I found the change (rather easily in the end - it's the last commit to parsmart.h). The magic phrase "Potential ABI/API change" prevented it being merged to the release branch.
Revision: a439d0fc1350a1c50e970f2467a0e837ec80a181 Author: Noel O'Boyle <baoille...@gmail.com> Date: 10/11/2011 09:20:42 Message: Potential ABI/API change: Check parsmart.h parsmart.h, parsmart.cpp: Add SMARTS support for all forms of tet stereo involving 4-coordinate atoms. (3-coordinate + lone pair not yet supported) test/CMakeLists.txt, test/testsmartssym.py: Associated test cases. ---- Modified: include/openbabel/parsmart.h Modified: src/parsmart.cpp Modified: test/CMakeLists.txt Added: test/testsmartssym.py - Noel On 7 August 2013 15:15, Noel O'Boyle <baoille...@gmail.com> wrote: > Not everything was merged to the OB 2.3.2 release branch if it fiddled > with the API. I have a feeling that this was the case here, but I'd > have to go back and doublecheck. Without checking, I think that the > underlying cause is that the SMARTS internals are exposed in the API > (by accident or whatever) and so any work to correct things can cause > the API to change. > > - Noel > > On 7 August 2013 14:59, Craig James <cja...@emolecules.com> wrote: >> Does anybody remember disabling chirality features for SMARTS matching >> somewhere around version 2.3.2? I have a vague memory that using '@' for a >> tetrahedral center was just broken and gave wrong answers prior to Tim's >> work, so it was disabled and SMARTS were matched as though they had no >> chiral markings. I tested earlier versions (pre-2.3.2), and they use the >> '@' markings correctly. I also tested the latest dev branch and it works >> correctly. But the official 2.3.2 release seems to ignore the '@' flag. >> >> wrong: OpenBabel 2.3.2: >> >> $ echo "C[C@H](Br)Cl" | babel -s"C[C@H](Br)Cl" -i smi -o can >> C[C@H](Br)Cl >> 1 molecule converted >> >> $ echo "C[C@H](Br)Cl" | babel -s"C[C@@H](Br)Cl" -i smi -o can >> C[C@H](Br)Cl >> 1 molecule converted >> >> correct: OpenBabel 2.3.x-dev (HEAD) >> >> $ echo "C[C@H](Br)Cl" | obabel -s"C[C@H](Br)Cl" -i smi -o can >> C[C@H](Br)Cl >> 1 molecule converted >> [radon:emi] ~ $ echo "C[C@H](Br)Cl" | obabel -s"C[C@@H](Br)Cl" -i smi -o can >> 0 molecules converted >> >> I'm in the middle of deploying a new version of some of our internal >> software, and ran into this while testing. Normally I like to use the >> production version, but in this case I'll probably want to use the fixed >> version. I tried to dig through the source to find something, but there are >> so many changes in the stereo code I couldn't pinpoint anything. >> >> Thanks, >> Craig >> >> ------------------------------------------------------------------------------ >> Get 100% visibility into Java/.NET code with AppDynamics Lite! >> It's a free troubleshooting tool designed for production. >> Get down to code-level detail for bottlenecks, with <2% overhead. >> Download for free and get started troubleshooting in minutes. >> http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk >> _______________________________________________ >> OpenBabel-Devel mailing list >> OpenBabel-Devel@lists.sourceforge.net >> https://lists.sourceforge.net/lists/listinfo/openbabel-devel >> ------------------------------------------------------------------------------ Get 100% visibility into Java/.NET code with AppDynamics Lite! It's a free troubleshooting tool designed for production. Get down to code-level detail for bottlenecks, with <2% overhead. Download for free and get started troubleshooting in minutes. http://pubads.g.doubleclick.net/gampad/clk?id=48897031&iu=/4140/ostg.clktrk _______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel