ViewVC Help

View File | Revision Log | Show Annotations | Root Listing

root/cvsroot/COMP/CRAB/python/cms_cmssw.py

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.155 by slacapra, Tue Feb 12 15:20:47 2008 UTC vs.
Revision 1.264 by slacapra, Tue Dec 16 14:23:07 2008 UTC

#	Line 2 | Line 2 | from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	<	from BlackWhiteListParser import BlackWhiteListParser
5	>	from WMCore.SiteScreening.BlackWhiteListParser import SEBlackWhiteListParser
6		import common
7		import Scram
8
9		import os, string, glob
10
11		class Cmssw(JobType):
12	<	def __init__(self, cfg_params, ncjobs):
12	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
13		JobType.__init__(self, 'CMSSW')
14		common.logger.debug(3,'CMSSW::__init__')
15	<
15	>	self.skip_blocks = skip_blocks
16		self.argsList = []
17
18		self._params = {}
19		self.cfg_params = cfg_params
20	–	# init BlackWhiteListParser
21	–	self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
20
21	<	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
21	>	# init BlackWhiteListParser
22	>	seWhiteList = cfg_params.get('EDG.se_white_list',[])
23	>	seBlackList = cfg_params.get('EDG.se_black_list',[])
24	>	self.blackWhiteListParser = SEBlackWhiteListParser(seWhiteList, seBlackList, common.logger)
25	>
26	>	### Temporary patch to automatically skip the ISB size check:
27	>	server=self.cfg_params.get('CRAB.server_name',None)
28	>	size = 9.5
29	>	if server or common.scheduler.name().upper() in ['LSF','CAF']: size = 99999
30	>	### D.S.
31	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',size))
32
33		# number of jobs requested to be created, limit obj splitting
34		self.ncjobs = ncjobs
#	Line 33 | Line 41 | class Cmssw(JobType):
41		self.executable = ''
42		self.executable_arch = self.scram.getArch()
43		self.tgz_name = 'default.tgz'
36	–	self.additional_tgz_name = 'additional.tgz'
44		self.scriptName = 'CMSSW.sh'
45	<	self.pset = ''      #scrip use case Da
46	<	self.datasetPath = '' #scrip use case Da
45	>	self.pset = ''
46	>	self.datasetPath = ''
47
48		# set FJR file name
49		self.fjrFileName = 'crab_fjr.xml'
50
51		self.version = self.scram.getSWVersion()
52	<
53	<	#
54	<	# Try to block creation in case of arch/version mismatch
55	<	#
56	<
57	<	a = string.split(self.version, "_")
58	<
59	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
60	<	msg = "Warning: You are using %s version of CMSSW  with %s architecture. \n--> Did you compile your libraries with SLC3? Otherwise you can find some problems running on SLC4 Grid nodes.\n"%(self.version, self.executable_arch)
61	<	common.logger.message(msg)
55	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
56	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
52	>	version_array = self.version.split('_')
53	>	self.CMSSW_major = 0
54	>	self.CMSSW_minor = 0
55	>	self.CMSSW_patch = 0
56	>	try:
57	>	self.CMSSW_major = int(version_array[1])
58	>	self.CMSSW_minor = int(version_array[2])
59	>	self.CMSSW_patch = int(version_array[3])
60	>	except:
61	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
62		raise CrabException(msg)
63
59	–	common.taskDB.setDict('codeVersion',self.version)
60	–	self.setParam_('application', self.version)
61	–
64		### collect Data cards
65
66		if not cfg_params.has_key('CMSSW.datasetpath'):
67		msg = "Error: datasetpath not defined "
68		raise CrabException(msg)
69	+
70	+	### Temporary: added to remove input file control in the case of PU
71	+	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
72	+
73		tmp =  cfg_params['CMSSW.datasetpath']
74		log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	<	if string.lower(tmp)=='none':
75	>
76	>	if tmp =='':
77	>	msg = "Error: datasetpath not defined "
78	>	raise CrabException(msg)
79	>	elif string.lower(tmp)=='none':
80		self.datasetPath = None
81		self.selectNoInput = 1
82		else:
83		self.datasetPath = tmp
84		self.selectNoInput = 0
85
76	–	# ML monitoring
77	–	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
78	–	if not self.datasetPath:
79	–	self.setParam_('dataset', 'None')
80	–	self.setParam_('owner', 'None')
81	–	else:
82	–	## SL what is supposed to fail here?
83	–	try:
84	–	datasetpath_split = self.datasetPath.split("/")
85	–	# standard style
86	–	self.setParam_('datasetFull', self.datasetPath)
87	–	self.setParam_('dataset', datasetpath_split[1])
88	–	self.setParam_('owner', datasetpath_split[2])
89	–	except:
90	–	self.setParam_('dataset', self.datasetPath)
91	–	self.setParam_('owner', self.datasetPath)
92	–
93	–	self.setParam_('taskId', common.taskDB.dict('taskId'))
94	–
86		self.dataTiers = []
87	<
87	>	self.debugWrap = ''
88	>	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
89	>	if self.debug_wrapper: self.debugWrap='--debug'
90		## now the application
91	+	self.managedGenerators = ['madgraph','comphep']
92	+	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
93		self.executable = cfg_params.get('CMSSW.executable','cmsRun')
99	–	self.setParam_('exe', self.executable)
94		log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
95
96		if not cfg_params.has_key('CMSSW.pset'):
#	Line 117 | Line 111 | class Cmssw(JobType):
111		self.output_file_sandbox.append(self.fjrFileName)
112
113		# other output files to be returned via sandbox or copied to SE
114	+	outfileflag = False
115		self.output_file = []
116		tmp = cfg_params.get('CMSSW.output_file',None)
117		if tmp :
118	<	tmpOutFiles = string.split(tmp,',')
119	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
120	<	for tmp in tmpOutFiles:
121	<	tmp=string.strip(tmp)
127	<	self.output_file.append(tmp)
128	<	pass
129	<	else:
130	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
131	<	pass
118	>	self.output_file = [x.strip() for x in tmp.split(',')]
119	>	outfileflag = True #output found
120	>	#else:
121	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
122
123		# script_exe file as additional file in inputSandbox
124		self.scriptExe = cfg_params.get('USER.script_exe',None)
125		if self.scriptExe :
126	<	if not os.path.isfile(self.scriptExe):
127	<	msg ="ERROR. file "+self.scriptExe+" not found"
128	<	raise CrabException(msg)
129	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
126	>	if not os.path.isfile(self.scriptExe):
127	>	msg ="ERROR. file "+self.scriptExe+" not found"
128	>	raise CrabException(msg)
129	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
130
141	–	#CarlosDaniele
131		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
132	<	msg ="Error. script_exe  not defined"
133	<	raise CrabException(msg)
132	>	msg ="Error. script_exe  not defined"
133	>	raise CrabException(msg)
134	>
135	>	# use parent files...
136	>	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
137
138		## additional input files
139		if cfg_params.has_key('USER.additional_input_files'):
#	Line 161 | Line 153 | class Cmssw(JobType):
153		if not os.path.exists(file):
154		raise CrabException("Additional input file not found: "+file)
155		pass
164	–	# fname = string.split(file, '/')[-1]
165	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
166	–	# shutil.copyfile(file, storedFile)
156		self.additional_inbox_files.append(string.strip(file))
157		pass
158		pass
#	Line 189 | Line 178 | class Cmssw(JobType):
178		if cfg_params.has_key('CMSSW.total_number_of_events'):
179		self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
180		self.selectTotalNumberEvents = 1
181	+	if self.selectNumberOfJobs  == 1:
182	+	if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs):
183	+	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
184	+	raise CrabException(msg)
185		else:
186		self.total_number_of_events = 0
187		self.selectTotalNumberEvents = 0
188
189	<	if self.pset != None: #CarlosDaniele
189	>	if self.pset != None:
190		if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
191		msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
192		raise CrabException(msg)
#	Line 202 | Line 195 | class Cmssw(JobType):
195		msg = 'Must specify  number_of_jobs.'
