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 1.231 by spiga, Mon Aug 25 16:13:05 2008 UTC vs.
Revision 1.324 by spiga, Thu Jul 23 12:43:54 2009 UTC

#	Line 1 | Line 1
1		from JobType import JobType
2	–	from crab_logger import Logger
2		from crab_exceptions import *
3		from crab_util import *
5	–	from BlackWhiteListParser import SEBlackWhiteListParser
4		import common
5		import Scram
6	<	from LFNBaseName import *
6	>	from Splitter import JobSplitter
7
8	+	from IMProv.IMProvNode import IMProvNode
9		import os, string, glob
10
11		class Cmssw(JobType):
12		def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
13		JobType.__init__(self, 'CMSSW')
14	<	common.logger.debug(3,'CMSSW::__init__')
14	>	common.logger.debug('CMSSW::__init__')
15		self.skip_blocks = skip_blocks
16	<
17	<	self.argsList = []
19	<
16	>	self.argsList = 1
17	>	self.NumEvents=0
18		self._params = {}
19		self.cfg_params = cfg_params
22	–	# init BlackWhiteListParser
23	–	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
20
21	<	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
21	>	### Temporary patch to automatically skip the ISB size check:
22	>	self.server = self.cfg_params.get('CRAB.server_name',None) or \
23	>	self.cfg_params.get('CRAB.use_server',0)
24	>	size = 9.5
25	>	if self.server or common.scheduler.name().upper() in ['LSF','CAF']:
26	>	size = 99999
27	>	self.MaxTarBallSize = float(self.cfg_params.get('GRID.maxtarballsize',size))
28
29		# number of jobs requested to be created, limit obj splitting
30		self.ncjobs = ncjobs
31
30	–	log = common.logger
31	–
32		self.scram = Scram.Scram(cfg_params)
33		self.additional_inbox_files = []
34		self.scriptExe = ''
#	Line 39 | Line 39 | class Cmssw(JobType):
39		self.pset = ''
40		self.datasetPath = ''
41
42	+	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
43		# set FJR file name
44		self.fjrFileName = 'crab_fjr.xml'
45
46		self.version = self.scram.getSWVersion()
47	+	common.logger.log(10-1,"CMSSW version is: "+str(self.version))
48		version_array = self.version.split('_')
49		self.CMSSW_major = 0
50		self.CMSSW_minor = 0
#	Line 55 | Line 57 | class Cmssw(JobType):
57		msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
58		raise CrabException(msg)
59
60	+	if self.CMSSW_major < 1 or (self.CMSSW_major == 1 and self.CMSSW_minor < 5):
61	+	msg = "CRAB supports CMSSW >= 1_5_x only. Use an older CRAB version."
62	+	raise CrabException(msg)
63	+	"""
64	+	As CMSSW versions are dropped we can drop more code:
65	+	1.X dropped: drop support for running .cfg on WN
66	+	2.0 dropped: drop all support for cfg here and in writeCfg
67	+	2.0 dropped: Recheck the random number seed support
68	+	"""
69	+
70		### collect Data cards
71
60	–	if not cfg_params.has_key('CMSSW.datasetpath'):
61	–	msg = "Error: datasetpath not defined "
62	–	raise CrabException(msg)
72
73		### Temporary: added to remove input file control in the case of PU
74		self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
75
76		tmp =  cfg_params['CMSSW.datasetpath']
77	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
78	<	if string.lower(tmp)=='none':
77	>	common.logger.log(10-1, "CMSSW::CMSSW(): datasetPath = "+tmp)
78	>
79	>	if tmp =='':
80	>	msg = "Error: datasetpath not defined "
81	>	raise CrabException(msg)
82	>	elif string.lower(tmp)=='none':
83		self.datasetPath = None
84		self.selectNoInput = 1
85	+	self.primaryDataset = 'null'
86		else:
87		self.datasetPath = tmp
88		self.selectNoInput = 0
89	+	self.primaryDataset = self.datasetPath.split("/")[1]
90	+	self.dataTier = self.datasetPath.split("/")[2]
91
92		self.dataTiers = []
93	<	self.debugWrap = ''
94	<	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
95	<	if self.debug_wrapper: self.debugWrap='--debug'
93	>
94	>	self.debugWrap=''
95	>	self.debug_wrapper = int(cfg_params.get('USER.debug_wrapper',0))
96	>	if self.debug_wrapper == 1: self.debugWrap='--debug'
97	>
98		## now the application
99	+	self.managedGenerators = ['madgraph', 'comphep', 'lhe']
100	+	self.generator = cfg_params.get('CMSSW.generator','pythia').lower()
101		self.executable = cfg_params.get('CMSSW.executable','cmsRun')
102	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
102	>	common.logger.log(10-1, "CMSSW::CMSSW(): executable = "+self.executable)
103
104		if not cfg_params.has_key('CMSSW.pset'):
105		raise CrabException("PSet file missing. Cannot run cmsRun ")
106		self.pset = cfg_params['CMSSW.pset']
107	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
107	>	common.logger.log(10-1, "Cmssw::Cmssw(): PSet file = "+self.pset)
108		if self.pset.lower() != 'none' :
109		if (not os.path.exists(self.pset)):
110		raise CrabException("User defined PSet file "+self.pset+" does not exist")
#	Line 116 | Line 136 | class Cmssw(JobType):
136		raise CrabException(msg)
137		self.additional_inbox_files.append(string.strip(self.scriptExe))
138
139	+	self.AdditionalArgs = cfg_params.get('USER.script_arguments',None)
140	+	if self.AdditionalArgs : self.AdditionalArgs = string.replace(self.AdditionalArgs,',',' ')
141	+
142		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
143		msg ="Error. script_exe  not defined"
144		raise CrabException(msg)
145
146		# use parent files...
147	<	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
147	>	self.useParent = int(self.cfg_params.get('CMSSW.use_parent',0))
148
149		## additional input files
150		if cfg_params.has_key('USER.additional_input_files'):
#	Line 144 | Line 167 | class Cmssw(JobType):
167		self.additional_inbox_files.append(string.strip(file))
168		pass
169		pass
170	<	common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
170	>	common.logger.debug("Additional input files: "+str(self.additional_inbox_files))
171		pass
172
150	–	## Events per job
151	–	if cfg_params.has_key('CMSSW.events_per_job'):
152	–	self.eventsPerJob =int( cfg_params['CMSSW.events_per_job'])
153	–	self.selectEventsPerJob = 1
154	–	else:
155	–	self.eventsPerJob = -1
156	–	self.selectEventsPerJob = 0
157	–
158	–	## number of jobs
159	–	if cfg_params.has_key('CMSSW.number_of_jobs'):
160	–	self.theNumberOfJobs =int( cfg_params['CMSSW.number_of_jobs'])
161	–	self.selectNumberOfJobs = 1
162	–	else:
163	–	self.theNumberOfJobs = 0
164	–	self.selectNumberOfJobs = 0
165	–
166	–	if cfg_params.has_key('CMSSW.total_number_of_events'):
167	–	self.total_number_of_events = int(cfg_params['CMSSW.total_number_of_events'])
168	–	self.selectTotalNumberEvents = 1
169	–	if self.selectNumberOfJobs  == 1:
170	–	if (self.total_number_of_events != -1) and int(self.total_number_of_events) < int(self.theNumberOfJobs):
171	–	msg = 'Must specify at least one event per job. total_number_of_events > number_of_jobs '
172	–	raise CrabException(msg)
173	–	else:
174	–	self.total_number_of_events = 0
175	–	self.selectTotalNumberEvents = 0
176	–
177	–	if self.pset != None:
178	–	if ( (self.selectTotalNumberEvents + self.selectEventsPerJob + self.selectNumberOfJobs) != 2 ):
179	–	msg = 'Must define exactly two of total_number_of_events, events_per_job, or number_of_jobs.'
180	–	raise CrabException(msg)
181	–	else:
182	–	if (self.selectNumberOfJobs == 0):
183	–	msg = 'Must specify  number_of_jobs.'
184	–	raise CrabException(msg)
173
174		## New method of dealing with seeds
175		self.incrementSeeds = []
#	Line 197 | Line 185 | class Cmssw(JobType):
185		tmp.strip()
186		self.incrementSeeds.append(tmp)
187
200	–	## FUTURE: Can remove in CRAB 2.4.0
201	–	self.sourceSeed    = cfg_params.get('CMSSW.pythia_seed',None)
202	–	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
203	–	self.sourceSeedG4  = cfg_params.get('CMSSW.g4_seed',None)
204	–	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
205	–	if self.sourceSeed or self.sourceSeedVtx or self.sourceSeedG4 or self.sourceSeedMix:
206	–	msg = 'pythia_seed, vtx_seed, g4_seed, and mix_seed are no longer valid settings. You must use increment_seeds or preserve_seeds'
207	–	raise CrabException(msg)
208	–
188		self.firstRun = cfg_params.get('CMSSW.first_run',None)
189
190	<	# Copy/return
190	>	# Copy/return/publish
191		self.copy_data = int(cfg_params.get('USER.copy_data',0))
192		self.return_data = int(cfg_params.get('USER.return_data',0))
193	+	### FEDE ###
194	+	self.publish_data = int(cfg_params.get('USER.publish_data',0))
195	+	if (self.publish_data == 1):
196	+	if not cfg_params.has_key('USER.publish_data_name'):
197	+	raise CrabException('Cannot publish output data, because you did not specify USER.publish_data_name parameter in the crab.cfg file')
198	+	else:
199	+	self.processedDataset = cfg_params['USER.publish_data_name']
200	+	#### check of lenght of datasetname to publish ####
201	+	common.logger.debug("test 100 char limit on datasetname")
202	+	user = getUserName()
203	+	common.logger.debug("user = " + user)
204	+	len_user_name = len(user)
205	+	common.logger.debug("len_user_name = " + str(len_user_name))
206	+	len_processedDataset = len(self.processedDataset)
207	+	common.logger.debug("processedDataset " + self.processedDataset)
208	+	common.logger.debug("len_processedDataset = " + str(len_processedDataset))
209	+	if (self.datasetPath != None ):
210	+	len_primary = len(self.primaryDataset)
211	+	common.logger.debug("primaryDataset = " + self.primaryDataset)
212	+	common.logger.debug("len_primary = " + str(len_primary))
213	+	#common.logger.info("59 - len_user_name - len_primary = " + str(59 - len_user_name - len_primary))
214	+	if (len_processedDataset > (59 - len_user_name - len_primary)):
215	+	raise CrabException("Warning: publication name too long. USER.publish_data_name has to be < " + str(59 - len_user_name - len_primary) + " characters")
216	+	else:
217	+	if (len_processedDataset > (59 - len_user_name) / 2):
218	+	raise CrabException("Warning: publication name too long. USER.publish_data_name has to be < " + str((59 - len_user_name) / 2) + " characters")
219
220	+	self.conf = {}
221	+	self.conf['pubdata'] = None
222	+	# number of jobs requested to be created, limit obj splitting DD
223		#DBSDLS-start
224		## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
225		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
#	Line 223 | Line 231 | class Cmssw(JobType):
231		if self.datasetPath:
232		blockSites = self.DataDiscoveryAndLocation(cfg_params)
233		#DBSDLS-end
234	+	self.conf['blockSites']=blockSites
235
236		## Select Splitting
237	+	splitByRun = int(cfg_params.get('CMSSW.split_by_run',0))
238	+
239		if self.selectNoInput:
240		if self.pset == None:
241	<	self.jobSplittingForScript()
241	>	self.algo = 'ForScript'
242		else:
243	<	self.jobSplittingNoInput()
243	>	self.algo = 'NoInput'
244	>	self.conf['managedGenerators']=self.managedGenerators
245	>	self.conf['generator']=self.generator
246	>	elif splitByRun ==1:
247	>	self.algo = 'RunBased'
248		else:
249	<	self.jobSplittingByBlocks(blockSites)
249	>	self.algo = 'EventBased'
250	>
251	>	#        self.algo = 'LumiBased'
252	>	splitter = JobSplitter(self.cfg_params,self.conf)
253	>	self.dict = splitter.Algos()[self.algo]()
254
255	+	self.argsFile= '%s/arguments.xml'%common.work_space.shareDir()
256	+	self.rootArgsFilename= 'arguments'
257		# modify Pset only the first time
258		if isNew:
259	<	if self.pset != None:
260	<	import PsetManipulator as pp
240	<	PsetEdit = pp.PsetManipulator(self.pset)
241	<	try:
242	<	# Add FrameworkJobReport to parameter-set, set max events.
243	<	# Reset later for data jobs by writeCFG which does all modifications
244	<	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
245	<	PsetEdit.maxEvent(self.eventsPerJob)
246	<	PsetEdit.psetWriter(self.configFilename())
247	<	## If present, add TFileService to output files
248	<	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
249	<	tfsOutput = PsetEdit.getTFileService()
250	<	if tfsOutput:
251	<	if tfsOutput in self.output_file:
252	<	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
253	<	else:
254	<	outfileflag = True #output found
255	<	self.output_file.append(tfsOutput)
256	<	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
257	<	pass
258	<	pass
259	<	## If present and requested, add PoolOutputModule to output files
260	<	if int(cfg_params.get('CMSSW.get_edm_output',0)):
261	<	edmOutput = PsetEdit.getPoolOutputModule()
262	<	if edmOutput:
263	<	if edmOutput in self.output_file:
264	<	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
265	<	else:
266	<	self.output_file.append(edmOutput)
267	<	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
268	<	pass
269	<	pass
270	<	except CrabException:
271	<	msg='Error while manipulating ParameterSet: exiting...'
272	<	raise CrabException(msg)
259	>	if self.pset != None: self.ModifyPset()
260	>
261		## Prepare inputSandbox TarBall (only the first time)
262	<	self.tgzNameWithPath = self.getTarBall(self.executable)
262	>	self.tarNameWithPath = self.getTarBall(self.executable)
263	>
264	>
265	>	def ModifyPset(self):
266	>	import PsetManipulator as pp
267	>	PsetEdit = pp.PsetManipulator(self.pset)
268	>	try:
269	>	# Add FrameworkJobReport to parameter-set, set max events.
270	>	# Reset later for data jobs by writeCFG which does all modifications
271	>	PsetEdit.maxEvent(1)
272	>	PsetEdit.skipEvent(0)
273	>	PsetEdit.psetWriter(self.configFilename())
274	>	## If present, add TFileService to output files
275	>	if not int(self.cfg_params.get('CMSSW.skip_TFileService_output',0)):
276	>	tfsOutput = PsetEdit.getTFileService()
277	>	if tfsOutput:
278	>	if tfsOutput in self.output_file:
279	>	common.logger.debug("Output from TFileService "+tfsOutput+" already in output files")
280	>	else:
281	>	outfileflag = True #output found
282	>	self.output_file.append(tfsOutput)
283	>	common.logger.info("Adding "+tfsOutput+" (from TFileService) to list of output files")
284	>	pass
285	>	pass
286	>	# If present and requested, add PoolOutputModule to output files
287	>	edmOutput = PsetEdit.getPoolOutputModule()
288	>	if int(self.cfg_params.get('CMSSW.get_edm_output',0)):
289	>	if edmOutput:
290	>	for outputFile in edmOutput:
291	>	if outputFile in self.output_file:
292	>	common.logger.debug("Output from PoolOutputModule "+edmOutput+" already in output files")
293	>	else:
294	>	self.output_file.append(outputFile)
295	>	common.logger.info("Adding "+outputFile+" (from PoolOutputModule) to list of output files")
296	>	# not requested, check anyhow to avoid accidental T2 overload
297	>	else:
298	>	if edmOutput:
299	>	missedFiles = []
300	>	for outputFile in edmOutput:
301	>	if outputFile not in self.output_file:
302	>	missedFiles.append(outputFile)
303	>	if missedFiles:
304	>	msg  = "ERROR: PoolOutputModule(s) are present in your ParameteSet %s \n"%self.pset
305	>	msg += "    but the file(s) produced ( %s ) are not in the list of output files\n" % ', '.join(missedFiles)
306	>	msg += "WARNING: please remove them. If you want to keep them, add the file(s) to output_files or use CMSSW.get_edm_output = 1\n"
307	>	if int(self.cfg_params.get('CMSSW.ignore_edm_output',0)):
308	>	msg += "    CMSSW.ignore_edm_output==1 : Hope you know what you are doing...\n"
309	>	common.logger.info(msg)
310	>	else :
311	>	raise CrabException(msg)
312	>
313	>	if (PsetEdit.getBadFilesSetting()):
314	>	msg = "WARNING: You have set skipBadFiles to True. This will continue processing on some errors and you may not be notified."
315	>	common.logger.info(msg)
316	>
317	>	except CrabException, msg:
318	>	common.logger.info(str(msg))
319	>	msg='Error while manipulating ParameterSet (see previous message, if any): exiting...'
320	>	raise CrabException(msg)
321	>
322
323		def DataDiscoveryAndLocation(self, cfg_params):
324
325		import DataDiscovery
326		import DataLocation
327	<	common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
327	>	common.logger.log(10-1,"CMSSW::DataDiscoveryAndLocation()")
328
329		datasetPath=self.datasetPath
330
331		## Contact the DBS
332	<	common.logger.message("Contacting Data Discovery Services ...")
332	>	common.logger.info("Contacting Data Discovery Services ...")
333		try:
334		self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
335		self.pubdata.fetchDBSInfo()
#	Line 298 | Line 345 | class Cmssw(JobType):
345		raise CrabException(msg)
346
347		self.filesbyblock=self.pubdata.getFiles()
348	<	self.eventsbyblock=self.pubdata.getEventsPerBlock()
349	<	self.eventsbyfile=self.pubdata.getEventsPerFile()
303	<	self.parentFiles=self.pubdata.getParent()
348	>	#print self.filesbyblock
349	>	self.conf['pubdata']=self.pubdata
350
351		## get max number of events
352		self.maxEvents=self.pubdata.getMaxEvents()
#	Line 309 | Line 355 | class Cmssw(JobType):
355		try:
356		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
357		dataloc.fetchDLSInfo()
358	+
359		except DataLocation.DataLocationError , ex:
360		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
361		raise CrabException(msg)
362
363
364	<	sites = dataloc.getSites()
364	>	unsorted_sites = dataloc.getSites()
365	>	sites = self.filesbyblock.fromkeys(self.filesbyblock,'')
366	>	for lfn in self.filesbyblock.keys():
367	>	if unsorted_sites.has_key(lfn):
368	>	sites[lfn]=unsorted_sites[lfn]
369	>	else:
370	>	sites[lfn]=[]
371	>
372	>	if len(sites)==0:
373	>	msg = 'ERROR ***: no location for any of the blocks of this dataset: \n\t %s \n'%datasetPath
374	>	msg += "\tMaybe the dataset is located only at T1's (or at T0), where analysis jobs are not allowed\n"
375	>	msg += "\tPlease check DataDiscovery page https://cmsweb.cern.ch/dbs_discovery/\n"
376	>	raise CrabException(msg)
377	>
378		allSites = []
379		listSites = sites.values()
380		for listSite in listSites:
381		for oneSite in listSite:
382		allSites.append(oneSite)
383	<	allSites = self.uniquelist(allSites)
383	>	[allSites.append(it) for it in allSites if not allSites.count(it)]
384
325	–	# screen output
326	–	common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
327	–
328	–	return sites
329	–
330	–	def jobSplittingByBlocks(self, blockSites):
331	–	"""
332	–	Perform job splitting. Jobs run over an integer number of files
333	–	and no more than one block.
334	–	ARGUMENT: blockSites: dictionary with blocks as keys and list of host sites as values
335	–	REQUIRES: self.selectTotalNumberEvents, self.selectEventsPerJob, self.selectNumberofJobs,
336	–	self.total_number_of_events, self.eventsPerJob, self.theNumberOfJobs,
337	–	self.maxEvents, self.filesbyblock
338	–	SETS: self.jobDestination - Site destination(s) for each job (a list of lists)
339	–	self.total_number_of_jobs - Total # of jobs
340	–	self.list_of_args - File(s) job will run on (a list of lists)
341	–	"""
342	–
343	–	# ---- Handle the possible job splitting configurations ---- #
344	–	if (self.selectTotalNumberEvents):
345	–	totalEventsRequested = self.total_number_of_events
346	–	if (self.selectEventsPerJob):
347	–	eventsPerJobRequested = self.eventsPerJob
348	–	if (self.selectNumberOfJobs):
349	–	totalEventsRequested = self.theNumberOfJobs * self.eventsPerJob
350	–
351	–	# If user requested all the events in the dataset
352	–	if (totalEventsRequested == -1):
353	–	eventsRemaining=self.maxEvents
354	–	# If user requested more events than are in the dataset
355	–	elif (totalEventsRequested > self.maxEvents):
356	–	eventsRemaining = self.maxEvents
357	–	common.logger.message("Requested "+str(self.total_number_of_events)+ " events, but only "+str(self.maxEvents)+" events are available.")
358	–	# If user requested less events than are in the dataset
359	–	else:
360	–	eventsRemaining = totalEventsRequested
361	–
362	–	# If user requested more events per job than are in the dataset
363	–	if (self.selectEventsPerJob and eventsPerJobRequested > self.maxEvents):
364	–	eventsPerJobRequested = self.maxEvents
365	–
366	–	# For user info at end
367	–	totalEventCount = 0
368	–
369	–	if (self.selectTotalNumberEvents and self.selectNumberOfJobs):
370	–	eventsPerJobRequested = int(eventsRemaining/self.theNumberOfJobs)
371	–
372	–	if (self.selectNumberOfJobs):
373	–	common.logger.message("May not create the exact number_of_jobs requested.")
374	–
375	–	if ( self.ncjobs == 'all' ) :
376	–	totalNumberOfJobs = 999999999
377	–	else :
378	–	totalNumberOfJobs = self.ncjobs
379	–
380	–	blocks = blockSites.keys()
381	–	blockCount = 0
382	–	# Backup variable in case self.maxEvents counted events in a non-included block
383	–	numBlocksInDataset = len(blocks)
384	–
385	–	jobCount = 0
386	–	list_of_lists = []
387	–
388	–	# list tracking which jobs are in which jobs belong to which block
389	–	jobsOfBlock = {}
390	–
391	–	# ---- Iterate over the blocks in the dataset until ---- #
392	–	# ---- we've met the requested total # of events    ---- #
393	–	while ( (eventsRemaining > 0) and (blockCount < numBlocksInDataset) and (jobCount < totalNumberOfJobs)):
394	–	block = blocks[blockCount]
395	–	blockCount += 1
396	–	if block not in jobsOfBlock.keys() :
397	–	jobsOfBlock[block] = []
398	–
399	–	if self.eventsbyblock.has_key(block) :
400	–	numEventsInBlock = self.eventsbyblock[block]
401	–	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
402	–
403	–	files = self.filesbyblock[block]
404	–	numFilesInBlock = len(files)
405	–	if (numFilesInBlock <= 0):
406	–	continue
407	–	fileCount = 0
408	–
409	–	# ---- New block => New job ---- #
410	–	parString = ""
411	–	# counter for number of events in files currently worked on
412	–	filesEventCount = 0
413	–	# flag if next while loop should touch new file
414	–	newFile = 1
415	–	# job event counter
416	–	jobSkipEventCount = 0
417	–
418	–	# ---- Iterate over the files in the block until we've met the requested ---- #
419	–	# ---- total # of events or we've gone over all the files in this block  ---- #
420	–	pString=''
421	–	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
422	–	file = files[fileCount]
423	–	if self.useParent:
424	–	parent = self.parentFiles[file]
425	–	for f in parent :
426	–	pString += '\\\"' + f + '\\\"\,'
427	–	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
428	–	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
429	–	if newFile :
430	–	try:
431	–	numEventsInFile = self.eventsbyfile[file]
432	–	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
433	–	# increase filesEventCount
434	–	filesEventCount += numEventsInFile
435	–	# Add file to current job
436	–	parString += '\\\"' + file + '\\\"\,'
437	–	newFile = 0
438	–	except KeyError:
439	–	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
440	–
441	–	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
442	–	# if less events in file remain than eventsPerJobRequested
443	–	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
444	–	# if last file in block
445	–	if ( fileCount == numFilesInBlock-1 ) :
446	–	# end job using last file, use remaining events in block
447	–	# close job and touch new file
448	–	fullString = parString[:-2]
449	–	if self.useParent:
450	–	fullParentString = pString[:-2]
451	–	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
452	–	else:
453	–	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
454	–	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
455	–	self.jobDestination.append(blockSites[block])
456	–	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
457	–	# fill jobs of block dictionary
458	–	jobsOfBlock[block].append(jobCount+1)
459	–	# reset counter
460	–	jobCount = jobCount + 1
461	–	totalEventCount = totalEventCount + filesEventCount - jobSkipEventCount
462	–	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
463	–	jobSkipEventCount = 0
464	–	# reset file
465	–	pString = ""
466	–	parString = ""
467	–	filesEventCount = 0
468	–	newFile = 1
469	–	fileCount += 1
470	–	else :
471	–	# go to next file
472	–	newFile = 1
473	–	fileCount += 1
474	–	# if events in file equal to eventsPerJobRequested
475	–	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
476	–	# close job and touch new file
477	–	fullString = parString[:-2]
478	–	if self.useParent:
479	–	fullParentString = pString[:-2]
480	–	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
481	–	else:
482	–	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
483	–	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
484	–	self.jobDestination.append(blockSites[block])
485	–	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
486	–	jobsOfBlock[block].append(jobCount+1)
487	–	# reset counter
488	–	jobCount = jobCount + 1
489	–	totalEventCount = totalEventCount + eventsPerJobRequested
490	–	eventsRemaining = eventsRemaining - eventsPerJobRequested
491	–	jobSkipEventCount = 0
492	–	# reset file
493	–	pString = ""
494	–	parString = ""
495	–	filesEventCount = 0
496	–	newFile = 1
497	–	fileCount += 1
498	–
499	–	# if more events in file remain than eventsPerJobRequested
500	–	else :
501	–	# close job but don't touch new file
502	–	fullString = parString[:-2]
503	–	if self.useParent:
504	–	fullParentString = pString[:-2]
505	–	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
506	–	else:
507	–	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
508	–	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
509	–	self.jobDestination.append(blockSites[block])
510	–	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
511	–	jobsOfBlock[block].append(jobCount+1)
512	–	# increase counter
513	–	jobCount = jobCount + 1
514	–	totalEventCount = totalEventCount + eventsPerJobRequested
515	–	eventsRemaining = eventsRemaining - eventsPerJobRequested
516	–	# calculate skip events for last file
517	–	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
518	–	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
519	–	# remove all but the last file
520	–	filesEventCount = self.eventsbyfile[file]
521	–	if self.useParent:
522	–	for f in parent : pString += '\\\"' + f + '\\\"\,'
