221,7 +221,7 @@ transformation inverse(const transformation &t)
 spacegroup::spacegroup(int nr)
 	: m_nr(nr)
 {
-	const size_t N = kSymopNrTableSize;
+	const std::size_t N = kSymopNrTableSize;
 	int32_t L = 0, R = static_cast<int32_t>(N - 1);
 	while (L <= R)
 	{
@@ -234,7 +234,7 @@ spacegroup::spacegroup(int nr)
 
 	m_index = L;
 
-	for (size_t i = L; i < N and kSymopNrTable[i].spacegroup() == m_nr; ++i)
+	for (std::size_t i = L; i < N and kSymopNrTable[i].spacegroup() == m_nr; ++i)
 		emplace_back(kSymopNrTable[i].symop().data());
 }
 
@@ -343,7 +343,7 @@ int get_space_group_number(std::string_view spacegroup)
 
 	int result = 0;
 
-	const size_t N = kNrOfSpaceGroups;
+	const std::size_t N = kNrOfSpaceGroups;
 	int32_t L = 0, R = static_cast<int32_t>(N - 1);
 	while (L <= R)
 	{
@@ -365,7 +365,7 @@ int get_space_group_number(std::string_view spacegroup)
 	// not found, see if we can find a match based on xHM name
 	if (result == 0)
 	{
-		for (size_t i = 0; i < kNrOfSpaceGroups; ++i)
+		for (std::size_t i = 0; i < kNrOfSpaceGroups; ++i)
 		{
 			auto &sp = kSpaceGroups[i];
 			if (sp.xHM == spacegroup)
@@ -395,7 +395,7 @@ int get_space_group_number(std::string_view spacegroup, space_group_name type)
 
 	if (type == space_group_name::full)
 	{
-		const size_t N = kNrOfSpaceGroups;
+		const std::size_t N = kNrOfSpaceGroups;
 		int32_t L = 0, R = static_cast<int32_t>(N - 1);
 		while (L <= R)
 		{
@@ -475,7 +475,7 @@ std::tuple<float,point,sym_op> crystal::closest_symmetry_copy(point a, point b)
 
 	a = orthogonal(fa, m_cell);
 
-	for (size_t i = 0; i < m_spacegroup.size(); ++i)
+	for (std::size_t i = 0; i < m_spacegroup.size(); ++i)
 	{
 		sym_op s(static_cast<uint8_t>(i + 1));
 		auto &t = m_spacegroup[i];

src/symop-map-generator.cpp

@@ -280,7 +280,7 @@ int main(int argc, char* const argv[])
 			if (std::isdigit(line[0]))	// start of new spacegroup
 			{
 				auto r = std::from_chars(line.data(), line.data() + line.length(), sgnr);
-				if (r.ec != std::errc())
+				if ((bool)r.ec)
 					throw std::runtime_error("Error parsing symop.lib file");
 				rnr = 1;
 				continue;
@@ -426,7 +426,7 @@ const space_group kSpaceGroups[] =
 out << R"(
 };
 
-const size_t kNrOfSpaceGroups = sizeof(kSpaceGroups) / sizeof(space_group);
+const std::size_t kNrOfSpaceGroups = sizeof(kSpaceGroups) / sizeof(space_group);
 
 const symop_datablock kSymopNrTable[] = {
 )";
@@ -450,7 +450,7 @@ const symop_datablock kSymopNrTable[] = {
 
 		out << R"(};
 
-const size_t kSymopNrTableSize = sizeof(kSymopNrTable) / sizeof(symop_datablock);
+const std::size_t kSymopNrTableSize = sizeof(kSymopNrTable) / sizeof(symop_datablock);
 
 } // namespace mmcif
 )";

src/symop_table_data.hpp

@@ -361,7 +361,7 @@ const space_group kSpaceGroups[] =
 
 };
 
-const size_t kNrOfSpaceGroups = sizeof(kSpaceGroups) / sizeof(space_group);
+const std::size_t kNrOfSpaceGroups = sizeof(kSpaceGroups) / sizeof(space_group);
 
 const symop_datablock kSymopNrTable[] = {
     // P 1
@@ -5286,6 +5286,6 @@ const symop_datablock kSymopNrTable[] = {
     { 5005,   4, { -1, 0, 0, 0, 1, 0, 0, 0,-1, 1, 2, 0, 0, 1, 2, } },
 };
 
-const size_t kSymopNrTableSize = sizeof(kSymopNrTable) / sizeof(symop_datablock);
+const std::size_t kSymopNrTableSize = sizeof(kSymopNrTable) / sizeof(symop_datablock);
 
 } // namespace mmcif

src/text.cpp

@@ -35,24 +35,24 @@ namespace cif
 // --------------------------------------------------------------------
 // This really makes a difference, having our own tolower routines
 
-const uint8_t kCharToLowerMap[256] =
-	{
-		0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
-		0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
-		0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28, 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f,
-		0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, 0x3c, 0x3d, 0x3e, 0x3f,
-		0x40, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
-		0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x5b, 0x5c, 0x5d, 0x5e, 0x5f,
-		0x60, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
-		0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x7b, 0x7c, 0x7d, 0x7e, 0x7f,
-		0x80, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89, 0x8a, 0x8b, 0x8c, 0x8d, 0x8e, 0x8f,
-		0x90, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, 0x98, 0x99, 0x9a, 0x9b, 0x9c, 0x9d, 0x9e, 0x9f,
-		0xa0, 0xa1, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0xaa, 0xab, 0xac, 0xad, 0xae, 0xaf,
-		0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, 0xb8, 0xb9, 0xba, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf,
-		0xc0, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, 0xc8, 0xc9, 0xca, 0xcb, 0xcc, 0xcd, 0xce, 0xcf,
-		0xd0, 0xd1, 0xd2, 0xd3, 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0xdd, 0xde, 0xdf,
-		0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef,
-		0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff};
+const uint8_t kCharToLowerMap[256] = {
+	0x00, 0x01, 0x02, 0x03, 0x04, 0x05, 0x06, 0x07, 0x08, 0x09, 0x0a, 0x0b, 0x0c, 0x0d, 0x0e, 0x0f,
+	0x10, 0x11, 0x12, 0x13, 0x14, 0x15, 0x16, 0x17, 0x18, 0x19, 0x1a, 0x1b, 0x1c, 0x1d, 0x1e, 0x1f,
+	0x20, 0x21, 0x22, 0x23, 0x24, 0x25, 0x26, 0x27, 0x28, 0x29, 0x2a, 0x2b, 0x2c, 0x2d, 0x2e, 0x2f,
+	0x30, 0x31, 0x32, 0x33, 0x34, 0x35, 0x36, 0x37, 0x38, 0x39, 0x3a, 0x3b, 0x3c, 0x3d, 0x3e, 0x3f,
+	0x40, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
+	0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x5b, 0x5c, 0x5d, 0x5e, 0x5f,
+	0x60, 0x61, 0x62, 0x63, 0x64, 0x65, 0x66, 0x67, 0x68, 0x69, 0x6a, 0x6b, 0x6c, 0x6d, 0x6e, 0x6f,
+	0x70, 0x71, 0x72, 0x73, 0x74, 0x75, 0x76, 0x77, 0x78, 0x79, 0x7a, 0x7b, 0x7c, 0x7d, 0x7e, 0x7f,
+	0x80, 0x81, 0x82, 0x83, 0x84, 0x85, 0x86, 0x87, 0x88, 0x89, 0x8a, 0x8b, 0x8c, 0x8d, 0x8e, 0x8f,
+	0x90, 0x91, 0x92, 0x93, 0x94, 0x95, 0x96, 0x97, 0x98, 0x99, 0x9a, 0x9b, 0x9c, 0x9d, 0x9e, 0x9f,
+	0xa0, 0xa1, 0xa2, 0xa3, 0xa4, 0xa5, 0xa6, 0xa7, 0xa8, 0xa9, 0xaa, 0xab, 0xac, 0xad, 0xae, 0xaf,
+	0xb0, 0xb1, 0xb2, 0xb3, 0xb4, 0xb5, 0xb6, 0xb7, 0xb8, 0xb9, 0xba, 0xbb, 0xbc, 0xbd, 0xbe, 0xbf,
+	0xc0, 0xc1, 0xc2, 0xc3, 0xc4, 0xc5, 0xc6, 0xc7, 0xc8, 0xc9, 0xca, 0xcb, 0xcc, 0xcd, 0xce, 0xcf,
+	0xd0, 0xd1, 0xd2, 0xd3, 0xd4, 0xd5, 0xd6, 0xd7, 0xd8, 0xd9, 0xda, 0xdb, 0xdc, 0xdd, 0xde, 0xdf,
+	0xe0, 0xe1, 0xe2, 0xe3, 0xe4, 0xe5, 0xe6, 0xe7, 0xe8, 0xe9, 0xea, 0xeb, 0xec, 0xed, 0xee, 0xef,
+	0xf0, 0xf1, 0xf2, 0xf3, 0xf4, 0xf5, 0xf6, 0xf7, 0xf8, 0xf9, 0xfa, 0xfb, 0xfc, 0xfd, 0xfe, 0xff
+};
 
 // --------------------------------------------------------------------
 
@@ -171,7 +171,7 @@ std::string trim_right_copy(std::string_view s)
 		e = pe;
 	}
 
-	return {s.begin(), e};
+	return { s.begin(), e };
 }
 
 std::string trim_left_copy(std::string_view s)
@@ -185,27 +185,46 @@ std::string trim_left_copy(std::string_view s)
 		b = std::next(b);
 	}
 
-	return {b, s.end()};
+	return { b, s.end() };
 }
 
 void trim_left(std::string &s)
 {
-	auto b = s.begin();
-	while (b != s.end())
+	auto in = s.begin(), out = s.begin();
+
+	while (in != s.end() and std::isspace(*in))
+		++in;
+	
+	if (in == s.end())
+		s.clear();
+	else if (in != out)
 	{
-		if (not std::isspace(*b))
-			break;
-
-		b = std::next(b);
+		while (in != s.end())
+			*out++ = *in++;
+		s.erase(out, s.end());
 	}
-
-	s.erase(s.begin(), b);
 }
 
 void trim(std::string &s)
 {
-	trim_right(s);
-	trim_left(s);
+	auto in = s.begin(), out = s.begin(), end = s.end();
+
+	while (end != s.begin() and std::isspace(*(end - 1)))
+		--end;
+
+	while (in != end and std::isspace(*in))
+		++in;
+	
+	if (in == end)
+		s.clear();
+	else if (in != out)
+	{
+		while (in != end)
+			*out++ = *in++;
+		s.erase(out, s.end());
+	}
+	else if (end != s.end())
+		s.erase(end, s.end());
 }
 
 std::string trim_copy(std::string_view s)
@@ -220,14 +239,14 @@ std::tuple<std::string, std::string> split_item_name(std::string_view item_name)
 	if (item_name.empty())
 		throw std::runtime_error("empty item_name");
 	if (item_name[0] != '_')
-		throw std::runtime_error("item_name '" + std::string { item_name } + "' does not start with underscore");
+		throw std::runtime_error("item_name '" + std::string{ item_name } + "' does not start with underscore");
 
 	auto s = item_name.find('.');
 	if (s == std::string::npos)
 		// throw std::runtime_error("item_name does not contain dot (" + std::string{ item_name } + ')');
 		return std::tuple<std::string, std::string>{ "", item_name.substr(1) };
 	else
-		return std::tuple<std::string, std::string>{item_name.substr(1, s - 1), item_name.substr(s + 1)};
+		return std::tuple<std::string, std::string>{ item_name.substr(1, s - 1), item_name.substr(s + 1) };
 }
 
 // --------------------------------------------------------------------
@@ -242,8 +261,7 @@ std::string cif_id_for_number(int number)
 		result += static_cast<char>('A' + r);
 
 		number = (number - r) / 26 - 1;
-	}
-	while (number >= 0);
+	} while (number >= 0);
 
 	std::reverse(result.begin(), result.end());
 
@@ -298,29 +316,29 @@ enum LineBreakClass
 	kLBC_Unknown
 };
 