196		raise CrabException(msg)
197
198	<	## source seed for pythia
199	<	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
198	>	## New method of dealing with seeds
199	>	self.incrementSeeds = []
200	>	self.preserveSeeds = []
201	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
202	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
203	>	for tmp in tmpList:
204	>	tmp.strip()
205	>	self.preserveSeeds.append(tmp)
206	>	if cfg_params.has_key('CMSSW.increment_seeds'):
207	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
208	>	for tmp in tmpList:
209	>	tmp.strip()
210	>	self.incrementSeeds.append(tmp)
211
212	+	## FUTURE: Can remove in CRAB 2.4.0
213	+	self.sourceSeed    = cfg_params.get('CMSSW.pythia_seed',None)
214		self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
215	<
210	<	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
211	<
215	>	self.sourceSeedG4  = cfg_params.get('CMSSW.g4_seed',None)
216		self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
217	+	if self.sourceSeed or self.sourceSeedVtx or self.sourceSeedG4 or self.sourceSeedMix:
218	+	msg = 'pythia_seed, vtx_seed, g4_seed, and mix_seed are no longer valid settings. You must use increment_seeds or preserve_seeds'
219	+	raise CrabException(msg)
220
221		self.firstRun = cfg_params.get('CMSSW.first_run',None)
222
216	–	if self.pset != None: #CarlosDaniele
217	–	import PsetManipulator as pp
218	–	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
219	–
223		# Copy/return
221	–
224		self.copy_data = int(cfg_params.get('USER.copy_data',0))
225		self.return_data = int(cfg_params.get('USER.return_data',0))
226
#	Line 234 | Line 236 | class Cmssw(JobType):
236		blockSites = self.DataDiscoveryAndLocation(cfg_params)
237		#DBSDLS-end
238
237	–	self.tgzNameWithPath = self.getTarBall(self.executable)
238	–
239		## Select Splitting
240		if self.selectNoInput:
241	<	if self.pset == None: #CarlosDaniele
241	>	if self.pset == None:
242		self.jobSplittingForScript()
243		else:
244		self.jobSplittingNoInput()
245	+	elif (cfg_params.get('CMSSW.noblockboundary',0)):
246	+	self.jobSplittingNoBlockBoundary(blockSites)
247		else:
248		self.jobSplittingByBlocks(blockSites)
249
250	<	# modify Pset
251	<	if self.pset != None: #CarlosDaniele
252	<	try:
253	<	if (self.datasetPath): # standard job
254	<	# allow to processa a fraction of events in a file
255	<	PsetEdit.inputModule("INPUTFILE")
256	<	PsetEdit.maxEvent(0)
257	<	PsetEdit.skipEvent(0)
258	<	else:  # pythia like job
250	>	# modify Pset only the first time
251	>	if isNew:
252	>	if self.pset != None:
253	>	import PsetManipulator as pp
254	>	PsetEdit = pp.PsetManipulator(self.pset)
255	>	try:
256	>	# Add FrameworkJobReport to parameter-set, set max events.
257	>	# Reset later for data jobs by writeCFG which does all modifications
258	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
259		PsetEdit.maxEvent(self.eventsPerJob)
260	<	if (self.firstRun):
261	<	PsetEdit.pythiaFirstRun(0)  #First Run
262	<	if (self.sourceSeed) :
263	<	PsetEdit.pythiaSeed(0)
264	<	if (self.sourceSeedVtx) :
265	<	PsetEdit.vtxSeed(0)
266	<	if (self.sourceSeedG4) :
267	<	PsetEdit.g4Seed(0)
268	<	if (self.sourceSeedMix) :
269	<	PsetEdit.mixSeed(0)
270	<	# add FrameworkJobReport to parameter-set
271	<	PsetEdit.addCrabFJR(self.fjrFileName)
272	<	PsetEdit.psetWriter(self.configFilename())
273	<	except:
274	<	msg='Error while manipuliating ParameterSet: exiting...'
275	<	raise CrabException(msg)
260	>	PsetEdit.psetWriter(self.configFilename())
261	>	## If present, add TFileService to output files
262	>	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
263	>	tfsOutput = PsetEdit.getTFileService()
264	>	if tfsOutput:
265	>	if tfsOutput in self.output_file:
266	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
267	>	else:
268	>	outfileflag = True #output found
269	>	self.output_file.append(tfsOutput)
270	>	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
271	>	pass
272	>	pass
273	>	## If present and requested, add PoolOutputModule to output files
274	>	if int(cfg_params.get('CMSSW.get_edm_output',0)):
275	>	edmOutput = PsetEdit.getPoolOutputModule()
276	>	if edmOutput:
277	>	if edmOutput in self.output_file:
278	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
279	>	else:
280	>	self.output_file.append(edmOutput)
281	>	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
282	>	pass
283	>	pass
284	>	except CrabException:
285	>	msg='Error while manipulating ParameterSet: exiting...'
286	>	raise CrabException(msg)
287	>	## Prepare inputSandbox TarBall (only the first time)
288	>	self.tgzNameWithPath = self.getTarBall(self.executable)
289
290		def DataDiscoveryAndLocation(self, cfg_params):
291
#	Line 283 | Line 298 | class Cmssw(JobType):
298		## Contact the DBS
299		common.logger.message("Contacting Data Discovery Services ...")
300		try:
301	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
301	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
302		self.pubdata.fetchDBSInfo()
303
304		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 299 | Line 314 | class Cmssw(JobType):
314		self.filesbyblock=self.pubdata.getFiles()
315		self.eventsbyblock=self.pubdata.getEventsPerBlock()
316		self.eventsbyfile=self.pubdata.getEventsPerFile()
317	+	self.parentFiles=self.pubdata.getParent()
318
319		## get max number of events
320	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
320	>	self.maxEvents=self.pubdata.getMaxEvents()
321
322		## Contact the DLS and build a list of sites hosting the fileblocks
323		try:
324		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
325		dataloc.fetchDLSInfo()
326	+
327		except DataLocation.DataLocationError , ex:
328		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
329		raise CrabException(msg)
330
331
332		sites = dataloc.getSites()
333	+	if len(sites)==0:
334	+	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n %s \n'%datasetPath
335	+	msg += "Maybe the dataset is located only at T1's, where analysis jobs are not allowed\n"
336	+	msg += "Please check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
337	+	raise CrabException(msg)
338	+
339		allSites = []
340		listSites = sites.values()
341		for listSite in listSites:
#	Line 325 | Line 348 | class Cmssw(JobType):
348
349		return sites
350
328	–	def setArgsList(self, argsList):
329	–	self.argsList = argsList
330	–
351		def jobSplittingByBlocks(self, blockSites):
352		"""
353		Perform job splitting. Jobs run over an integer number of files
#	Line 378 | Line 398 | class Cmssw(JobType):
398		else :
399		totalNumberOfJobs = self.ncjobs
400
381	–
401		blocks = blockSites.keys()
402		blockCount = 0
403		# Backup variable in case self.maxEvents counted events in a non-included block
#	Line 419 | Line 438 | class Cmssw(JobType):
438
439		# ---- Iterate over the files in the block until we've met the requested ---- #
440		# ---- total # of events or we've gone over all the files in this block  ---- #
441	+	pString=''
442		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
443		file = files[fileCount]
444	+	if self.useParent:
445	+	parent = self.parentFiles[file]
446	+	for f in parent :
447	+	pString += '\\\"' + f + '\\\"\,'
448	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
449	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
450		if newFile :
451		try:
452		numEventsInFile = self.eventsbyfile[file]
#	Line 433 | Line 459 | class Cmssw(JobType):
459		except KeyError:
460		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
461
462	<
462	>	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
463		# if less events in file remain than eventsPerJobRequested
464	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
464	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
465		# if last file in block
466		if ( fileCount == numFilesInBlock-1 ) :
467		# end job using last file, use remaining events in block
468		# close job and touch new file
469		fullString = parString[:-2]
470	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
470	>	if self.useParent:
471	>	fullParentString = pString[:-2]
472	>	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
473	>	else:
474	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
475		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
476		self.jobDestination.append(blockSites[block])
477		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 453 | Line 483 | class Cmssw(JobType):
483		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
484		jobSkipEventCount = 0
485		# reset file
486	+	pString = ""
487		parString = ""
488		filesEventCount = 0
489		newFile = 1
#	Line 465 | Line 496 | class Cmssw(JobType):
496		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