523	–	parString = '\\\"' + file + '\\\"\,'
524	–	pass # END if
525	–	pass # END while (iterate over files in the block)
526	–	pass # END while (iterate over blocks in the dataset)
527	–	self.ncjobs = self.total_number_of_jobs = jobCount
528	–	if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
529	–	common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
530	–	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
385
386		# screen output
387	<	screenOutput = "List of jobs and available destination sites:\n\n"
534	<
535	<	# keep trace of block with no sites to print a warning at the end
536	<	noSiteBlock = []
537	<	bloskNoSite = []
538	<
539	<	blockCounter = 0
540	<	for block in blocks:
541	<	if block in jobsOfBlock.keys() :
542	<	blockCounter += 1
543	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
544	<	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
545	<	if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
546	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
547	<	bloskNoSite.append( blockCounter )
548	<
549	<	common.logger.message(screenOutput)
550	<	if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
551	<	msg = 'WARNING: No sites are hosting any part of data for block:\n                '
552	<	virgola = ""
553	<	if len(bloskNoSite) > 1:
554	<	virgola = ","
555	<	for block in bloskNoSite:
556	<	msg += ' ' + str(block) + virgola
557	<	msg += '\n               Related jobs:\n                 '
558	<	virgola = ""
559	<	if len(noSiteBlock) > 1:
560	<	virgola = ","
561	<	for range_jobs in noSiteBlock:
562	<	msg += str(range_jobs) + virgola
563	<	msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
564	<	if self.cfg_params.has_key('EDG.se_white_list'):
565	<	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
566	<	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
567	<	msg += 'Please check if the dataset is available at this site!)\n'
568	<	if self.cfg_params.has_key('EDG.ce_white_list'):
569	<	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
570	<	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
571	<	msg += 'Please check if the dataset is available at this site!)\n'
572	<
573	<	common.logger.message(msg)
574	<
575	<	self.list_of_args = list_of_lists
576	<	return
577	<
578	<	def jobSplittingNoInput(self):
579	<	"""
580	<	Perform job splitting based on number of event per job
581	<	"""
582	<	common.logger.debug(5,'Splitting per events')
583	<
584	<	if (self.selectEventsPerJob):
585	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
586	<	if (self.selectNumberOfJobs):
587	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
588	<	if (self.selectTotalNumberEvents):
589	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
590	<
591	<	if (self.total_number_of_events < 0):
592	<	msg='Cannot split jobs per Events with "-1" as total number of events'
593	<	raise CrabException(msg)
594	<
595	<	if (self.selectEventsPerJob):
596	<	if (self.selectTotalNumberEvents):
597	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
598	<	elif(self.selectNumberOfJobs) :
599	<	self.total_number_of_jobs =self.theNumberOfJobs
600	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
601	<
602	<	elif (self.selectNumberOfJobs) :
603	<	self.total_number_of_jobs = self.theNumberOfJobs
604	<	self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
605	<
606	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
607	<
608	<	# is there any remainder?
609	<	check = int(self.total_number_of_events) - (int(self.total_number_of_jobs)*self.eventsPerJob)
610	<
611	<	common.logger.debug(5,'Check  '+str(check))
612	<
613	<	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created, each for '+str(self.eventsPerJob)+' for a total of '+str(self.total_number_of_jobs*self.eventsPerJob)+' events')
614	<	if check > 0:
615	<	common.logger.message('Warning: asked '+str(self.total_number_of_events)+' but can do only '+str(int(self.total_number_of_jobs)*self.eventsPerJob))
616	<
617	<	# argument is seed number.$i
618	<	self.list_of_args = []
619	<	for i in range(self.total_number_of_jobs):
620	<	## Since there is no input, any site is good
621	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
622	<	args=[]
623	<	if (self.firstRun):
624	<	## pythia first run
625	<	args.append(str(self.firstRun)+str(i))
626	<	self.list_of_args.append(args)
627	<
628	<	return
629	<
630	<
631	<	def jobSplittingForScript(self):
632	<	"""
633	<	Perform job splitting based on number of job
634	<	"""
635	<	common.logger.debug(5,'Splitting per job')
636	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
387	>	common.logger.info("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
388
389	<	self.total_number_of_jobs = self.theNumberOfJobs
639	<
640	<	common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
389	>	return sites
390
642	–	common.logger.message(str(self.total_number_of_jobs)+' jobs can be created')
643	–
644	–	# argument is seed number.$i
645	–	self.list_of_args = []
646	–	for i in range(self.total_number_of_jobs):
647	–	self.jobDestination.append([""])
648	–	self.list_of_args.append([str(i)])
649	–	return
391
392		def split(self, jobParams,firstJobID):
393
394	<	njobs = self.total_number_of_jobs
395	<	arglist = self.list_of_args
394	>	jobParams = self.dict['args']
395	>	njobs = self.dict['njobs']
396	>	self.jobDestination = self.dict['jobDestination']
397	>
398	>	if njobs==0:
399	>	raise CrabException("Ask to split "+str(njobs)+" jobs: aborting")
400	>	if not self.server and njobs > 500:
401	>	raise CrabException("The CRAB client will not submit more than 500 jobs. You must use the server mode.")