-const LineBreakClass kASCII_LBTable[128] =
-	{
-		kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
-		kLBC_CombiningMark, kLBC_BreakAfter, kLBC_LineFeed, kLBC_MandatoryBreak, kLBC_MandatoryBreak, kLBC_CarriageReturn, kLBC_CombiningMark, kLBC_CombiningMark,
-		kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
-		kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
-		kLBC_Space, kLBC_Exlamation, kLBC_Quotation, kLBC_Alphabetic, kLBC_PrefixNumeric, kLBC_PostfixNumeric, kLBC_Alphabetic, kLBC_Quotation,
-		kLBC_OpenPunctuation, kLBC_CloseParenthesis, kLBC_Alphabetic, kLBC_PrefixNumeric,
+const LineBreakClass kASCII_LBTable[128] = {
+	kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
+	kLBC_CombiningMark, kLBC_BreakAfter, kLBC_LineFeed, kLBC_MandatoryBreak, kLBC_MandatoryBreak, kLBC_CarriageReturn, kLBC_CombiningMark, kLBC_CombiningMark,
+	kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
+	kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark, kLBC_CombiningMark,
+	kLBC_Space, kLBC_Exlamation, kLBC_Quotation, kLBC_Alphabetic, kLBC_PrefixNumeric, kLBC_PostfixNumeric, kLBC_Alphabetic, kLBC_Quotation,
+	kLBC_OpenPunctuation, kLBC_CloseParenthesis, kLBC_Alphabetic, kLBC_PrefixNumeric,
 
-		// comma treated differently here, it is not a numeric separator in PDB
-		kLBC_SymbolAllowingBreakAfter /*	kLBC_InfixNumericSeparator */,
+	// comma treated differently here, it is not a numeric separator in PDB
+	kLBC_SymbolAllowingBreakAfter /*	kLBC_InfixNumericSeparator */,
 
-		kLBC_Hyphen, kLBC_InfixNumericSeparator, kLBC_SymbolAllowingBreakAfter,
-		kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric,
-		kLBC_Numeric, kLBC_Numeric, kLBC_InfixNumericSeparator, kLBC_InfixNumericSeparator, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Exlamation,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_OpenPunctuation, kLBC_PrefixNumeric, kLBC_CloseParenthesis, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
-		kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_OpenPunctuation, kLBC_BreakAfter, kLBC_ClosePunctuation, kLBC_Alphabetic, kLBC_CombiningMark};
+	kLBC_Hyphen, kLBC_InfixNumericSeparator, kLBC_SymbolAllowingBreakAfter,
+	kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric, kLBC_Numeric,
+	kLBC_Numeric, kLBC_Numeric, kLBC_InfixNumericSeparator, kLBC_InfixNumericSeparator, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Exlamation,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_OpenPunctuation, kLBC_PrefixNumeric, kLBC_CloseParenthesis, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic,
+	kLBC_Alphabetic, kLBC_Alphabetic, kLBC_Alphabetic, kLBC_OpenPunctuation, kLBC_BreakAfter, kLBC_ClosePunctuation, kLBC_Alphabetic, kLBC_CombiningMark
+};
 
 std::string::const_iterator nextLineBreak(std::string::const_iterator text, std::string::const_iterator end)
 {
@@ -338,33 +356,33 @@ std::string::const_iterator nextLineBreak(std::string::const_iterator text, std:
 
 	const breakAction brkTable[27][27] = {
 		//        OP   CL   CP   QU   GL   NS   EX   SY   IS   PR   PO   NU   AL   ID   IN   HY   BA   BB   B2   ZW   CM   WJ   H2   H3   JL   JV   JT
-		/* OP */ {PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, CPB, PBK, PBK, PBK, PBK, PBK, PBK},
-		/* CL */ {DBK, PBK, PBK, IBK, IBK, PBK, PBK, PBK, PBK, IBK, IBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* CP */ {DBK, PBK, PBK, IBK, IBK, PBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* QU */ {PBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK},
-		/* GL */ {IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK},
-		/* NS */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* EX */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* SY */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* IS */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* PR */ {IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, IBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK},
-		/* PO */ {IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* NU */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* AL */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* ID */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* IN */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* HY */ {DBK, PBK, PBK, IBK, DBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* BA */ {DBK, PBK, PBK, IBK, DBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* BB */ {IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK},
-		/* B2 */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, PBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* ZW */ {DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK},
-		/* CM */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK},
-		/* WJ */ {IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK},
-		/* H2 */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, IBK, IBK},
-		/* H3 */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, IBK},
-		/* JL */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, DBK},
-		/* JV */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, IBK, IBK},
-		/* JT */ {DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, IBK},
+		/* OP */ { PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, PBK, CPB, PBK, PBK, PBK, PBK, PBK, PBK },
+		/* CL */ { DBK, PBK, PBK, IBK, IBK, PBK, PBK, PBK, PBK, IBK, IBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* CP */ { DBK, PBK, PBK, IBK, IBK, PBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* QU */ { PBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK },
+		/* GL */ { IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK },
+		/* NS */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* EX */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* SY */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* IS */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* PR */ { IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, IBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK },
+		/* PO */ { IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* NU */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* AL */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* ID */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* IN */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* HY */ { DBK, PBK, PBK, IBK, DBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* BA */ { DBK, PBK, PBK, IBK, DBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* BB */ { IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK },
+		/* B2 */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, IBK, IBK, DBK, PBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* ZW */ { DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK, PBK, DBK, DBK, DBK, DBK, DBK, DBK, DBK },
+		/* CM */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, DBK, IBK, IBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, DBK },
+		/* WJ */ { IBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, IBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, IBK },
+		/* H2 */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, IBK, IBK },
+		/* H3 */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, IBK },
+		/* JL */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, IBK, IBK, IBK, IBK, DBK },
+		/* JV */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, IBK, IBK },
+		/* JT */ { DBK, PBK, PBK, IBK, IBK, IBK, PBK, PBK, PBK, DBK, IBK, DBK, DBK, DBK, IBK, IBK, IBK, DBK, DBK, PBK, CIB, PBK, DBK, DBK, DBK, DBK, IBK },
 	};
 
 	uint8_t ch = static_cast<uint8_t>(*text);
@@ -415,10 +433,10 @@ std::string::const_iterator nextLineBreak(std::string::const_iterator text, std:
 	return text;
 }
 