497		# close job and touch new file
498		fullString = parString[:-2]
499	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
499	>	if self.useParent:
500	>	fullParentString = pString[:-2]
501	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
502	>	else:
503	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
504		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
505		self.jobDestination.append(blockSites[block])
506		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 476 | Line 511 | class Cmssw(JobType):
511		eventsRemaining = eventsRemaining - eventsPerJobRequested
512		jobSkipEventCount = 0
513		# reset file
514	+	pString = ""
515		parString = ""
516		filesEventCount = 0
517		newFile = 1
#	Line 485 | Line 521 | class Cmssw(JobType):
521		else :
522		# close job but don't touch new file
523		fullString = parString[:-2]
524	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
524	>	if self.useParent:
525	>	fullParentString = pString[:-2]
526	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
527	>	else:
528	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
529		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
530		self.jobDestination.append(blockSites[block])
531		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 499 | Line 539 | class Cmssw(JobType):
539		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
540		# remove all but the last file
541		filesEventCount = self.eventsbyfile[file]
542	<	parString = ""
543	<	parString += '\\\"' + file + '\\\"\,'
542	>	if self.useParent:
543	>	for f in parent : pString += '\\\"' + f + '\\\"\,'
544	>	parString = '\\\"' + file + '\\\"\,'
545		pass # END if
546		pass # END while (iterate over files in the block)
547		pass # END while (iterate over blocks in the dataset)
#	Line 520 | Line 561 | class Cmssw(JobType):
561		for block in blocks:
562		if block in jobsOfBlock.keys() :
563		blockCounter += 1
564	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
564	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
565	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
566		if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
567		noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
568		bloskNoSite.append( blockCounter )
#	Line 554 | Line 596 | class Cmssw(JobType):
596		self.list_of_args = list_of_lists
597		return
598
599	+	def jobSplittingNoBlockBoundary(self,blockSites):
600	+	"""
601	+	"""
602	+	# ---- Handle the possible job splitting configurations ---- #
603	+	if (self.selectTotalNumberEvents):
604	+	totalEventsRequested = self.total_number_of_events
605	+	if (self.selectEventsPerJob):
606	+	eventsPerJobRequested = self.eventsPerJob
607	+	if (self.selectNumberOfJobs):
608	+	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
609	+
610	+	# If user requested all the events in the dataset
611	+	if (totalEventsRequested == -1):
612	+	eventsRemaining=self.maxEvents
613	+	# If user requested more events than are in the dataset
614	+	elif (totalEventsRequested > self.maxEvents):
615	+	eventsRemaining = self.maxEvents
616	+	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
617	+	# If user requested less events than are in the dataset
618	+	else:
619	+	eventsRemaining = totalEventsRequested
620	+
621	+	# If user requested more events per job than are in the dataset
622	+	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
623	+	eventsPerJobRequested = self.maxEvents
624	+
625	+	# For user info at end
626	+	totalEventCount = 0
627	+
628	+	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
629	+	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
630	+
631	+	if (self.selectNumberOfJobs):
632	+	common.logger.message("May not create the exact number_of_jobs requested.")
633	+
634	+	if ( self.ncjobs == 'all' ) :
635	+	totalNumberOfJobs = 999999999
636	+	else :
637	+	totalNumberOfJobs = self.ncjobs
638	+
639	+	blocks = blockSites.keys()
640	+	blockCount = 0
641	+	# Backup variable in case self.maxEvents counted events in a non-included block
642	+	numBlocksInDataset = len(blocks)
643	+
644	+	jobCount = 0
645	+	list_of_lists = []
646	+
647	+	#AF
648	+	#AF do not reset input files and event count on block boundary
649	+	#AF
650	+	parString=""
651	+	filesEventCount = 0
652	+	#AF
653	+
654	+	# list tracking which jobs are in which jobs belong to which block
655	+	jobsOfBlock = {}
656	+	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
657	+	block = blocks[blockCount]
658	+	blockCount += 1
659	+	if block not in jobsOfBlock.keys() :
660	+	jobsOfBlock[block] = []
661	+
662	+	if self.eventsbyblock.has_key(block) :
663	+	numEventsInBlock = self.eventsbyblock[block]
664	+	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
665	+	files = self.filesbyblock[block]
666	+	numFilesInBlock = len(files)
667	+	if (numFilesInBlock <= 0):
668	+	continue
669	+	fileCount = 0
670	+	#AF
671	+	#AF do not reset input files and event count of block boundary
672	+	#AF
673	+	## ---- New block => New job ---- #
674	+	#parString = ""
675	+	# counter for number of events in files currently worked on
676	+	#filesEventCount = 0
677	+	#AF
678	+	# flag if next while loop should touch new file
679	+	newFile = 1
680	+	# job event counter
681	+	jobSkipEventCount = 0
682	+
683	+	# ---- Iterate over the files in the block until we've met the requested ---- #
684	+	# ---- total # of events or we've gone over all the files in this block  ---- #
685	+	pString=''
686	+	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
687	+	file = files[fileCount]
688	+	if self.useParent:
689	+	parent = self.parentFiles[file]
690	+	for f in parent :
691	+	pString += '\\\"' + f + '\\\"\,'
692	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
693	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
694	+	if newFile :
695	+	try:
696	+	numEventsInFile = self.eventsbyfile[file]
697	+	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
698	+	# increase filesEventCount
699	+	filesEventCount += numEventsInFile
700	+	# Add file to current job
701	+	parString += '\\\"' + file + '\\\"\,'
702	+	newFile = 0
703	+	except KeyError:
704	+	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
705	+	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
706	+	#common.logger.message("AF filesEventCount %s - jobSkipEventCount %s "%(filesEventCount,jobSkipEventCount))
707	+	# if less events in file remain than eventsPerJobRequested
708	+	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
709	+	#AF
710	+	#AF skip fileboundary part
711	+	#AF
712	+	# go to next file
713	+	newFile = 1
714	+	fileCount += 1
715	+	# if events in file equal to eventsPerJobRequested
716	+	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
717	+	# close job and touch new file
718	+	fullString = parString[:-2]
719	+	if self.useParent:
720	+	fullParentString = pString[:-2]
721	+	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
722	+	else:
723	+	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
724	+	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
725	+	self.jobDestination.append(blockSites[block])
726	+	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
727	+	jobsOfBlock[block].append(jobCount+1)
728	+	# reset counter
729	+	jobCount = jobCount + 1
730	+	totalEventCount = totalEventCount + eventsPerJobRequested
731	+	eventsRemaining = eventsRemaining - eventsPerJobRequested
732	+	jobSkipEventCount = 0
733	+	# reset file
734	+	pString = ""
735	+	parString = ""
736	+	filesEventCount = 0
737	+	newFile = 1
738	+	fileCount += 1
739	+
740	+	# if more events in file remain than eventsPerJobRequested
741	+	else :
742	+	# close job but don't touch new file
743	+	fullString = parString[:-2]
744	+	if self.useParent:
745	+	fullParentString = pString[:-2]
746	+	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
747	+	else:
748	+	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
749	+	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
750	+	self.jobDestination.append(blockSites[block])
751	+	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
752	+	jobsOfBlock[block].append(jobCount+1)
753	+	# increase counter
754	+	jobCount = jobCount + 1
755	+	totalEventCount = totalEventCount + eventsPerJobRequested
756	+	eventsRemaining = eventsRemaining - eventsPerJobRequested
757	+	# calculate skip events for last file
758	+	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
759	+	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
760	+	# remove all but the last file
761	+	filesEventCount = self.eventsbyfile[file]
762	+	if self.useParent:
763	+	for f in parent : pString += '\\\"' + f + '\\\"\,'
764	+	parString = '\\\"' + file + '\\\"\,'
765	+	pass # END if