402	>
403		# create the empty structure
404		for i in range(njobs):
405		jobParams.append("")
406
407		listID=[]
408		listField=[]
409	+	listDictions=[]
410	+	exist= os.path.exists(self.argsFile)
411		for id in range(njobs):
412		job = id + int(firstJobID)
663	–	jobParams[id] = arglist[id]
413		listID.append(job+1)
414		job_ToSave ={}
415		concString = ' '
416		argu=''
417	+	str_argu = str(job+1)
418		if len(jobParams[id]):
419	<	argu +=   concString.join(jobParams[id] )
420	<	job_ToSave['arguments']= str(job+1)+' '+argu
419	>	argu = {'JobID': job+1}
420	>	for i in range(len(jobParams[id])):
421	>	argu[self.dict['params'][i]]=jobParams[id][i]
422	>	if len(jobParams[id])==1: self.NumEvents = jobParams[id][i]
423	>	# just for debug
424	>	str_argu += concString.join(jobParams[id])
425	>	if argu != '': listDictions.append(argu)
426	>	job_ToSave['arguments']= str(job+1)
427		job_ToSave['dlsDestination']= self.jobDestination[id]
428		listField.append(job_ToSave)
429	<	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
430	<	+"                     Destination: "+str(self.jobDestination[id])
431	<	common.logger.debug(5,msg)
429	>	from ProdCommon.SiteDB.CmsSiteMapper import CmsSEMap
430	>	cms_se = CmsSEMap()
431	>	msg="Job  %s  Arguments:  %s\n"%(str(job+1),str_argu)
432	>	msg+="\t  Destination: %s "%(str(self.jobDestination[id]))
433	>	SEDestination = [cms_se[dest] for dest in self.jobDestination[id]]
434	>	msg+="\t  CMSDestination: %s "%(str(SEDestination))
435	>	common.logger.log(10-1,msg)
436	>	# write xml
437	>	if len(listDictions):
438	>	if exist==False: self.CreateXML()
439	>	self.addEntry(listDictions)
440	>	# self.zipXMLfile()
441		common._db.updateJob_(listID,listField)
442	<	self.argsList = (len(jobParams[0])+1)
442	>	return
443	>
444	>	# def zipXMLfile(self):
445
446	+	#    import tarfile
447	+	#    try:
448	+	#        tar = tarfile.open(self.tarNameWithPath, "a")
449	+	#        tar.add(self.argsFile, os.path.basename(self.argsFile))
450	+	#        tar.close()
451	+	#    except IOError, exc:
452	+	#        msg = 'Could not add %s to %s \n'%(self.argsFile,self.tarNameWithPath)
453	+	#        msg += str(exc)
454	+	#        raise CrabException(msg)
455	+	#    except tarfile.TarError, exc:
456	+	#        msg = 'Could not add %s to %s \n'%(self.argsFile,self.tarNameWithPath)
457	+	#        msg += str(exc)
458	+	#        raise CrabException(msg)
459	+
460	+	def CreateXML(self):
461	+	"""
462	+	"""
463	+	result = IMProvNode( self.rootArgsFilename )
464	+	outfile = file( self.argsFile, 'w').write(str(result))
465	+	return
466	+
467	+	def addEntry(self, listDictions):
468	+	"""
469	+	_addEntry_
470	+
471	+	add an entry to the xml file
472	+	"""
473	+	from IMProv.IMProvLoader import loadIMProvFile
474	+	## load xml
475	+	improvDoc = loadIMProvFile(self.argsFile)
476	+	entrname= 'Job'
477	+	for dictions in listDictions:
478	+	report = IMProvNode(entrname , None, **dictions)
479	+	improvDoc.addNode(report)
480	+	outfile = file( self.argsFile, 'w').write(str(improvDoc))
481		return
482
483		def numberOfJobs(self):
484	<	return self.total_number_of_jobs
484	>	return self.dict['njobs']
485
486		def getTarBall(self, exe):
487		"""
488		Return the TarBall with lib and exe
489		"""
490	<	self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
490	>	self.tgzNameWithPath = common.work_space.pathForTgz()+self.tgz_name
491		if os.path.exists(self.tgzNameWithPath):
492		return self.tgzNameWithPath
493
#	Line 702 | Line 504 | class Cmssw(JobType):
504
505		## check if working area is release top
506		if swReleaseTop == '' or swArea == swReleaseTop:
507	<	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
507	>	common.logger.debug("swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
508		return
509
510		import tarfile
#	Line 717 | Line 519 | class Cmssw(JobType):
519		## then check if it's private or not
520		if exeWithPath.find(swReleaseTop) == -1:
521		# the exe is private, so we must ship
522	<	common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
522	>	common.logger.debug("Exe "+exeWithPath+" to be tarred")
523		path = swArea+'/'
524		# distinguish case when script is in user project area or given by full path somewhere else
525		if exeWithPath.find(path) >= 0 :
#	Line 731 | Line 533 | class Cmssw(JobType):
533		pass
534
535		## Now get the libraries: only those in local working area
536	+	tar.dereference=True
537		libDir = 'lib'
538		lib = swArea+'/' +libDir
539	<	common.logger.debug(5,"lib "+lib+" to be tarred")
539	>	common.logger.debug("lib "+lib+" to be tarred")
540		if os.path.exists(lib):
541		tar.add(lib,libDir)
542
#	Line 742 | Line 545 | class Cmssw(JobType):
545		module = swArea + '/' + moduleDir
546		if os.path.isdir(module):
547		tar.add(module,moduleDir)
548	+	tar.dereference=False
549
550		## Now check if any data dir(s) is present
551		self.dataExist = False
#	Line 755 | Line 559 | class Cmssw(JobType):
559		todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
560		if name == 'data':
561		self.dataExist=True
562	<	common.logger.debug(5,"data "+entry+" to be tarred")
562	>	common.logger.debug("data "+entry+" to be tarred")
563		tar.add(swArea+"/src/"+entry,"src/"+entry)
564		pass
565		pass
#	Line 764 | Line 568 | class Cmssw(JobType):
568		if not self.pset is None:
569		cfg_file = common.work_space.jobDir()+self.configFilename()
570		tar.add(cfg_file,self.configFilename())
767	–	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
571
572	+	try:
573	+	crab_cfg_file = common.work_space.shareDir()+'/crab.cfg'
574	+	tar.add(crab_cfg_file,'crab.cfg')
575	+	except:
576	+	pass
577
578		## Add ProdCommon dir to tar
579		prodcommonDir = './'
580		prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
581	<	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv']
581	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools', \
582	>	'ProdCommon/Core', 'ProdCommon/MCPayloads', 'IMProv', 'ProdCommon/Storage', \
583	>	'WMCore/__init__.py','WMCore/Algorithms']
584		for file in neededStuff:
585		tar.add(prodcommonPath+file,prodcommonDir+file)
776	–	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
586
587		##### ML stuff
588		ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
589		path=os.environ['CRABDIR'] + '/python/'
590		for file in ML_file_list:
591		tar.add(path+file,file)
783	–	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
592
593		##### Utils
594	<	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py']
594	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py','cmscp.py']
595		for file in Utils_file_list:
596		tar.add(path+file,file)
789	–	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
597
598		##### AdditionalFiles
599	+	tar.dereference=True
600		for file in self.additional_inbox_files:
601		tar.add(file,string.split(file,'/')[-1])
602	<	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
602	>	tar.dereference=False
603	>	common.logger.log(10-1,"Files in "+self.tgzNameWithPath+" : "+str(tar.getnames()))
604
605		tar.close()
606	<	except IOError:
607	<	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
608	<	except tarfile.TarError:
609	<	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
606	>	except IOError, exc:
607	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
608	>	msg += str(exc)
609	>	raise CrabException(msg)
610	>	except tarfile.TarError, exc:
611	>	msg = 'Could not create tar-ball %s \n'%self.tgzNameWithPath
612	>	msg += str(exc)
613	>	raise CrabException(msg)
614
802	–	## check for tarball size
615		tarballinfo = os.stat(self.tgzNameWithPath)
616		if ( tarballinfo.st_size > self.MaxTarBallSize*1024*1024 ) :
617	<	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.')
