[Rdkit-devel] Beta of Q1 2012 release available

[Greg L. <gre...@gm...>]

Dear all,

This morning I tagged the beta for the 2012.03 (Q1 2012 in the old
numbering) release in svn:
http://rdkit.svn.sourceforge.net/viewvc/rdkit/tags/Release_2012_03_1beta1/

and uploaded a source distribution to the google code site:
http://code.google.com/p/rdkit/downloads/detail?name=RDKit_2012_03_1beta1.tgz
If there's demand for it, I will also put up a windows binary.

As usual: if no show-stopper bugs appear, I will do the release itself
in about a week.

Excerpts from the release notes are below.

Best Regards,
-greg

******  Release_2012.03.1 *******
(Changes relative to Release_2011.12.1)

!!!!!! IMPORTANT !!!!!!
 - The atom-atom match behavior for non-query atoms has been changed.
    This affects the results of doing substructure matches using
    query molecules that are not constructed from SMARTS.

Acknowledgements:
JP Ebejer, Paul Emsley, Roger Sayle, Adrian Schreyer, Gianluca Sforna,
Riccardo Vianello

Bug Fixes:
- the older form of group evaluations in Mol blocks is now correctly
  parsed. (Issue 3477283)
- some problems with handling aromatic boron were fixed. (Issue 3480481)
- the SD writer no longer adds an extra $$$$ when molecule parsing
  fails (Issue 3480790)
- molecules in SD files that don't contain atoms are now parsed
  without warnings and their properties are read in. (Issue 3482695)
- it's now possible to embed molecules despite failures in the triangle
  bounds checking (Issue 3483968)
- Isotope information in Mol blocks is now written to M ISO lines
  instead of going in the atom block. (Issue 3494552)
- Better 2D coordinates are now generated for neighbors of atoms with
  unspecified hybridization. (Issue 3487469)
- Dummy atoms and query atoms are now assigned UNSPECIFIED hybridization
  instead of SP. (Issue 3487473)
- Error reporting for SMARTS involving recursion has been improved.
  (Issue 3495296)
- Some problems of queries and generating SMARTS for queries were resolved.
  (Issues 3496759, 3496799, 3496800)
- It's now possible to do database queries with SMARTS that use the index.
  (Issue 3493156).
- A series of problems related to thread safety were fixed.
- Tracking the lifetime of owning molecules across the C++/Python
border is now being handled better (Issue 3510149)
- The AllChem module now imports successfully even if the SLN parser
hasn't been built.

New Features:
- The molecular sanitization is now configurable using an optional
  command-line argument.
- It's now possible to get information from the sanitization routine
  about which operation failed.
- Suppliers support GetLastItemText()
- ComputeDihedralAngle() and ComputeSignedDihedralAngle() were added
  to the rdkit.Geometry module.
- computeSignedDihedralAngle() was added to the C++ API
- ChemicalReactions now support a GetReactingAtoms() method
- the Mol file and Mol block parsers, as well as the SD suppliers,
  now support an optional "strictParsing" argument.
  When this is set to False, problems in the structure of the
  input file are ignored when possible
- EditableMols return the index of the atom/bond added by AddAtom/AddBond
- rdkit.Chem.Draw.MolToImage() now supports an optional "legend" argument
- The MolToSmiles function now supports an optional "allBondsExplicit"
argument.

New Database Cartridge Features:
- the functions mol_from_smiles() and mol_from_smarts() were added

New Java Wrapper Features:
- the diversity picker now supports an optional random-number seed

Deprecated modules (to be removed in next release):
- rdkit.Excel

Removed modules:
- rdkit.ML.Descriptors.DescriptorsCOM
- rdkit.ML.Composite.CompositeCOM

Other:
- Assigning/cleaning up stereochemistry is now considerably
faster. This makes standard molecule construction faster.