766	+	pass # END while (iterate over files in the block)
767	+	pass # END while (iterate over blocks in the dataset)
768	+	self.ncjobs = self.total_number_of_jobs = jobCount
769	+	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
770	+	common.logger.message("eventsRemaining "+str(eventsRemaining))
771	+	common.logger.message("jobCount "+str(jobCount))
772	+	common.logger.message(" totalNumberOfJobs "+str(totalNumberOfJobs))
773	+	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
774	+	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
775	+
776	+	# screen output
777	+	screenOutput = "List of jobs and available destination sites:\n\n"
778	+
779	+	#AF
780	+	#AF   skip check on  block with no sites
781	+	#AF
782	+	self.list_of_args = list_of_lists
783	+
784	+	return
785	+
786	+
787	+
788		def jobSplittingNoInput(self):
789		"""
790		Perform job splitting based on number of event per job
#	Line 602 | Line 833 | class Cmssw(JobType):
833		if (self.firstRun):
834		## pythia first run
835		args.append(str(self.firstRun)+str(i))
836	<	if (self.sourceSeed):
837	<	args.append(str(self.sourceSeed)+str(i))
838	<	if (self.sourceSeedVtx):
839	<	## + vtx random seed
840	<	args.append(str(self.sourceSeedVtx)+str(i))
841	<	if (self.sourceSeedG4):
611	<	## + G4 random seed
612	<	args.append(str(self.sourceSeedG4)+str(i))
613	<	if (self.sourceSeedMix):
614	<	## + Mix random seed
615	<	args.append(str(self.sourceSeedMix)+str(i))
616	<	pass
617	<	pass
836	>	if (self.generator in self.managedGenerators):
837	>	if (self.generator == 'comphep' and i == 0):
838	>	# COMPHEP is brain-dead and wants event #'s like 1,100,200,300
839	>	args.append('1')
840	>	else:
841	>	args.append(str(i*self.eventsPerJob))
842		self.list_of_args.append(args)
619	–	pass
620	–
843		return
844
845
846	<	def jobSplittingForScript(self):#CarlosDaniele
846	>	def jobSplittingForScript(self):
847		"""
848		Perform job splitting based on number of job
849		"""
#	Line 637 | Line 859 | class Cmssw(JobType):
859		# argument is seed number.$i
860		self.list_of_args = []
861		for i in range(self.total_number_of_jobs):
640	–	## Since there is no input, any site is good
641	–	# self.jobDestination.append(["Any"])
862		self.jobDestination.append([""])
643	–	## no random seed
863		self.list_of_args.append([str(i)])
864		return
865
866	<	def split(self, jobParams):
866	>	def split(self, jobParams,firstJobID):
867
649	–	common.jobDB.load()
650	–	#### Fabio
868		njobs = self.total_number_of_jobs
869		arglist = self.list_of_args
870	+	if njobs==0:
871	+	raise CrabException("Ask to split "+str(njobs)+" jobs: aborting")
872	+
873		# create the empty structure
874		for i in range(njobs):
875		jobParams.append("")
876
877	<	for job in range(njobs):
878	<	jobParams[job] = arglist[job]
879	<	# print str(arglist[job])
880	<	# print jobParams[job]
881	<	common.jobDB.setArguments(job, jobParams[job])
882	<	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
883	<	common.jobDB.setDestination(job, self.jobDestination[job])
877	>	listID=[]
878	>	listField=[]
879	>	for id in range(njobs):
880	>	job = id + int(firstJobID)
881	>	jobParams[id] = arglist[id]
882	>	listID.append(job+1)
883	>	job_ToSave ={}
884	>	concString = ' '
885	>	argu=''
886	>	if len(jobParams[id]):
887	>	argu +=   concString.join(jobParams[id] )
888	>	job_ToSave['arguments']= str(job+1)+' '+argu
889	>	job_ToSave['dlsDestination']= self.jobDestination[id]
890	>	listField.append(job_ToSave)
891	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
892	>	+"                     Destination: "+str(self.jobDestination[id])
893	>	common.logger.debug(5,msg)
894	>	common._db.updateJob_(listID,listField)
895	>	self.argsList = (len(jobParams[0])+1)
896
665	–	common.jobDB.save()
897		return
898
668	–	def getJobTypeArguments(self, nj, sched):
669	–	result = ''
670	–	for i in common.jobDB.arguments(nj):
671	–	result=result+str(i)+" "
672	–	return result
673	–
899		def numberOfJobs(self):
675	–	# Fabio
900		return self.total_number_of_jobs
901
902		def getTarBall(self, exe):
903		"""
904		Return the TarBall with lib and exe
905		"""
906	<
683	<	# if it exist, just return it
684	<	#
685	<	# Marco. Let's start to use relative path for Boss XML files
686	<	#
687	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
906	>	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
907		if os.path.exists(self.tgzNameWithPath):
908		return self.tgzNameWithPath
909
#	Line 697 | Line 916 | class Cmssw(JobType):
916
917		# First of all declare the user Scram area
918		swArea = self.scram.getSWArea_()
700	–	#print "swArea = ", swArea
701	–	# swVersion = self.scram.getSWVersion()
702	–	# print "swVersion = ", swVersion
919		swReleaseTop = self.scram.getReleaseTop_()
704	–	#print "swReleaseTop = ", swReleaseTop
920
921		## check if working area is release top
922		if swReleaseTop == '' or swArea == swReleaseTop:
923	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
924		return
925
926		import tarfile
#	Line 733 | Line 949 | class Cmssw(JobType):
949		pass
950
951		## Now get the libraries: only those in local working area
952	+	tar.dereference=True
953		libDir = 'lib'
954		lib = swArea+'/' +libDir
955		common.logger.debug(5,"lib "+lib+" to be tarred")
#	Line 744 | Line 961 | class Cmssw(JobType):
961		module = swArea + '/' + moduleDir
962		if os.path.isdir(module):
963		tar.add(module,moduleDir)
964	+	tar.dereference=False
965
966		## Now check if any data dir(s) is present
967	<	swAreaLen=len(swArea)
968	<	for root, dirs, files in os.walk(swArea):
969	<	if "data" in dirs:
970	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
971	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
972	<
973	<	### Removed ProdAgent Api dependencies ###
974	<	### Add ProdAgent dir to tar
975	<	#paDir = 'ProdAgentApi'
976	<	#pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
977	<	#if os.path.isdir(pa):
978	<	#    tar.add(pa,paDir)
967	>	self.dataExist = False
968	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
969	>	while len(todo_list):
970	>	entry, name = todo_list.pop()
971	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
972	>	continue
973	>	if os.path.isdir(swArea+"/src/"+entry):
974	>	entryPath = entry + '/'
975	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
976	>	if name == 'data':
977	>	self.dataExist=True
978	>	common.logger.debug(5,"data "+entry+" to be tarred")
979	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
980	>	pass
981	>	pass
982	>
983	>	### CMSSW ParameterSet
984	>	if not self.pset is None:
985	>	cfg_file = common.work_space.jobDir()+self.configFilename()
986	>	tar.add(cfg_file,self.configFilename())
987	>
988
989		## Add ProdCommon dir to tar
990	<	prodcommonDir = 'ProdCommon'
991	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
992	<	if os.path.isdir(prodcommonPath):
993	<	tar.add(prodcommonPath,prodcommonDir)
990	>	prodcommonDir = './'
991	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
992	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
993	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage']
994	>	for file in neededStuff:
995	>	tar.add(prodcommonPath+file,prodcommonDir+file)
996	>
997	>	##### ML stuff
998	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
999	>	path=os.environ['CRABDIR'] + '/python/'
1000	>	for file in ML_file_list:
1001	>	tar.add(path+file,file)
1002	>
1003	>	##### Utils
1004	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
1005	>	for file in Utils_file_list:
1006	>	tar.add(path+file,file)
1007	>
1008	>	##### AdditionalFiles
1009	>	tar.dereference=True
1010	>	for file in self.additional_inbox_files:
1011	>	tar.add(file,string.split(file,'/')[-1])
1012	>	tar.dereference=False
1013	>	common.logger.debug(5,"Files in "+self.tgzNameWithPath+" : "+str(tar.getnames()))
1014
768	–	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
1015		tar.close()
1016	<	except :
1017	<	raise CrabException('Could not create tar-ball')
1016	>	except IOError, exc:
1017	>	common.logger.write(str(exc))
1018	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
1019	>	except tarfile.TarError, exc:
1020	>	common.logger.write(str(exc))
1021	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
1022
1023		## check for tarball size
1024		tarballinfo = os.stat(self.tgzNameWithPath)
1025		if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
1026	<	raise CrabException('Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) + ' MB input sandbox limit and not supported by the used GRID submission system. Please make sure that no unnecessary files are in all data directories in your local CMSSW project area as they are automatically packed into the input sandbox.')
