Re: [Open Babel] Multiple-Molecule File Formats

[Peter Murray-R. <pm...@ca...>]

At 09:40 24/04/2003 -0500, Geoff Hutchison wrote:


>>I would suggest the latter since that would apply to any OB output 
>>format. The major difficulty would be naming the output files - which 
>>relates to names and IDs within the files (as above) or counting 
>>(mol1.mol, mol2.mol etc.)
>
>No. IMHO, the default should be to output a multiple-molecule file if the 
>output format supports it. This would keep the behavior of Babel 1.x. If 
>you wanted to split a file, you needed to add the -split option (which 
>goes with numbers for splits).

Fine by me - does OB support -split?

If you output multiple molecules in CML you will need a toplevel (root) 
element to keep the document well-formed. Traditionally this has been 
"<list>", as in:
<list id="mollist23" title="steroids" dictRef="mydict:steroids">
   <molecule id="m1" title="testosterone">...</molecule>
   ..
   <molecule id="m5" title="oestradiol">...</molecule>
</list>

(the attributes are optional, and at present there isn't a mechanism for 
adding all of them in OB. I strongly recommend generating them at least for 
the molecules, e.g. from sequence number, other IDs and/or title).

P.