-std::vector<std::string> wrapLine(const std::string &text, size_t width)
+std::vector<std::string> wrapLine(const std::string &text, std::size_t width)
 {
 	std::vector<std::string> result;
-	std::vector<size_t> offsets = {0};
+	std::vector<std::size_t> offsets = { 0 };
 
 	auto b = text.begin();
 	while (b != text.end())
@@ -430,18 +448,18 @@ std::vector<std::string> wrapLine(const std::string &text, size_t width)
 		b = e;
 	}
 
-	size_t count = offsets.size() - 1;
+	std::size_t count = offsets.size() - 1;
 
-	std::vector<size_t> minima(count + 1, 1000000);
+	std::vector<std::size_t> minima(count + 1, 1000000);
 	minima[0] = 0;
-	std::vector<size_t> breaks(count + 1, 0);
+	std::vector<std::size_t> breaks(count + 1, 0);
 
-	for (size_t i = 0; i < count; ++i)
+	for (std::size_t i = 0; i < count; ++i)
 	{
-		size_t j = i + 1;
+		std::size_t j = i + 1;
 		while (j <= count)
 		{
-			size_t w = offsets[j] - offsets[i];
+			std::size_t w = offsets[j] - offsets[i];
 
 			if (w > width)
 				break;
@@ -449,7 +467,7 @@ std::vector<std::string> wrapLine(const std::string &text, size_t width)
 			while (w > 0 and isspace(text[offsets[i] + w - 1]))
 				--w;
 
-			size_t cost = minima[i];
+			std::size_t cost = minima[i];
 			if (j < count) // last line may be shorter
 				cost += (width - w) * (width - w);
 
@@ -463,10 +481,10 @@ std::vector<std::string> wrapLine(const std::string &text, size_t width)
 		}
 	}
 
-	size_t j = count;
+	std::size_t j = count;
 	while (j > 0)
 	{
-		size_t i = breaks[j];
+		std::size_t i = breaks[j];
 		result.push_back(text.substr(offsets[i], offsets[j] - offsets[i]));
 		j = i;
 	}
@@ -476,7 +494,7 @@ std::vector<std::string> wrapLine(const std::string &text, size_t width)
 	return result;
 }
 
-std::vector<std::string> word_wrap(const std::string &text, size_t width)
+std::vector<std::string> word_wrap(const std::string &text, std::size_t width)
 {
 	std::vector<std::string> result;
 	for (auto p : cif::split<std::string>(text, "\n"))

src/utilities.cpp

@@ -215,7 +215,7 @@ const char* kSpinner[] = {
 	 ".", "o", "O", "0", "@", "*", " "
 };
 
-const size_t kSpinnerCount = sizeof(kSpinner) / sizeof(char*);
+const std::size_t kSpinnerCount = sizeof(kSpinner) / sizeof(char*);
 
 const int kSpinnerTimeInterval = 100;
 
@@ -235,7 +235,7 @@ void progress_bar_impl::print_progress()
 	float progress = static_cast<float>(m_consumed) / m_max_value;
 	
 	if (width < kMinBarWidth)
-		std::cout << (100 * progress) << '%' << std::endl;
+		std::cout << (100 * progress) << "%\n";
 	else
 	{
 		uint32_t bar_width = 7 * width / 10;
@@ -329,7 +329,7 @@ void progress_bar_impl::print_done()
 	if (msg.length() < width)
 		msg += std::string(width - msg.length(), ' ');
 
-	std::cout << '\r' << msg << std::endl;
+	std::cout << '\r' << msg << '\n';
 }
 
 progress_bar::progress_bar(int64_t inMax, const std::string &inAction)
@@ -877,6 +877,21 @@ class resource_pool
 
 resource_pool::resource_pool()
 {
+	// directories are searched in reverse order
+
+	// As a last resort, try the location that might have been
+	// used during installation, works only when running on an
+	// OS with a proc file system.
+
+	std::error_code ec;
+	if (auto exefile = fs::read_symlink("/proc/self/exe", ec); not ec and exefile.parent_path().filename() == "bin")
+	{
+		auto install_prefix = exefile.parent_path().parent_path();
+		auto data_dir = install_prefix / "share" / "libcifpp";
+		if (fs::exists(data_dir, ec))
+			pushDir(data_dir);
+	}
+
 #if defined(DATA_DIR)
 	pushDir(DATA_DIR);
 #endif

src/validate.cpp

@@ -55,14 +55,12 @@ validation_exception::validation_exception(std::error_code ec)
 }
 
 validation_exception::validation_exception(std::error_code ec, std::string_view category)
-	: runtime_error(
-		  (std::ostringstream{} << ec.message() << "; category: " << std::quoted(category)).str())
+	: runtime_error((ec.message() + "; category: ").append(category))
 {
 }
 
 validation_exception::validation_exception(std::error_code ec, std::string_view category, std::string_view item)
-	: runtime_error(
-		  (std::ostringstream{} << ec.message() << "; category: " << std::quoted(category) << "; item: " << std::quoted(item)).str())
+	: runtime_error((ec.message() + "; category: ").append(category).append("; item: ").append(item))
 {
 }
 
@@ -81,7 +79,7 @@ struct regex_impl : public regex
 DDL_PrimitiveType map_to_primitive_type(std::string_view s, std::error_code &ec) noexcept
 {
 	ec = {};
-	DDL_PrimitiveType result;
+	DDL_PrimitiveType result = DDL_PrimitiveType::Char;
 	if (iequals(s, "char"))
 		result = DDL_PrimitiveType::Char;
 	else if (iequals(s, "uchar"))
@@ -140,7 +138,7 @@ int type_validator::compare(std::string_view a, std::string_view b) const
 				ra = selected_charconv<double>::from_chars(a.data(), a.data() + a.length(), da);
 				rb = selected_charconv<double>::from_chars(b.data(), b.data() + b.length(), db);
 
-				if (ra.ec == std::errc() and rb.ec == std::errc())
+				if (not (bool)ra.ec and not (bool)rb.ec)
 				{
 					auto d = da - db;
 					if (std::abs(d) > std::numeric_limits<double>::epsilon())
@@ -151,7 +149,7 @@ int type_validator::compare(std::string_view a, std::string_view b) const
 							result = -1;
 					}
 				}
-				else if (ra.ec == std::errc())
+				else if ((bool)ra.ec)
 					result = 1;
 				else
 					result = -1;
@@ -224,7 +222,7 @@ void item_validator::operator()(std::string_view value) const
 
 bool item_validator::validate_value(std::string_view value, std::error_code &ec) const noexcept
 {
-	ec = {};
+	ec.clear();
 
 	if (not value.empty() and value != "?" and value != ".")
 	{
@@ -234,7 +232,7 @@ bool item_validator::validate_value(std::string_view value, std::error_code &ec)
 			ec = make_error_code(validation_error::value_is_not_in_enumeration_list);
 	}
 
-	return ec == std::errc();
+	return not (bool)ec;
 }
 
 // --------------------------------------------------------------------
@@ -355,7 +353,7 @@ void validator::add_link_validator(link_validator &&v)
 	if (ccv == nullptr)
 		throw std::runtime_error("unknown child category " + v.m_child_category);
 
-	for (size_t i = 0; i < v.m_parent_keys.size(); ++i)
+	for (std::size_t i = 0; i < v.m_parent_keys.size(); ++i)
 	{
 		auto piv = pcv->get_validator_for_item(v.m_parent_keys[i]);
 

test/CMakeLists.txt

@@ -0,0 +1,75 @@
+# We're using the older version 2 of Catch2
+
+if(NOT(Catch2_FOUND OR TARGET Catch2))
+	find_package(Catch2 QUIET)
+
+	if(NOT Catch2_FOUND)
+		include(FetchContent)
+
+		FetchContent_Declare(
+			Catch2
+			GIT_REPOSITORY https://github.com/catchorg/Catch2.git
+			GIT_TAG v2.13.9)
+
+		FetchContent_MakeAvailable(Catch2)
+
+		set(Catch2_VERSION "2.13.9")
+	endif()
+endif()
+
+list(
+	APPEND
+	CIFPP_tests
+	unit-v2
+	unit-3d
+	format
+	model
+	rename-compound
+	sugar
+	spinner
+	# reconstruction
+	validate-pdbx)
+
+add_library(test-main OBJECT "${CMAKE_CURRENT_SOURCE_DIR}/test-main.cpp")
+
+target_link_libraries(test-main cifpp::cifpp Catch2::Catch2)
+
+if(${Catch2_VERSION} VERSION_GREATER_EQUAL 3.0.0)
+	target_compile_definitions(test-main PUBLIC CATCH22=0)
+else()
+	target_compile_definitions(test-main PUBLIC CATCH22=1)
+endif()
+
+foreach(CIFPP_TEST IN LISTS CIFPP_tests)
+	set(CIFPP_TEST "${CIFPP_TEST}-test")
+	set(CIFPP_TEST_SOURCE "${CMAKE_CURRENT_SOURCE_DIR}/${CIFPP_TEST}.cpp")
+
+	add_executable(
+		${CIFPP_TEST} ${CIFPP_TEST_SOURCE} $<TARGET_OBJECTS:test-main>)
+
+	if(${Catch2_VERSION} VERSION_GREATER_EQUAL 3.0.0)
+		target_compile_definitions(${CIFPP_TEST} PUBLIC CATCH22=0)
+	else()
+		target_compile_definitions(${CIFPP_TEST} PUBLIC CATCH22=1)
+	endif()
+
+	target_link_libraries(${CIFPP_TEST} PRIVATE cifpp::cifpp Catch2::Catch2)
+	target_include_directories(${CIFPP_TEST} PRIVATE "${EIGEN_INCLUDE_DIR}")
+
+	if(MSVC)
+		# Specify unwind semantics so that MSVC knowns how to handle exceptions
+		target_compile_options(${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}> --data-dir
+		${CMAKE_CURRENT_SOURCE_DIR})
+
+	add_test(NAME ${CIFPP_TEST} COMMAND $<TARGET_FILE:${CIFPP_TEST}> --data-dir
+		${CMAKE_CURRENT_SOURCE_DIR})
+endforeach()

test/format-test.cpp

@@ -24,7 +24,8 @@
  * SOFTWARE, EVEN IF ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.
  */
 
-#include <catch2/catch.hpp>
+
+#include "test-main.hpp"
 
 #include <stdexcept>
 

test/model-test.cpp

@@ -26,19 +26,17 @@
 
 #include "test-main.hpp"
 
-#include <catch2/catch.hpp>
-
 #include <stdexcept>
 
 #include <cif++.hpp>
 
 // --------------------------------------------------------------------
 
-cif::file operator""_cf(const char *text, size_t length)
+cif::file operator""_cf(const char *text, 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);
 		}
@@ -165,10 +163,10 @@ _atom_type.symbol   C
 
 	if (not(expected.front() == structure.get_datablock()))
 	{
-		REQUIRE(false);
-		std::cout << expected.front() << '\n'
+		std::cerr << expected.front() << '\n'
 				  << '\n'
 				  << structure.get_datablock() << '\n';
+		REQUIRE(false);
 	}
 }
 

test/reconstruction-test.cpp

@@ -28,8 +28,6 @@
 
 #include <cif++.hpp>
 
-#include <catch2/catch.hpp>
-
 #include <iostream>
 #include <fstream>
 
@@ -56,7 +54,7 @@ TEST_CASE("reconstruct")
 
 			std::error_code ec;
 			CHECK_FALSE(cif::pdb::is_valid_pdbx_file(f, ec));
-			CHECK(ec != std::errc{});
+			CHECK((bool)ec);
 
