* Removes ATOM/BOND_SPTR in boost::graph in favor of raw pointers
* Actually delete atoms and bonds...
* RWMol::clear now calls destroy to handle atom/bond deletion
* Changes broken Atom lookup for windows/gcc
* Adds tests for running with valgrind
* Adds test designed for valgrind and molecule deletions
* Removes RNG, actually tests bond deletions
* update swig wrappers
* deal with most recent changes on the main branch
* add a SmilesParserParams object to prepare for this
* add a SmilesParserParams object to prepare for this
* add tests for the SmilesParseParmas
* support name parsing, should it be the default?
* rename CXNSmiles to CXSmiles;
add a spirit parser for CXSmiles coordinate that is at least syntax correct
* abandon boost::spirit for now; crude atom token parser
* support params in smiles parser (not tested, may not build)
* can read coords and atom labels along with mol names; crude, but works
* read coordinate bonds
* remove some compiler warnings with VS2015
* remove a bunch of compiler warnings on windows
* remove more warnings on windows
* remove more warnings on windows
* backup commit: first pass at parsing query features
* radical spec parsing
* handle attachment points using atom mapping
* switch to a named property for atom labels
* fix handling of the "A" atom query
* add functions to construct A and Q queries (needs more work)
* fix a problem created while cleaning up warnings earlier
* add some additional convenience functions for making generic atoms.
Still need M and to recognize these while writing CXSMILES
* add M queries; update some tests
* fix a linux compile problem
* get the cxsmiles stuff working in python; basic testing
* support "M" in CXSMILES
* Add support for dative bonds to molfile reader plus writer and to SMILES reader (SMILES writer already supports it).
Initial V3000 molfile reader and writer dative bond support by Esben Jannik Bjerrum.
* first pass at adding -> as dative bond in smiles/smarts
more testing required
* all tests pass
* update .cmake files for lex
This makes things a little less clear (which is why I left it out originally) but it's more efficient. Now, rather than updating vfragsmi with the sorted fragment smiles, the final result smiles is generated directly.
Handle disconnected components correctly when generating the _smilesAtomOutputOrder property. This requires:
(a) the mapping from the fragment molecules back to the original molecule
(b) the fragment maps to be sorted at the same time as the fragment smiles (if canonical)
o rdkit gains a RDKit::common_properties namespace that contains common string value properties
o Dict.h and below gain getPropIfPresent that attempts to retrieve a property and returns
true/false on success or failure. This is used to optimize access.
o rdkit learns how to pass property keys by reference, not value.
A new namespace has been added to RDKit, common_properties
that contains the std::string values for commonly used
properties. This helps to avoid typos in string values
but also avoids a creation of std::strings from character
values. All accessors (has/get/clear and getPropIfPresent) now pass
the key by reference.
Additionally, getPropIfPresent removes the double lookup
of hasProp/getProp which can be a significant speedup
in the smiles and smarts parsers (10-20%)
