[Cod-bugs] A certain CIF is crashing the new version of Mercury

[Saulius Gražulis]

Hi,

On 2018-04-03 16:07, Timothy D Siegler wrote:
> I just downloaded the new version of Mercury. In the new version, CIF
> #1545320 will not load.I tried using both my old copy and a newly
> downloaded copy of this CIF file, but both crash my program. 
> 
> Other CIF files I have tried from the database load perfectly well. 

Since you have just downloaded the new Mercury, it seems like the most
probable reason is in the Mercury itself – you may want to submit a bug
report to CCDC.

I've just looked at the current COD 1545320 entry, it looks fairly OK:
the syntax is correct (checked with cod-tools cifparse, vcif, pycifrw),
validator has only minor complaints (_space_group_symop_id is missing,
usually not a problem for most programs), Jmol, Rasmol and Pymol on
Ubuntu 16.04 read and display the file just fine, and cif_molecule
processes it OK. The calculated cell contents matches the declared one.

The problem might be a large _shelx_res_file data item (you may want to
remove it from the file and try Mercury again), but it is smaller than
many other instances of this data item – it would be strange if it
causes problems...

You can also try the previous revision of this file from
http://www.crystallography.net/cod/1545320.cif@194862 , but the
differences are minor indeed, just bibliography info was added.

Other than that I have really no ideas why Mercury crashes. Unless CCDC
publish Mercury source code under a F/LOSS license and I can run the
code under debugger, we have virtually no means to tell what is going
inside the program.

Regards,
Saulius

-- 
Dr. Saulius Gražulis
Vilnius University Institute of Biotechnology, Saulėtekio al. 7
LT-10257 Vilnius, Lietuva (Lithuania)
fax: (+370-5)-2234367 / phone (office): (+370-5)-2234353
mobile: (+370-684)-49802, (+370-614)-36366