 			CHECK(cif::pdb::reconstruct_pdbx(f));
 		}

test/rename-compound-test.cpp

@@ -28,8 +28,6 @@
 
 #include <cif++.hpp>
 
-#include <catch2/catch.hpp>
-
 #include <iostream>
 #include <fstream>
 

test/spinner-test.cpp

@@ -1,6 +1,32 @@
-#include "cif++/utilities.hpp"
+/*-
+ * SPDX-License-Identifier: BSD-2-Clause
+ * 
+ * Copyright (c) 2024 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.
+ */
 
-#include <catch2/catch.hpp>
+#include "test-main.hpp"
+
+#include "cif++/utilities.hpp"
 
 #include <random>
 #include <thread>

test/sugar-test.cpp

@@ -26,19 +26,17 @@
 
 #include "test-main.hpp"
 
-#include <catch2/catch.hpp>
-
 #include <stdexcept>
 
 #include <cif++.hpp>
 
 // --------------------------------------------------------------------
 
-cif::file operator""_cf(const char* text, size_t length)
+cif::file operator""_cf(const char* text, 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);
         }
@@ -133,7 +131,7 @@ TEST_CASE("sugar_name_1")
 // 	auto &db = s.get_datablock();
 // 	auto &as = db["atom_site"];
 
-// 	for (size_t i = 0; i < 2; ++i)
+// 	for (std::size_t i = 0; i < 2; ++i)
 // 	{
 // 		for (auto r : as.find("label_asym_id"_key == "H" and "auth_seq_id"_key == i + 1))
 // 			/*auto &ri = */ai[i].emplace_back(r);

test/test-main.cpp

@@ -1,10 +1,9 @@
+#define CATCH_CONFIG_RUNNER 1
+
 #include "test-main.hpp"
 
 #include <cif++.hpp>
 
-#define CATCH_CONFIG_RUNNER
-#include <catch2/catch.hpp>
-
 std::filesystem::path gTestDir = std::filesystem::current_path();
 
 int main(int argc, char *argv[])
@@ -12,7 +11,13 @@ int main(int argc, char *argv[])
 	Catch::Session session; // There must be exactly one instance
 
 	// Build a new parser on top of Catch2's
+#if CATCH22
 	using namespace Catch::clara;
+#else
+	// Build a new parser on top of Catch2's
+	using namespace Catch::Clara;
+#endif
+
 	auto cli = session.cli()                               // Get Catch2's command line parser
 	           | Opt(gTestDir, "data-dir")                 // bind variable to a new option, with a hint string
 	                 ["-D"]["--data-dir"]                  // the option names it will respond to

test/test-main.hpp

@@ -1,3 +1,37 @@
+/*-
+ * SPDX-License-Identifier: BSD-2-Clause
+ * 
+ * Copyright (c) 2024 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
+
+#if CATCH22
+#include <catch2/catch.hpp>
+#else
+#include <catch2/catch_all.hpp>
+#endif
+
 #include <filesystem>
 
 extern std::filesystem::path gTestDir;

test/unit-3d-test.cpp

@@ -26,8 +26,6 @@
 
 #include "test-main.hpp"
 
-#include <catch2/catch.hpp>
-
 #include <stdexcept>
 
 #include <cif++.hpp>
@@ -36,11 +34,11 @@
 
 // --------------------------------------------------------------------
 
-cif::file operator""_cf(const char *text, size_t length)
+cif::file operator""_cf(const char *text, 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);
 		}
@@ -159,29 +157,29 @@ TEST_CASE("dh_q_0")
 	};
 
 	auto a = cif::dihedral_angle(t[0], t[1], t[2], p);
-	REQUIRE_THAT(a, Catch::Matchers::WithinRel(0, 0.01f));
+	REQUIRE_THAT(a, Catch::Matchers::WithinRel(0.f, 0.01f));
 
 	auto q = cif::construct_from_angle_axis(90, axis);
 
 	p.rotate(q);
 
-	REQUIRE_THAT(p.m_x, Catch::Matchers::WithinRel(1, 0.01f));
-	REQUIRE_THAT(p.m_y, Catch::Matchers::WithinRel(0, 0.01f));
-	REQUIRE_THAT(p.m_z, Catch::Matchers::WithinRel(1, 0.01f));
+	REQUIRE(std::abs(p.m_x - 1.f) < 0.01f);
+	REQUIRE(std::abs(p.m_y - 0.f) < 0.01f);
+	REQUIRE(std::abs(p.m_z - 1.f) < 0.01f);
 
 	a = cif::dihedral_angle(t[0], t[1], t[2], p);
-	REQUIRE_THAT(a, Catch::Matchers::WithinRel(90, 0.01f));
+	REQUIRE(std::abs(a - 90.f) < 0.01f);
 
 	q = cif::construct_from_angle_axis(-90, axis);
 
 	p.rotate(q);
 
-	REQUIRE_THAT(p.m_x, Catch::Matchers::WithinRel(1, 0.01f));
-	REQUIRE_THAT(p.m_y, Catch::Matchers::WithinRel(1, 0.01f));
-	REQUIRE_THAT(p.m_z, Catch::Matchers::WithinRel(0, 0.01f));
+	REQUIRE(std::abs(p.m_x - 1.f) < 0.01f);
+	REQUIRE(std::abs(p.m_y - 1.f) < 0.01f);
+	REQUIRE(std::abs(p.m_z - 0.f) < 0.01f);
 
 	a = cif::dihedral_angle(t[0], t[1], t[2], p);
-	REQUIRE_THAT(a, Catch::Matchers::WithinRel(0, 0.01f));
+	REQUIRE(std::abs(a - 0.f) < 0.01f);
 
 }
 
@@ -232,7 +230,7 @@ TEST_CASE("dh_q_1")
 
 TEST_CASE("m2q_0, *utf::tolerance(0.001f)")
 {
-	for (size_t i = 0; i < cif::kSymopNrTableSize; ++i)
+	for (std::size_t i = 0; i < cif::kSymopNrTableSize; ++i)
 	{
 		auto d = cif::kSymopNrTable[i].symop().data();
 
@@ -258,10 +256,10 @@ TEST_CASE("m2q_0, *utf::tolerance(0.001f)")
 
 		auto ev = es.eigenvalues();
 
