On 06/10/2010 17:11, Geoffrey Hutchison wrote: >> In AliasData::FromNameLookup allow multiply bonded placeholder atoms, which >> disappeared when FormulaParse was deleted. Should fix fail in inchitest. > > This allows the inchi test to run, but we're getting a weird failure for > molecule #1 (ferrocene, I think): > > # > InChI=1S/2C5H5.10CH.2Fe/c2*1-2-4-5-3-1;;;;;;;;;;;;/h2*1-5H;10*1H;;/q;;;;;;;;;;2*-1;;+2 > # InChI=1S/4C5H5.2Fe/c4*1-2-4-5-3-1;;/h4*1-5H;;/q;;2*-1;;+2 > This was how it was before r4137 (whose comments are quoted above) but the "Continuous" builds show the bug is now fixed: Manganese.timvdm 2010-10-05T14:46:24 EDT failed on inchiSamples_Test 2010-10-06T11:38:33 EDT no failures
> BTW, I think we should probably augment our InChI tests using the InChI > executable and run it on one of our OB test files. You pointed out that the > InChI test set uses non-standard SDF options, but we can clearly augment with > our stereo test files, etc. I agree. The InChI sample files are not mainstream examples: some are edge cases and some use inappropriate alias format, the interpretation of which is what is really being tested here. Noel's recent test on his blog is the sort of test that OB should use (scaled down). I'm expect that he is expecting to do this... Chris ------------------------------------------------------------------------------ Beautiful is writing same markup. Internet Explorer 9 supports standards for HTML5, CSS3, SVG 1.1, ECMAScript5, and DOM L2 & L3. Spend less time writing and rewriting code and more time creating great experiences on the web. Be a part of the beta today. http://p.sf.net/sfu/beautyoftheweb _______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel
