File Changes for angie

v244_base to v245_preview (2010-12-07 to 2010-12-14) v245

• src/hg/makeDb/doc/hg18.txt
  • lines changed 4, context: html, text, full: html, text
    Update to Track #1656 (GWAS Catalog): reloaded hg18 and hg19gwasCatalog with latest data.
• src/hg/makeDb/doc/hg19.txt
  • lines changed 3, context: html, text, full: html, text
    Update to Track #1656 (GWAS Catalog): reloaded hg18 and hg19gwasCatalog with latest data.
• src/hg/makeDb/trackDb/human/hg19/trackDb.ra
  • lines changed 1, context: html, text, full: html, text
    Removed stray "release alpha" from snp131 -- thanks to Brooke fornoticing that some options were missing!
  • lines changed 11, context: html, text, full: html, text
    hg19's snp131 override does need release alpha for visibility hide.Adding a release beta,public stanza without the vis change, like I should have back in August.
• src/hg/utils/tdbQuery/tdbQuery.c
  • lines changed 4, context: html, text, full: html, text
    Moved filterOnRelease into tdbsForDbPath, calling it just after weload stanzas from .ra file(s) at each level, instead of calling it after tdbsForDbPath. The reason for filtering earlier is that stanzas for incompatible releases are hard not to trip on somewhere in the complex logic of merging records. The specific case that was failing, mentioned in MLQ #1963: snp131 has no release specified in the top-level trackDb.ra, but in human/hg19/trackDb.ra, it has two overrides, one for release alpha and the other for beta,public. The beta,public one was processed before the alpha one. Since the top-level had no release, it didn't clash with beta,public, so the records were merged, leaving a singe beta,public record. When we encountered the alpha record, it was incompatible with the merged top-level+beta,public record so it was not merged and had only override params, not top-level. Since it was missing the type param, it failed the doRecordChecks test. Filtering out the record for other releases allowed no-release and alpha to merge properly.