The 'cif_fix_values' script with an option '--fix-only-weighting-scheme' and 'cif_filter' script were run over CIF files which included an incorrect value of the data item '_refine_ls_weighting_scheme'. Read the log for more details. There were more than 18k CIF files with values to modify. Usually these incorrect values include comment with formulas, but values of this data item must be an enumerator. The script creates new data item '_refine_ls_weighting_details' to store these comments with formulas, and writes the enumerator 'calc' as a value of data item '_refine_ls_weighting_scheme'. There exists some exceptions in the script.
The problem was that the reprocessed files from IUCr, although were more complete, "resurrected" data errors that were corrected in the previous recisions. The additional data that have been extracted during the reprocessing will have to be merged without overwriting the existing values in COD files, except some explicit and controlled cases (such as publication titles and chemical names).
cif/2/ Adding reporocessed IUCr CIFs for the years 1980-2010. The CIFs now contain more complete bibliography, substance names, correctly unfolded text fields (using the CIF line unfolding protocol), explicitely set cell angles, and _cod_data_source_block fields.
Redepositing COD files reprocessed in the Acta-Cryst-C-1991-1993-redeposition/ directory.
Although all structures were present in COD already, this redeposition enhances and improves the deposited files in several aspects:
1) full bibliographies, with diacrytic marks, are added from the IUCr BibTeX files;
2) chemical formula information is checked and fixed were necessary;
3) symmetry operators were checked for syntactic correctness, fixed were necessary, and spacegroup information was derived from these operators whenever possible; the original spacegroup was checked against the determined spacegroup and the differences were marked if present;
4) missing cell constants were restored,
5) at least in one entry, corrent coordinates were restored,
6) in entries where problems were detected, _code_depositor_comments and _cod_error_... tags were added,
7) information necessary to generate Acta Cryst. paper back-links was added, and the comment header agreed with IUCr was inserted.
Changing volume number in numerous Acta Cryst. C published CIFs from 'C[0-9]?[0-9]' into '[0-9]?[0-9]'. This makes the COD CIF collection more uniform, and also helps the deposition scripts to spot potential duplicates more accurately.
The command for this transformation was:
for i in 2/*.cif; \ do ( \ if grep -q '_journal_volume *C' $i && \ grep -q '_journal_name_full .*Acta *Cryst' $i; \ then \ echo === $i ===; \ perl -i -pe 's/_journal_volume *C/_journal_volume /' $i; \ perl -i -pe \ "s/_journal_name_full .*/_journal_name_full 'Acta Crystallographica Section C'/" $i ; \ fi \ ) done