-		size_t bestJ = 0;
+		std::size_t bestJ = 0;
 		float bestEV = -1;
 
-		for (size_t j = 0; j < 4; ++j)
+		for (std::size_t j = 0; j < 4; ++j)
 		{
 			if (bestEV < ev[j].real())
 			{
@@ -295,7 +293,7 @@ TEST_CASE("m2q_0, *utf::tolerance(0.001f)")
 
 // "TEST_CASE(m2q_1, *utf::tolerance(0.001f)")
 // {
-// 	for (size_t i = 0; i < cif::kSymopNrTableSize; ++i)
+// 	for (std::size_t i = 0; i < cif::kSymopNrTableSize; ++i)
 // 	{
 // 		auto d = cif::kSymopNrTable[i].symop().data();
 
@@ -319,10 +317,10 @@ TEST_CASE("m2q_0, *utf::tolerance(0.001f)")
 
 // 		auto &&[ev, em] = cif::eigen(m * (1/3.0f), false);
 
-// 		size_t bestJ = 0;
+// 		std::size_t bestJ = 0;
 // 		float bestEV = -1;
 
-// 		for (size_t j = 0; j < 4; ++j)
+// 		for (std::size_t j = 0; j < 4; ++j)
 // 		{
 // 			if (bestEV < ev[j])
 // 			{

test/unit-v2-test.cpp

@@ -26,8 +26,6 @@
 
 #include "test-main.hpp"
 
-#include <catch2/catch.hpp>
-
 #include <cif++.hpp>
 
 #include "cif++/dictionary_parser.hpp"
@@ -36,11 +34,11 @@
 
 // --------------------------------------------------------------------
 
-cif::file operator""_cf(const char *text, size_t length)
+cif::file operator""_cf(const char *text, 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);
 		}
@@ -52,6 +50,32 @@ cif::file operator""_cf(const char *text, size_t length)
 
 // --------------------------------------------------------------------
 
+TEST_CASE("from_chars_1")
+{
+auto f = R"(data_TEST
+#
+loop_
+_test.v
+616.487
+616.487000
+    )"_cf;
+
+	auto &db = f.front();
+	auto &c = db.front();
+
+	auto r1 = c.front();
+	REQUIRE(r1.get<double>("v") == 616.487);
+	REQUIRE(r1["v"].compare(616.487) == 0);
+
+	auto r2 = c.back();
+	REQUIRE(r2.get<double>("v") == 616.487);
+	REQUIRE(r2["v"].compare(616.487) == 0);
+
+
+}
+
+// --------------------------------------------------------------------
+
 TEST_CASE("id_1")
 {
 	REQUIRE(cif::cif_id_for_number(0) == "A");
@@ -103,7 +127,7 @@ TEST_CASE("cc_1")
 		float tv;
 		const auto &[ptr, ec] = cif::from_chars(txt.data(), txt.data() + txt.length(), tv);
 
-		REQUIRE(ec == std::errc());
+		CHECK_FALSE((bool)ec);
 		REQUIRE(tv == val);
 		if (ch != 0)
 			REQUIRE(*ptr == ch);
@@ -121,7 +145,7 @@ TEST_CASE("cc_2")
 		char buffer[64];
 		const auto &[ptr, ec] = cif::to_chars(buffer, buffer + sizeof(buffer), val, cif::chars_format::fixed, prec);
 
-		REQUIRE(ec == std::errc());
+		CHECK_FALSE((bool)ec);
 
 		REQUIRE(buffer == test);
 	}
@@ -242,7 +266,7 @@ TEST_CASE("r_2")
 {
 	cif::category c("foo");
 
-	for (size_t i = 1; i < 256; ++i)
+	for (std::size_t i = 1; i < 256; ++i)
 	{
 		c.emplace({ { "id", i },
 			{ "txt", std::string(i, 'x') } });
@@ -539,7 +563,7 @@ _test.value
 	REQUIRE(not t.empty());
 	REQUIRE(t.front()["name"].as<std::string>() == "aap");
 
-	auto t2 = test.find(cif::key("value") == 1.2f);
+	auto t2 = test.find(cif::key("value") == 1.2);
 	REQUIRE(not t2.empty());
 	REQUIRE(t2.front()["name"].as<std::string>() == "mies");
 }
@@ -574,6 +598,28 @@ _test.value
 
 // --------------------------------------------------------------------
 
+TEST_CASE("compare-with-float")
+{
+	auto f = R"(data_TEST
+#
+loop_
+_test.id
+_test.value
+1 1.0
+2 2.0
+3 3.0
+4 ?
+5 .
+    )"_cf;
+
+	auto &db = f.front();
+	auto &cat = db["test"];
+
+	CHECK(cat.find_first<int>(cif::key("value") == 1.0, "id") == 1);
+}
+
+// --------------------------------------------------------------------
+
 TEST_CASE("sw_1")
 {
 	using namespace cif::literals;
@@ -702,7 +748,7 @@ save__cat_2.desc
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -737,7 +783,7 @@ _cat_2.desc
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -746,27 +792,51 @@ _cat_2.desc
 	std::istream is_data(&data_buffer);
 	f.load(is_data);
 
-	auto &cat1 = f.front()["cat_1"];
-	auto &cat2 = f.front()["cat_2"];
+	SECTION("one")
+	{
+		auto &cat1 = f.front()["cat_1"];
+		auto &cat2 = f.front()["cat_2"];
 
-	REQUIRE(cat1.size() == 3);
-	REQUIRE(cat2.size() == 3);
+		REQUIRE(cat1.size() == 3);
+		REQUIRE(cat2.size() == 3);
 
-	cat1.erase(cif::key("id") == 1);
+		cat1.erase(cif::key("id") == 1);
 
-	REQUIRE(cat1.size() == 2);
-	REQUIRE(cat2.size() == 1);
+		REQUIRE(cat1.size() == 2);
+		REQUIRE(cat2.size() == 1);
 
-	// REQUIRE_THROWS_AS(cat2.emplace({
-	//     { "id", 4 },
-	//     { "parent_id", 4 },
-	//     { "desc", "moet fout gaan" }
-	// }), std::exception);
+		// REQUIRE_THROWS_AS(cat2.emplace({
+		//     { "id", 4 },
+		//     { "parent_id", 4 },
+		//     { "desc", "moet fout gaan" }
+		// }), std::exception);
 