1026	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) \
1027	>	+'MB input sandbox limit \n'
1028	>	msg += '      and not supported by the direct GRID submission system.\n'
1029	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
1030	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/CrabServer#CRABSERVER_for_Users'
1031	>	raise CrabException(msg)
1032
1033		## create tar-ball with ML stuff
779	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
780	–	try:
781	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
782	–	path=os.environ['CRABDIR'] + '/python/'
783	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
784	–	tar.add(path+file,file)
785	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
786	–	tar.close()
787	–	except :
788	–	raise CrabException('Could not create ML files tar-ball')
789	–
790	–	return
1034
1035	<	def additionalInputFileTgz(self):
793	<	"""
794	<	Put all additional files into a tar ball and return its name
795	<	"""
796	<	import tarfile
797	<	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
798	<	tar = tarfile.open(tarName, "w:gz")
799	<	for file in self.additional_inbox_files:
800	<	tar.add(file,string.split(file,'/')[-1])
801	<	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
802	<	tar.close()
803	<	return tarName
804	<
805	<	def wsSetupEnvironment(self, nj):
1035	>	def wsSetupEnvironment(self, nj=0):
1036		"""
1037		Returns part of a job script which prepares
1038		the execution environment for the job 'nj'.
1039		"""
1040	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1041	+	psetName = 'pset.py'
1042	+	else:
1043	+	psetName = 'pset.cfg'
1044		# Prepare JobType-independent part
1045	<	txt = ''
1045	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
1046		txt += 'echo ">>> setup environment"\n'
1047		txt += 'if [ $middleware == LCG ]; then \n'
1048		txt += self.wsSetupCMSLCGEnvironment_()
1049		txt += 'elif [ $middleware == OSG ]; then\n'
1050		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
1051		txt += '    if [ ! $? == 0 ] ;then\n'
1052	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
1053	<	txt += '        echo "JOB_EXIT_STATUS = 10016"\n'
1054	<	txt += '        echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
821	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
822	<	txt += '        exit 1\n'
1052	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
1053	>	txt += '        job_exit_code=10016\n'
1054	>	txt += '        func_exit\n'
1055		txt += '    fi\n'
1056		txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
1057		txt += '\n'
#	Line 827 | Line 1059 | class Cmssw(JobType):
1059		txt += '    cd $WORKING_DIR\n'
1060		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
1061		txt += self.wsSetupCMSOSGEnvironment_()
830	–	#txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
831	–	#txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1062		txt += 'fi\n'
1063
1064		# Prepare JobType-specific part
#	Line 839 | Line 1069 | class Cmssw(JobType):
1069		txt += scram+' project CMSSW '+self.version+'\n'
1070		txt += 'status=$?\n'
1071		txt += 'if [ $status != 0 ] ; then\n'
1072	<	txt += '    echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
1073	<	txt += '    echo "JOB_EXIT_STATUS = 10034"\n'
1074	<	txt += '    echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
845	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
846	<	txt += '    if [ $middleware == OSG ]; then \n'
847	<	txt += '        cd $RUNTIME_AREA\n'
848	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
849	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
850	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
851	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
852	<	txt += '            echo "SET_CMS_ENV 10018 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after CMSSW CMSSW_0_6_1 not found on `hostname`"\n'
853	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
854	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
855	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
856	<	txt += '        fi\n'
857	<	txt += '    fi \n'
858	<	txt += '    exit 1 \n'
1072	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
1073	>	txt += '    job_exit_code=10034\n'
1074	>	txt += '    func_exit\n'
1075		txt += 'fi \n'
1076		txt += 'cd '+self.version+'\n'
861	–	########## FEDE FOR DBS2 ######################
1077		txt += 'SOFTWARE_DIR=`pwd`\n'
1078		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
864	–	###############################################
865	–	### needed grep for bug in scramv1 ###
1079		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
1080	+	txt += 'if [ $? != 0 ] ; then\n'
1081	+	txt += '    echo "ERROR ==> Problem with the command: "\n'
1082	+	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
1083	+	txt += '    job_exit_code=10034\n'
1084	+	txt += '    func_exit\n'
1085	+	txt += 'fi \n'
1086		# Handle the arguments:
1087		txt += "\n"
1088		txt += "## number of arguments (first argument always jobnumber)\n"
1089		txt += "\n"
1090	<	txt += "if [ $nargs -lt "+str(len(self.argsList[nj].split()))+" ]\n"
1090	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
1091		txt += "then\n"
1092	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
1093	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
1094	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
876	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
877	<	txt += '    if [ $middleware == OSG ]; then \n'
878	<	txt += '        cd $RUNTIME_AREA\n'
879	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
880	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
881	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
882	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
883	<	txt += '            echo "SET_EXE_ENV 50114 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Too few arguments for CRAB job wrapper"\n'
884	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
885	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
886	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
887	<	txt += '        fi\n'
888	<	txt += '    fi \n'
889	<	txt += "    exit 1\n"
1092	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
1093	>	txt += '    job_exit_code=50113\n'
1094	>	txt += "    func_exit\n"
1095		txt += "fi\n"
1096		txt += "\n"
1097
1098		# Prepare job-specific part
1099		job = common.job_list[nj]
895	–	### FEDE FOR DBS OUTPUT PUBLICATION
1100		if (self.datasetPath):
1101	+	self.primaryDataset = self.datasetPath.split("/")[1]
1102	+	DataTier = self.datasetPath.split("/")[2]
1103		txt += '\n'
1104		txt += 'DatasetPath='+self.datasetPath+'\n'
1105
1106	<	datasetpath_split = self.datasetPath.split("/")
1107	<
902	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
903	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
1106	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
1107	>	txt += 'DataTier='+DataTier+'\n'
1108		txt += 'ApplicationFamily=cmsRun\n'
1109
1110		else:
1111	+	self.primaryDataset = 'null'
1112		txt += 'DatasetPath=MCDataTier\n'
1113		txt += 'PrimaryDataset=null\n'
1114		txt += 'DataTier=null\n'
1115		txt += 'ApplicationFamily=MCDataTier\n'
1116	<	if self.pset != None: #CarlosDaniele
1116	>	if self.pset != None:
1117		pset = os.path.basename(job.configFilename())
1118		txt += '\n'
1119		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
1120		if (self.datasetPath): # standard job
1121	<	txt += 'InputFiles=${args[1]}\n'
1122	<	txt += 'MaxEvents=${args[2]}\n'
1123	<	txt += 'SkipEvents=${args[3]}\n'
1121	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
1122	>	if (self.useParent):
1123	>	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
1124	>	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
1125	>	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
1126	>	else:
1127	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
1128	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
1129		txt += 'echo "Inputfiles:<$InputFiles>"\n'
1130	<	txt += 'sed "s#\'INPUTFILE\'#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1130	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
1131		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
922	–	txt += 'sed "s#int32 input = 0#int32 input = $MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1132		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
924	–	txt += 'sed "s#uint32 skipEvents = 0#uint32 skipEvents = $SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1133		else:  # pythia like job
1134	<	seedIndex=1
1134	>	argNum = 1
1135	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
1136	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
1137	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
1138	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
1139		if (self.firstRun):
1140	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
1140	>	txt += 'export FirstRun=${args[%s]}\n' % argNum
1141		txt += 'echo "FirstRun: <$FirstRun>"\n'
1142	<	txt += 'sed "s#uint32 firstRun = 0#uint32 firstRun = $FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1143	<	seedIndex=seedIndex+1
1142	>	argNum += 1
1143	>	if (self.generator == 'madgraph'):
1144	>	txt += 'export FirstEvent=${args[%s]}\n' % argNum
1145	>	txt += 'echo "FirstEvent:<$FirstEvent>"\n'
1146	>	argNum += 1
1147	>	elif (self.generator == 'comphep'):
1148	>	txt += 'export CompHEPFirstEvent=${args[%s]}\n' % argNum
1149	>	txt += 'echo "CompHEPFirstEvent:<$CompHEPFirstEvent>"\n'
1150	>	argNum += 1
1151
1152	<	if (self.sourceSeed):
934	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
935	<	txt += 'sed "s#uint32 sourceSeed = 0#uint32 sourceSeed = $Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
936	<	seedIndex=seedIndex+1
937	<	## the following seeds are not always present
938	<	if (self.sourceSeedVtx):
939	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
940	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
941	<	txt += 'sed "s#uint32 VtxSmeared = 0#uint32 VtxSmeared = $VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
942	<	seedIndex += 1
943	<	if (self.sourceSeedG4):
944	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
945	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
946	<	txt += 'sed "s#uint32 g4SimHits = 0#uint32 g4SimHits = $G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
947	<	seedIndex += 1
948	<	if (self.sourceSeedMix):
949	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
950	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
951	<	txt += 'sed "s#uint32 mix = 0#uint32 mix = $mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
952	<	seedIndex += 1
953	<	pass
954	<	pass
955	<	txt += 'mv -f '+pset+' pset.cfg\n'
1152	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
1153
957	–	if len(self.additional_inbox_files) > 0:
958	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
959	–	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
960	–	txt += 'fi\n'
961	–	pass
1154
1155	<	if self.pset != None: #CarlosDaniele
1156	<	txt += '\n'
965	<	txt += 'echo "***** cat pset.cfg *********"\n'
966	<	txt += 'cat pset.cfg\n'
967	<	txt += 'echo "****** end pset.cfg ********"\n'
1155	>	if self.pset != None:
1156	>	# FUTURE: Can simply for 2_1_x and higher
1157		txt += '\n'
1158	<	### FEDE FOR DBS OUTPUT PUBLICATION
1159	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
1158	>	if self.debug_wrapper==True:
1159	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
1160	>	txt += 'cat ' + psetName + '\n'
1161	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
1162	>	txt += '\n'
1163	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1164	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
1165	>	else:
1166	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
1167		txt += 'echo "PSETHASH = $PSETHASH" \n'
972	–	##############
1168		txt += '\n'
1169		return txt
1170
1171	<	def wsBuildExe(self, nj=0):
1171	>	def wsUntarSoftware(self, nj=0):
1172		"""
1173		Put in the script the commands to build an executable
1174		or a library.
1175		"""
1176
1177	<	txt = ""
1177	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
1178
1179		if os.path.isfile(self.tgzNameWithPath):
1180		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
1181	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1181	>	txt += 'tar xzf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1182	>	if  self.debug_wrapper:
1183	>	txt += 'tar tzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
1184	>	txt += 'ls -Al \n'
1185		txt += 'untar_status=$? \n'
1186		txt += 'if [ $untar_status -ne 0 ]; then \n'
1187	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
1188	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
1189	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
992	<	txt += '   if [ $middleware == OSG ]; then \n'
993	<	txt += '       cd $RUNTIME_AREA\n'
994	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
995	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
996	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
997	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
998	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
999	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1000	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1001	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1002	<	txt += '       fi\n'
1003	<	txt += '   fi \n'
1004	<	txt += '   \n'
1005	<	txt += '   exit 1 \n'
1187	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
1188	>	txt += '   job_exit_code=$untar_status\n'
1189	>	txt += '   func_exit\n'
1190		txt += 'else \n'
1191		txt += '   echo "Successful untar" \n'
1192		txt += 'fi \n'
1193		txt += '\n'
1194	<	#### Removed ProdAgent API dependencies
1011	<	txt += 'echo ">>> Include ProdCommon in PYTHONPATH:"\n'
1194	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1195		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1196	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1014	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon\n'
1196	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1197		txt += 'else\n'
1198	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1198	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1199		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1018	–	###################
1200		txt += 'fi\n'
1201		txt += '\n'
1202
#	Line 1023 | Line 1204 | class Cmssw(JobType):
1204
1205		return txt
1206
1207	<	def modifySteeringCards(self, nj):
1207	>	def wsBuildExe(self, nj=0):
1208		"""
1209	<	modify the card provided by the user,
1210	<	writing a new card into share dir
1209	>	Put in the script the commands to build an executable
1210	>	or a library.
1211		"""
1212
1213	+	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
1214	+	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
1215	+
1216	+	txt += 'rm -r lib/ module/ \n'
1217	+	txt += 'mv $RUNTIME_AREA/lib/ . \n'
1218	+	txt += 'mv $RUNTIME_AREA/module/ . \n'
1219	+	if self.dataExist == True:
1220	+	txt += 'rm -r src/ \n'
1221	+	txt += 'mv $RUNTIME_AREA/src/ . \n'
1222	+	if len(self.additional_inbox_files)>0:
1223	+	for file in self.additional_inbox_files:
1224	+	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
1225	+	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
1226	+	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
1227	+
1228	+	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1229	+	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1230	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1231	+	txt += 'else\n'
1232	+	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1233	+	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1234	+	txt += 'fi\n'
1235	+	txt += '\n'
1236	+
1237	+	return txt
1238	+
1239	+
1240		def executableName(self):
1241	<	if self.scriptExe: #CarlosDaniele
1241	>	if self.scriptExe:
1242		return "sh "
1243		else:
1244		return self.executable
1245
1246		def executableArgs(self):
1247	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1248		if self.scriptExe:#CarlosDaniele
1249		return   self.scriptExe + " $NJob"
1250		else:
1251	<	# if >= CMSSW_1_5_X, add -j crab_fjr.xml
1252	<	version_array = self.scram.getSWVersion().split('_')
1253	<	major = 0
1254	<	minor = 0
1255	<	try:
1256	<	major = int(version_array[1])
1257	<	minor = int(version_array[2])
1258	<	except:
1050	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1051	<	raise CrabException(msg)
1052	<	if major >= 1 and minor >= 5 :
1053	<	return " -j " + self.fjrFileName + " -p pset.cfg"
1251	>	ex_args = ""
1252	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1253	>	# Framework job report
1254	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1255	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1256	>	# Type of config file
1257	>	if self.CMSSW_major >= 2 :
1258	>	ex_args += " -p pset.py"
1259		else:
1260	<	return " -p pset.cfg"
1260	>	ex_args += " -p pset.cfg"
1261	>	return ex_args
1262
1263		def inputSandbox(self, nj):
1264		"""
1265		Returns a list of filenames to be put in JDL input sandbox.
1266		"""
1267		inp_box = []
1062	–	# # dict added to delete duplicate from input sandbox file list
1063	–	# seen = {}
1064	–	## code
1268		if os.path.isfile(self.tgzNameWithPath):
1269		inp_box.append(self.tgzNameWithPath)
1270	<	if os.path.isfile(self.MLtgzfile):
1068	<	inp_box.append(self.MLtgzfile)
1069	<	## config
1070	<	if not self.pset is None:
1071	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1072	<	## additional input files
1073	<	tgz = self.additionalInputFileTgz()
1074	<	inp_box.append(tgz)
1270	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
1271		return inp_box
1272
1273		def outputSandbox(self, nj):
#	Line 1083 | Line 1279 | class Cmssw(JobType):
1279		## User Declared output files
1280		for out in (self.output_file+self.output_file_sandbox):
1281		n_out = nj + 1
1282	<	out_box.append(self.numberFile_(out,str(n_out)))
1282	>	out_box.append(numberFile(out,str(n_out)))
1283		return out_box
1284
1089	–	def prepareSteeringCards(self):
1090	–	"""
1091	–	Make initial modifications of the user's steering card file.
