opensource/galaxy
3
0
Fork 0
mirror of https://github.com/galaxyproject/galaxy.git synced 2026-09-24 16:30:27 +08:00
Code Issues Packages Projects Releases Wiki Activity

Handle merging bad BLAST XML files better

Browse Source
This commit is contained in:
Peter Cock
2012-04-18 10:36:23 +01:00
parent e6628b9b66
commit a361e7d9cf
1 changed files with 31 additions and 3 deletions
Download Patch File Download Diff File Expand all files Collapse all files
lib/galaxy/datatypes/xml.py
+31 -3
View File
@@ -104,13 +104,41 @@ class BlastXml( GenericXml ):
for f in split_files:
h = open(f)
body = False
header = []
header = h.readline()
if not header:
out.close()
h.close()
raise ValueError("BLAST XML file %s was empty" % f)
if header.strip() != '<?xml version="1.0"?>':
out.write(header) #for diagnosis
out.close()
h.close()
raise ValueError("%s is not an XML file!" % f)
line = h.readline()
header += line
if line.strip() not in ['<!DOCTYPE BlastOutput PUBLIC "-//NCBI//NCBI BlastOutput/EN" "http://www.ncbi.nlm.nih.gov/dtd/NCBI_BlastOutput.dtd">',
'<!DOCTYPE BlastOutput PUBLIC "-//NCBI//NCBI BlastOutput/EN" "NCBI_BlastOutput.dtd">']:
out.write(header) #for diagnosis
out.close()
h.close()
raise ValueError("%s is not a BLAST XML file!" % f)
while True:
line = h.readline()
header.append(line)
if not line:
out.write(header) #for diagnosis
out.close()
h.close()
raise ValueError("BLAST XML file %s ended prematurely" % f)
header += line
if "<Iteration>" in line:
break
header = "".join(header)
if len(header) > 10000:
#Something has gone wrong, don't load too much into memory!
#Write what we have to the merged file for diagnostics
out.write(header)
out.close()
h.close()
raise ValueError("BLAST XML file %s has too long a header!" % f)
if "<BlastOutput>" not in header:
out.close()
h.close()