-	REQUIRE_THROWS_AS(cat2.emplace({ { "id", "vijf" }, // <- invalid value
-						  { "parent_id", 2 },
-						  { "desc", "moet fout gaan" } }),
-		std::exception);
+		REQUIRE_THROWS_AS(cat2.emplace({ { "id", "vijf" }, // <- invalid value
+							{ "parent_id", 2 },
+							{ "desc", "moet fout gaan" } }),
+			std::exception);
+	}
+
+	// SECTION("two")
+	// {
+	// 	auto &cat1 = f.front()["cat_1"];
+	// 	auto &cat2 = f.front()["cat_2"];
+
+	// 	cat1.update_value(cif::all(), "id", [](std::string_view v) -> std::string
+	// 	{
+	// 		int vi;
+	// 		auto [ec, ptr] = std::from_chars(v.data(), v.data() + v.length(), vi);
+	// 		return std::to_string(vi + 1);
+	// 	});
+
+	// 	REQUIRE(cat1.find1<std::string>(cif::key("id") == 2, "name") == "Aap");
+	// 	REQUIRE(cat1.find1<std::string>(cif::key("id") == 3, "name") == "Noot");
+	// 	REQUIRE(cat1.find1<std::string>(cif::key("id") == 4, "name") == "Mies");
+
+	// 	REQUIRE(cat2.find1<int>(cif::key("id") == 1, "parent_id") == 2);
+	// 	REQUIRE(cat2.find1<int>(cif::key("id") == 2, "parent_id") == 2);
+	// 	REQUIRE(cat2.find1<int>(cif::key("id") == 3, "parent_id") == 3);
+	// }
 }
 
 // --------------------------------------------------------------------
@@ -831,7 +901,7 @@ save__cat_1.c
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -858,7 +928,7 @@ mies Mies
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -993,7 +1063,7 @@ save__cat_2.desc
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1031,7 +1101,7 @@ _cat_2.desc
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1196,7 +1266,7 @@ save__cat_2.parent_id3
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1244,7 +1314,7 @@ _cat_2.parent_id3
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1417,7 +1487,7 @@ cat_2 3 cat_2:cat_1:3
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1458,7 +1528,7 @@ _cat_2.parent_id3
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1657,7 +1727,7 @@ cat_2 1 cat_2:cat_1:1
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -1698,7 +1768,7 @@ _cat_2.parent_id_2
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2053,7 +2123,7 @@ cat_2 1 '_cat_2.num'  '_cat_3.num'  cat_3
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2099,7 +2169,7 @@ _cat_3.num
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2338,7 +2408,7 @@ cat_2 1 '_cat_2.num'  '_cat_3.num'  cat_3
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2384,7 +2454,7 @@ _cat_3.num
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2545,11 +2615,11 @@ _cat_3.num
 // 				bonded.insert({atom_id_1, atom_id_2});
 // 		}
 
-// 		for (size_t i = 0; i + 1 < atoms.size(); ++i)
+// 		for (std::size_t i = 0; i + 1 < atoms.size(); ++i)
 // 		{
 // 			auto label_i = atoms[i].labelAtomID();
 
-// 			for (size_t j = i + 1; j < atoms.size(); ++j)
+// 			for (std::size_t j = i + 1; j < atoms.size(); ++j)
 // 			{
 // 				auto label_j = atoms[j].labelAtomID();
 
@@ -2570,7 +2640,7 @@ _cat_3.num
 
 // 	auto &poly = structure.polymers().front();
 
-// 	for (size_t i = 0; i + 1 < poly.size(); ++i)
+// 	for (std::size_t i = 0; i + 1 < poly.size(); ++i)
 // 	{
 // 		auto C = poly[i].atomByID("C");
 // 		auto N = poly[i + 1].atomByID("N");
@@ -2672,7 +2742,7 @@ boo.data_.whatever
 
 	REQUIRE(test1.size() == sizeof(kS) / sizeof(T));
 
-	size_t i = 0;
+	std::size_t i = 0;
 	for (auto r : test1)
 	{
 		auto text = r.get<std::string>("text");
@@ -2728,7 +2798,7 @@ There it was!)",
 
 	REQUIRE(test1.size() == sizeof(kS) / sizeof(T));
 
-	size_t i = 0;
+	std::size_t i = 0;
 	for (auto r : test1)
 	{
 		auto text = r.get<std::string>("text");
@@ -2935,7 +3005,7 @@ save__cat_1.name
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -2962,7 +3032,7 @@ _cat_1.name
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3130,7 +3200,7 @@ save__cat_1.name
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3161,7 +3231,7 @@ _cat_1.name
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3237,7 +3307,7 @@ save__cat_1.id_2
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3266,7 +3336,7 @@ _cat_1.id_2
 
 	struct data_membuf : public std::streambuf
 	{
-		data_membuf(char *text, size_t length)
+		data_membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3450,7 +3520,7 @@ ATOM      7  CD  PRO A   1      15.762  13.216  43.724  1.00 30.71           C)"
 
 	struct membuf : public std::streambuf
 	{
-		membuf(char *text, size_t length)
+		membuf(char *text, std::size_t length)
 		{
 			this->setg(text, text, text + length);
 		}
@@ -3486,4 +3556,11 @@ TEST_CASE("pdb2cif_formula_weight")
 
 	fw = a.front()["entity"].find1<float>(cif::key("id") == 3, "formula_weight");
 	CHECK(fw == 18.015f);
+}
+
+// --------------------------------------------------------------------
+
+TEST_CASE("update_values_with_provider")
+{
+
 }

test/validate-pdbx-test.cpp

@@ -26,19 +26,17 @@
 
 #include "test-main.hpp"
 
-#include <catch2/catch.hpp>
-
 #include <cif++.hpp>
 
 #include <stdexcept>
 
 // --------------------------------------------------------------------
 
-cif::file operator""_cf(const char *text, size_t length)
+cif::file operator""_cf(const char *text, 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);
 		}