Understood. I think the current behaviour can be preserved. I just want c1n[nH]cc1 to have the double bonds in the same place as c1nn([H])cc1. Right now the code does this by making all hydrogens inside square brackets explicit, which isn't what SMILES is supposed to do.
- Noel On 12 December 2012 16:55, Geoffrey Hutchison <geo...@pitt.edu> wrote: > > On Dec 12, 2012, at 11:44 AM, Noel O'Boyle wrote: > >> I can't find it on the Daylight website, or in the OpenSMILES spec, >> but Roger has told me (the classic "appeal to authority" argument) > ... >> Certainly Daylight's Depict (http://www.daylight.com/daycgi/depict) >> does not accept c1nncc1, but is happy with c1n[nH]cc1. > > The problem is not the appeal to authority or whether or not Daylight depict > accepts certain things. What we know from experience is that SMILES > collections exist, which include such non-standard patterns. IIRC, there are > a few such things in PubChem and ZINC. I remember them bitterly in the > efforts in 2.2 and 2.3 to make the canonical SMILES round-trip. > > I’m sure you remember them -- I always complain about aromatic compounds with > lots of nitrogens -- this is the reason. > > Now, our stated intent with OB has long been “accept quirks, but warn”. > Certainly the PDB parser does this. Several threads exist on the mailing list > on how the SMILES parser is too silent. > > -Geoff ------------------------------------------------------------------------------ LogMeIn Rescue: Anywhere, Anytime Remote support for IT. Free Trial Remotely access PCs and mobile devices and provide instant support Improve your efficiency, and focus on delivering more value-add services Discover what IT Professionals Know. Rescue delivers http://p.sf.net/sfu/logmein_12329d2d _______________________________________________ OpenBabel-Devel mailing list OpenBabel-Devel@lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/openbabel-devel
