Re: CIF-JSON draft 2017-05-15

Dear Bob,

we at the COD tend to make as little assumptions about input CIFs as possible, thus concatenated CCDC CIFs are retained correctly in COD-JSON. In addition to that, COD-JSON is also able to preserve:

1) data block order (I can imagine an animation of molecular motion where each frame is stored in a separate CIF data block),

2) CIF loops having data items from different categories (AFAIK, a discouraged practice, but still used, for example, see _ccdc_geom_bond_type tag in http://www.crystallography.net/cod/4507300.cif),

3) loops with lists as values, all possible combinations.

Of course, our method has some 2x-3x overhead, but JSON here is used as a "carrier" which is supposed to contain almost everything from the original CIFs (hence the "high-fidelity").

Best,
Andrius

CAF_YUvU1RKe9bGWr4HKkVaSzxZ+aNyiO9wxMZ6zz4zbOGKuuXg@mail.gmail.com" type="cite">

I'd like to also hear from people interested in the issue of concatenation. Obviously CCDC doesn't do this properly for their own multi-structure CIF files. Providing an easy concatenation mechanism that does not require going in and changing data_ names is an added benefit of this idea.

Note that that idea just says that the very top level could be a serial array of CIF-JSON data. Again, a royal pain to figure out if there is no key information. But, perhaps Jmol is unique in its ability to decipher data types without fully reading the data. I don't know. Still, important for me....

Bob

_______________________________________________
cif-developers mailing list
cif-developers@iucr.org
http://mailman.iucr.org/cgi-bin/mailman/listinfo/cif-developers

-- 
Andrius Merkys
PhD student at Vilnius University Institute of Biotechnology, Saulėtekio al. 7, room V325
LT-10257 Vilnius, Lithuania
Lecturer at Vilnius University Faculty of Mathematics and Informatics, Naugarduko g. 24
LT-03225 Vilnius, Lithuania