1092	–	"""
1093	–	return
1285
1286		def wsRenameOutput(self, nj):
1287		"""
1288		Returns part of a job script which renames the produced files.
1289		"""
1290
1291	<	txt = '\n'
1291	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1292		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1293		txt += 'echo ">>> current directory content:"\n'
1294	<	txt += 'ls \n'
1294	>	if self.debug_wrapper:
1295	>	txt += 'ls -Al\n'
1296		txt += '\n'
1297
1106	–	txt += 'output_exit_status=0\n'
1107	–
1108	–	for fileWithSuffix in (self.output_file_sandbox):
1109	–	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1110	–	txt += '\n'
1111	–	txt += '# check output file\n'
1112	–	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1113	–	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1114	–	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1115	–	txt += 'else\n'
1116	–	txt += '    exit_status=60302\n'
1117	–	txt += '    echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1118	–	if common.scheduler.name() == 'CONDOR_G':
1119	–	txt += '    if [ $middleware == OSG ]; then \n'
1120	–	txt += '        echo "prepare dummy output file"\n'
1121	–	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1122	–	txt += '    fi \n'
1123	–	txt += 'fi\n'
1124	–
1298		for fileWithSuffix in (self.output_file):
1299	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1299	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1300		txt += '\n'
1301		txt += '# check output file\n'
1302		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
#	Line 1134 | Line 1307 | class Cmssw(JobType):
1307		txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1308		txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1309		txt += 'else\n'
1310	<	txt += '    exit_status=60302\n'
1311	<	txt += '    echo "ERROR: Output file '+fileWithSuffix+' not found"\n'
1312	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1140	<	txt += '    output_exit_status=$exit_status\n'
1141	<	if common.scheduler.name() == 'CONDOR_G':
1310	>	txt += '    job_exit_code=60302\n'
1311	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1312	>	if common.scheduler.name().upper() == 'CONDOR_G':
1313		txt += '    if [ $middleware == OSG ]; then \n'
1314		txt += '        echo "prepare dummy output file"\n'
1315		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1146 | Line 1317 | class Cmssw(JobType):
1317		txt += 'fi\n'
1318		file_list = []
1319		for fileWithSuffix in (self.output_file):
1320	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1320	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$NJob'))
1321
1322	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1322	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
1323		txt += '\n'
1324		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1325		txt += 'echo ">>> current directory content:"\n'
1326	<	txt += 'ls \n'
1326	>	if self.debug_wrapper:
1327	>	txt += 'ls -Al\n'
1328		txt += '\n'
1329		txt += 'cd $RUNTIME_AREA\n'
1330		txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1331		return txt
1332
1161	–	def numberFile_(self, file, txt):
1162	–	"""
1163	–	append _'txt' before last extension of a file
1164	–	"""
1165	–	p = string.split(file,".")
1166	–	# take away last extension
1167	–	name = p[0]
1168	–	for x in p[1:-1]:
1169	–	name=name+"."+x
1170	–	# add "_txt"
1171	–	if len(p)>1:
1172	–	ext = p[len(p)-1]
1173	–	result = name + '_' + txt + "." + ext
1174	–	else:
1175	–	result = name + '_' + txt
1176	–
1177	–	return result
1178	–
1333		def getRequirements(self, nj=[]):
1334		"""
1335		return job requirements to add to jdl files
#	Line 1185 | Line 1339 | class Cmssw(JobType):
1339		req='Member("VO-cms-' + \
1340		self.version + \
1341		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1342	<	## SL add requirement for OS version only if SL4
1189	<	#reSL4 = re.compile( r'slc4' )
1190	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
1342	>	if self.executable_arch:
1343		req+=' && Member("VO-cms-' + \
1344		self.executable_arch + \
1345		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1346
1347		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1348	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
1349	+	req += ' && other.GlueCEStateStatus == "Production" '
1350
1351		return req
1352
1353		def configFilename(self):
1354		""" return the config filename """
1355	<	return self.name()+'.cfg'
1355	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1356	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1357	>	return self.name()+'.py'
1358	>	else:
1359	>	return self.name()+'.cfg'
1360
1361		def wsSetupCMSOSGEnvironment_(self):
1362		"""
1363		Returns part of a job script which is prepares
1364		the execution environment and which is common for all CMS jobs.
1365		"""
1366	<	txt = '    echo ">>> setup CMS OSG environment:"\n'
1366	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1367	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1368		txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1369		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1370		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
#	Line 1213 | Line 1372 | class Cmssw(JobType):
1372		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1373		txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1374		txt += '    else\n'
1375	<	txt += '        echo "SET_CMS_ENV 10020 ==> ERROR $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1376	<	txt += '        echo "JOB_EXIT_STATUS = 10020"\n'
1377	<	txt += '        echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1219	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1220	<	txt += '\n'
1221	<	txt += '        cd $RUNTIME_AREA\n'
1222	<	txt += '        echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1223	<	txt += '        echo ">>> Remove working directory: $WORKING_DIR"\n'
1224	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
1225	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
1226	<	txt += '            echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1227	<	txt += '            echo "JOB_EXIT_STATUS = 10017"\n'
1228	<	txt += '            echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1229	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1230	<	txt += '        fi\n'
1231	<	txt += '\n'
1232	<	txt += '        exit 1\n'
1375	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1376	>	txt += '        job_exit_code=10020\n'
1377	>	txt += '        func_exit\n'
1378		txt += '    fi\n'
1379		txt += '\n'
1380	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1380	>	txt += '    echo "==> setup cms environment ok"\n'
1381		txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1382
1383		return txt
1384
1240	–	### OLI_DANIELE
1385		def wsSetupCMSLCGEnvironment_(self):
1386		"""
1387		Returns part of a job script which is prepares
1388		the execution environment and which is common for all CMS jobs.
1389		"""
1390	<	txt = '    echo ">>> setup CMS LCG environment:"\n'
1390	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1391	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1392		txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1393		txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1394		txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1395		txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1396	<	txt += '        echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1397	<	txt += '        echo "JOB_EXIT_STATUS = 10031" \n'
1398	<	txt += '        echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1254	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1255	<	txt += '        exit 1\n'
1396	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1397	>	txt += '        job_exit_code=10031\n'
1398	>	txt += '        func_exit\n'
1399		txt += '    else\n'
1400		txt += '        echo "Sourcing environment... "\n'
1401		txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1402	<	txt += '            echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1403	<	txt += '            echo "JOB_EXIT_STATUS = 10020"\n'
1404	<	txt += '            echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1262	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1263	<	txt += '            exit 1\n'
1402	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1403	>	txt += '            job_exit_code=10020\n'
1404	>	txt += '            func_exit\n'
1405		txt += '        fi\n'
1406		txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1407		txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1408		txt += '        result=$?\n'
1409		txt += '        if [ $result -ne 0 ]; then\n'
1410	<	txt += '            echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1411	<	txt += '            echo "JOB_EXIT_STATUS = 10032"\n'
1412	<	txt += '            echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1272	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
1273	<	txt += '            exit 1\n'
1410	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1411	>	txt += '            job_exit_code=10032\n'
1412	>	txt += '            func_exit\n'
1413		txt += '        fi\n'
1414		txt += '    fi\n'
1415		txt += '    \n'
1416	<	txt += '    echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1416	>	txt += '    echo "==> setup cms environment ok"\n'
1417		return txt
1418
1419	<	### FEDE FOR DBS OUTPUT PUBLICATION
1281	<	def modifyReport(self, nj):
1419	>	def wsModifyReport(self, nj):
1420		"""
1421		insert the part of the script that modifies the FrameworkJob Report
1422		"""
1423	<
1424	<	txt = ''
1287	<	try:
1288	<	publish_data = int(self.cfg_params['USER.publish_data'])
1289	<	except KeyError:
1290	<	publish_data = 0
1423	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
1424	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1425		if (publish_data == 1):
1292	–	txt += 'echo ">>> Modify Job Report:" \n'
1293	–	################ FEDE FOR DBS2 #############################################
1294	–	#txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1295	–	txt += 'chmod a+x $SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1296	–	#############################################################################
1297	–
1298	–	txt += 'if [ -z "$SE" ]; then\n'
1299	–	txt += '    SE="" \n'
1300	–	txt += 'fi \n'
1301	–	txt += 'if [ -z "$SE_PATH" ]; then\n'
