def main():
jip = JobInfoProcessor()
fip = FileInfoProcessor()
(workDir, nJobs, jobList) = getWorkJobs(sys.argv[1:])
for jobNum in sorted(jobList):
if jip.process(os.path.join(workDir, 'output', 'job_%d' % jobNum))[1] == 0:
for fileInfo in fip.process(os.path.join(workDir, 'output', 'job_%d' % jobNum)):
pathSE = fileInfo[OutputFileInfo.Path].replace('file://', '').replace('dir://', '')
print('%s %s/%s' % (fileInfo[OutputFileInfo.Hash], pathSE, fileInfo[OutputFileInfo.NameDest]))
def main():
if opts.save_jobjson or opts.save_jobgc or opts.get_events:
(workDir, nJobs, jobList) = getWorkJobs(args)
(log, incomplete, splitter, splitInfo) = (None, False, None, {})
(lumiDict, readDict, writeDict) = ({}, {}, {})
try:
splitter = DataSplitter.loadState(os.path.join(workDir, 'datamap.tar'))
except Exception:
pass
jobList = sorted(jobList)
for jobNum in jobList:
del log
log = utils.ActivityLog('Reading job logs - [%d / %d]' % (jobNum, jobList[-1]))
jobInfo = getJobInfo(workDir, jobNum, lambda retCode: retCode == 0)
if not jobInfo:
if not incomplete:
print 'WARNING: Not all jobs have finished - results will be incomplete!'
incomplete = True
continue
if not parameterized:
if splitter:
splitInfo = splitter.getSplitInfo(jobNum)
outputName = splitInfo.get(DataSplitter.Nickname, splitInfo.get(DataSplitter.DatasetID, 0))
else:
outputName = jobInfo['file'].split()[2].replace("_%d_" % jobNum, '_').replace('/', '_').replace('__', '_')
# Read framework report files to get number of events
try:
outputDir = os.path.join(workDir, 'output', 'job_' + str(jobNum))
for fwkXML in getCMSSWInfo(os.path.join(outputDir, 'cmssw.dbs.tar.gz')):
for run in fwkXML.getElementsByTagName('Run'):
for lumi in run.getElementsByTagName('LumiSection'):
run_id = int(run.getAttribute('ID'))
lumi_id = int(lumi.getAttribute('ID'))
lumiDict.setdefault(outputName, {}).setdefault(run_id, set()).add(lumi_id)
for outFile in fwkXML.getElementsByTagName('File'):
pfn = outFile.getElementsByTagName('PFN')[0].childNodes[0].data
if pfn not in writeDict.setdefault(outputName, {}):
writeDict[outputName][pfn] = 0
writeDict[outputName][pfn] += int(outFile.getElementsByTagName('TotalEvents')[0].childNodes[0].data)
for inFile in fwkXML.getElementsByTagName('InputFile'):
if outputName not in readDict:
readDict[outputName] = 0
readDict[outputName] += int(inFile.getElementsByTagName('EventsRead')[0].childNodes[0].data)
except KeyboardInterrupt:
sys.exit(os.EX_OK)
except Exception:
raise
print 'Error while parsing framework output of job %s!' % jobNum
continue
del log
log = utils.ActivityLog('Simplifying lumi sections')
lumis = {}
for sample in lumiDict:
for run in lumiDict[sample]:
for lumi in lumiDict[sample][run]:
lumis.setdefault(sample, []).append(([run, lumi], [run, lumi]))
for sample in lumiDict:
lumis[sample] = mergeLumi(lumis[sample])
del log
for sample, lumis in lumis.items():
print 'Sample:', sample
print '========================================='
print 'Number of events processed: %12d' % readDict[sample]
print ' Number of events written: %12d' % sum(writeDict.get(sample, {}).values())
if writeDict.get(sample, None):
print
head = [(0, ' Output filename'), (1, 'Events')]
utils.printTabular(head, map(lambda pfn: {0: pfn, 1: writeDict[sample][pfn]}, writeDict[sample]))
if opts.save_jobjson:
outputJSON(lumis, open(os.path.join(workDir, 'processed_%s.json' % sample), 'w'))
print 'Saved processed lumi sections in', os.path.join(workDir, 'processed_%s.json' % sample)
if opts.save_jobgc:
print
print 'List of processed lumisections:'
print '-----------------------------------------'
outputGC(lumis)
print
###########################
# Lumi filter manuipulation
###########################
if opts.save_exprgc or opts.save_exprjson or opts.save_exprfull:
if len(args) == 0:
raise Exception('No arguments given!')
try:
lumis = parseLumiFilter(str.join(' ', args))
except Exception:
raise Exception('Could not parse: %s' % str.join(' ', args))
if opts.save_exprgc:
outputGC(lumis)
if opts.save_exprjson:
outputJSON(lumis)
if opts.save_exprfull:
result = {}
for rlrange in lumis:
start, end = rlrange
assert(start[0] == end[0])
llist = result.setdefault(start[0], []).extend(range(start[1], end[1] + 1))
print result
