Commits for angie

v302_base to v303_preview (2014-06-30 to 2014-07-07) v303

• Fixed SEGV in factorSourceLoad: track->items was dereferenced withouta NULL check. Do the NULL check ASAP so we don't even bother with
  all the filtering and joining setup if there are no items in window.
  
  • src/hg/hgTracks/factorSource.c - lines changed 2, context: html, text, full: html, text
• Ugh, c73b60407 introduced a bug: I assumed that if tdb->parent is non-NULL,tdb->parentName would be too -- nope, tdb->parentName is apparently only for
  children of superTracks, not children of views/composites.
  refs #13477
  
  • src/hg/lib/hgFind.c - lines changed 3, context: html, text, full: html, text
• Small bugfix: when reference allele had a padding base but alt allelewas symbolic, the first char was trimmed from the symbolic allele too, oops.
  
  • src/lib/vcf.c - lines changed 2, context: html, text, full: html, text
• hgGene was missing a call to setUdcCacheDir, so it was using /tmp/udcCachewhere it could have ownership conflicts with files left there by command-line
  programs etc.
  
  • src/hg/hgGene/hgGene.c - lines changed 1, context: html, text, full: html, text
• One more spot to convert from regular allocation to localmem.
  • src/hg/lib/variant.c - lines changed 1, context: html, text, full: html, text
• Fixed two bugs in the way cdsBasesAdded was computed -- that causedsome strange functional assignments for "<DEL>" (large deletion).
  Also fixed a miscall of frameshift instead of stop_gained (from
  insertion) brought to my attention by GoldenHelix's comparison of VEP,
  Annovar and SnpEff: http://blog.goldenhelix.com/?p=2486
  
  • src/hg/lib/gpFx.c - lines changed 20, context: html, text, full: html, text