1302	–	txt += '    SE_PATH="" \n'
1303	–	txt += 'fi \n'
1304	–	txt += 'echo "SE = $SE"\n'
1305	–	txt += 'echo "SE_PATH = $SE_PATH"\n'
1426
1427		processedDataset = self.cfg_params['USER.publish_data_name']
1428	<	txt += 'ProcessedDataset='+processedDataset+'\n'
1429	<	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1430	<	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1431	<	#### FEDE: added slash in LFN ##############
1428	>
1429	>	txt += 'if [ $StageOutExitStatus -eq 0 ]; then\n'
1430	>	txt += '    FOR_LFN=$LFNBaseName\n'
1431	>	txt += 'else\n'
1432		txt += '    FOR_LFN=/copy_problems/ \n'
1433	<	txt += 'else \n'
1434	<	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1435	<	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1436	<	txt += '    FOR_LFN=/store$tmp \n'
1437	<	txt += 'fi \n'
1433	>	txt += '    SE=""\n'
1434	>	txt += '    SE_PATH=""\n'
1435	>	txt += 'fi\n'
1436	>
1437	>	txt += 'echo ">>> Modify Job Report:" \n'
1438	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1439	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1440	>	#txt += 'ProcessedDataset=$procDataset \n'
1441		txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1442	+	txt += 'echo "SE = $SE"\n'
1443	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1444		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1445		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1446	<	txt += 'echo "$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1447	<	txt += '$SOFTWARE_DIR/ProdCommon/ProdCommon/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1448	<
1446	>	args = '$RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' \
1447	>	'$USER-$ProcessedDataset-$PSETHASH $ApplicationFamily '+ \
1448	>	'  $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH'
1449	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1450	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1451		txt += 'modifyReport_result=$?\n'
1325	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1452		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1453	<	txt += '    exit_status=1\n'
1454	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1453	>	txt += '    modifyReport_result=70500\n'
1454	>	txt += '    job_exit_code=$modifyReport_result\n'
1455	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1456	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1457		txt += 'else\n'
1458	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1458	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1459		txt += 'fi\n'
1332	–	else:
1333	–	txt += 'echo "no data publication required"\n'
1460		return txt
1461
1462	<	def cleanEnv(self):
1463	<	txt = ''
1464	<	txt += 'if [ $middleware == OSG ]; then\n'
1465	<	txt += '    cd $RUNTIME_AREA\n'
1466	<	txt += '    echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1467	<	txt += '    echo ">>> Remove working directory: $WORKING_DIR"\n'
1468	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1469	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1470	<	txt += '        echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1471	<	txt += '        echo "JOB_EXIT_STATUS = 60999"\n'
1472	<	txt += '        echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1473	<	txt += '        dumpStatus $RUNTIME_AREA/$repo\n'
1462	>	def wsParseFJR(self):
1463	>	"""
1464	>	Parse the FrameworkJobReport to obtain useful infos
1465	>	"""
1466	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1467	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1468	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1469	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1470	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1471	>	if self.debug_wrapper :
1472	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1473	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1474	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1475	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1476	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1477	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1478	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1479	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1480	>	txt += '        else\n'
1481	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1482	>	txt += '        fi\n'
1483	>	txt += '    else\n'
1484	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1485	>	txt += '    fi\n'
1486	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1487	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1488	>	txt += '      echo ">>> Executable succeded  $executable_exit_status"\n'
1489	>	if (self.datasetPath and not (self.dataset_pu or self.useParent)) :
1490	>	# VERIFY PROCESSED DATA
1491	>	txt += '      echo ">>> Verify list of processed files:"\n'
1492	>	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1493	>	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1494	>	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1495	>	txt += '      mv tmp.txt input-files.txt\n'
1496	>	txt += '      echo "cat input-files.txt"\n'
1497	>	txt += '      echo "----------------------"\n'
1498	>	txt += '      cat input-files.txt\n'
1499	>	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1500	>	txt += '      mv tmp.txt processed-files.txt\n'
1501	>	txt += '      echo "----------------------"\n'
1502	>	txt += '      echo "cat processed-files.txt"\n'
1503	>	txt += '      echo "----------------------"\n'
1504	>	txt += '      cat processed-files.txt\n'
1505	>	txt += '      echo "----------------------"\n'
1506	>	txt += '      diff -q input-files.txt processed-files.txt\n'
1507	>	txt += '      fileverify_status=$?\n'
1508	>	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1509	>	txt += '         executable_exit_status=30001\n'
1510	>	txt += '         echo "ERROR ==> not all input files processed"\n'
1511	>	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1512	>	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1513	>	txt += '      fi\n'
1514	>	txt += '    elif [ $executable_exit_status -ne 0 ] || [ $executable_exit_status -ne 50015 ] || [ $executable_exit_status -ne 50017 ];then\n'
1515	>	txt += '      echo ">>> Executable failed  $executable_exit_status"\n'
1516	>	txt += '      echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1517	>	txt += '      echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1518	>	txt += '      job_exit_code=$executable_exit_status\n'
1519	>	txt += '      func_exit\n'
1520		txt += '    fi\n'
1521	+	txt += '\n'
1522	+	txt += 'else\n'
1523	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1524		txt += 'fi\n'
1525		txt += '\n'
1526	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1527	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1528	+	txt += 'job_exit_code=$executable_exit_status\n'
1529	+
1530		return txt
1531
1532		def setParam_(self, param, value):
#	Line 1365 | Line 1544 | class Cmssw(JobType):
1544		nd[e]=0
1545		return nd.keys()
1546
1547	<
1369	<	def checkOut(self, limit):
1547	>	def outList(self,list=False):
1548		"""
1549		check the dimension of the output files
1550		"""
1551	<	txt = 'echo ">>> Starting output sandbox limit check :"\n'
1552	<	allOutFiles = ""
1551	>	txt = ''
1552	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1553		listOutFiles = []
1554	<	txt += 'stdoutFile=`ls *stdout` \n'
1555	<	txt += 'stderrFile=`ls *stderr` \n'
1554	>	stdout = 'CMSSW_$NJob.stdout'
1555	>	stderr = 'CMSSW_$NJob.stderr'
1556		if (self.return_data == 1):
1557	<	for fileOut in (self.output_file+self.output_file_sandbox):
1558	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1559	<	else:
1560	<	for fileOut in (self.output_file_sandbox):
1561	<	txt += 'echo " '+fileOut+'";\n'
1562	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob') + " $stdoutFile $stderrFile"
1563	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1564	<	txt += 'ls -gGhrta;\n'
1565	<	txt += 'sum=0;\n'
1566	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1567	<	txt += '    if [ -e $file ]; then\n'
1568	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1569	<	txt += '        sum=`expr $sum + $tt`\n'
1392	<	txt += '    else\n'
1393	<	txt += '        echo "WARNING: output file $file not found!"\n'
1394	<	txt += '    fi\n'
1395	<	txt += 'done\n'
1396	<	txt += 'echo "Total Output dimension: $sum";\n'
1397	<	txt += 'limit='+str(limit)+';\n'
1398	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1399	<	txt += 'if [ $limit -lt $sum ]; then\n'
1400	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1401	<	txt += '    echo "         checking the output file sizes..."\n'
1402	<	txt += '    tot=0;\n'
1403	<	txt += '    for filefile in '+str(allOutFiles)+' ; do\n'
1404	<	txt += '        dimFile=`ls -gGrta $filefile | awk \'{ print $3 }\';`\n'
1405	<	txt += '        tot=`expr $tot + $tt`;\n'
1406	<	txt += '        if [ $limit -lt $dimFile ]; then\n'
1407	<	txt += '            echo "deleting file: $filefile";\n'
1408	<	txt += '            rm -f $filefile\n'
1409	<	txt += '        elif [ $limit -lt $tot ]; then\n'
1410	<	txt += '            echo "deleting file: $filefile";\n'
1411	<	txt += '            rm -f $filefile\n'
1412	<	txt += '        else\n'
1413	<	txt += '            echo "saving file: $filefile"\n'
1414	<	txt += '        fi\n'
1415	<	txt += '    done\n'
1416	<
1417	<	txt += '    ls -agGhrt;\n'
1418	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1419	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1420	<	txt += '    exit_status=70000;\n'
1421	<	txt += 'else'
1422	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1423	<	txt += 'fi\n'
1424	<	txt += 'echo "Ending output sandbox limit check"\n'
1557	>	for file in (self.output_file+self.output_file_sandbox):
1558	>	listOutFiles.append(numberFile(file, '$NJob'))
1559	>	listOutFiles.append(stdout)
1560	>	listOutFiles.append(stderr)
1561	>	else:
1562	>	for file in (self.output_file_sandbox):
1563	>	listOutFiles.append(numberFile(file, '$NJob'))
1564	>	listOutFiles.append(stdout)
1565	>	listOutFiles.append(stderr)
1566	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1567	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1568	>	txt += 'export filesToCheck\n'
1569	>	if list : return self.output_file
1570		return txt

Diff Legend

–	Removed lines
+	Added lines
<	Changed lines
>	Changed lines