617	>	msg  = 'Input sandbox size of ' + str(float(tarballinfo.st_size)/1024.0/1024.0) + ' MB is larger than the allowed ' + str(self.MaxTarBallSize) \
618	>	+'MB input sandbox limit \n'
619	>	msg += '      and not supported by the direct GRID submission system.\n'
620	>	msg += '      Please use the CRAB server mode by setting server_name=<NAME> in section [CRAB] of your crab.cfg.\n'
621	>	msg += '      For further infos please see https://twiki.cern.ch/twiki/bin/view/CMS/CrabServer#CRABSERVER_for_Users'
622	>	raise CrabException(msg)
623
624		## create tar-ball with ML stuff
625
#	Line 811 | Line 628 | class Cmssw(JobType):
628		Returns part of a job script which prepares
629		the execution environment for the job 'nj'.
630		"""
631	+	# FUTURE: Drop support for .cfg when possible
632		if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
633		psetName = 'pset.py'
634		else:
#	Line 818 | Line 636 | class Cmssw(JobType):
636		# Prepare JobType-independent part
637		txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
638		txt += 'echo ">>> setup environment"\n'
639	<	txt += 'if [ $middleware == LCG ]; then \n'
639	>	txt += 'if [ $middleware == LCG ] || [ $middleware == CAF ] || [ $middleware == LSF ]; then \n'
640		txt += self.wsSetupCMSLCGEnvironment_()
641		txt += 'elif [ $middleware == OSG ]; then\n'
642		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
#	Line 833 | Line 651 | class Cmssw(JobType):
651		txt += '    cd $WORKING_DIR\n'
652		txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
653		txt += self.wsSetupCMSOSGEnvironment_()
654	+	#Setup SGE Environment
655	+	txt += 'elif [ $middleware == SGE ]; then\n'
656	+	txt += self.wsSetupCMSLCGEnvironment_()
657	+
658	+	txt += 'elif [ $middleware == ARC ]; then\n'
659	+	txt += self.wsSetupCMSLCGEnvironment_()
660	+
661		txt += 'fi\n'
662
663		# Prepare JobType-specific part
#	Line 848 | Line 673 | class Cmssw(JobType):
673		txt += '    func_exit\n'
674		txt += 'fi \n'
675		txt += 'cd '+self.version+'\n'
676	<	txt += 'SOFTWARE_DIR=`pwd`\n'
676	>	txt += 'SOFTWARE_DIR=`pwd`; export SOFTWARE_DIR\n'
677		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
678		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
679		txt += 'if [ $? != 0 ] ; then\n'
#	Line 872 | Line 697 | class Cmssw(JobType):
697		# Prepare job-specific part
698		job = common.job_list[nj]
699		if (self.datasetPath):
700	+	#self.primaryDataset = self.datasetPath.split("/")[1]
701	+	#DataTier = self.datasetPath.split("/")[2]
702		txt += '\n'
703		txt += 'DatasetPath='+self.datasetPath+'\n'
704
705	<	datasetpath_split = self.datasetPath.split("/")
706	<	### FEDE FOR NEW LFN ###
880	<	self.primaryDataset = datasetpath_split[1]
881	<	########################
882	<	txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
883	<	txt += 'DataTier='+datasetpath_split[2]+'\n'
705	>	txt += 'PrimaryDataset='+self.primaryDataset +'\n'
706	>	txt += 'DataTier='+self.dataTier+'\n'
707		txt += 'ApplicationFamily=cmsRun\n'
708
709		else:
710	+	#self.primaryDataset = 'null'
711		txt += 'DatasetPath=MCDataTier\n'
888	–	### FEDE FOR NEW LFN ###
889	–	self.primaryDataset = 'null'
890	–	########################
712		txt += 'PrimaryDataset=null\n'
713		txt += 'DataTier=null\n'
714		txt += 'ApplicationFamily=MCDataTier\n'
#	Line 895 | Line 716 | class Cmssw(JobType):
716		pset = os.path.basename(job.configFilename())
717		txt += '\n'
718		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
898	–	if (self.datasetPath): # standard job
899	–	txt += 'InputFiles=${args[1]}; export InputFiles\n'
900	–	if (self.useParent):
901	–	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
902	–	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
903	–	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
904	–	else:
905	–	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
906	–	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
907	–	txt += 'echo "Inputfiles:<$InputFiles>"\n'
908	–	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
909	–	txt += 'echo "MaxEvents:<$MaxEvents>"\n'
910	–	txt += 'echo "SkipEvents:<$SkipEvents>"\n'
911	–	else:  # pythia like job
912	–	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
913	–	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
914	–	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
915	–	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
916	–	if (self.firstRun):
917	–	txt += 'FirstRun=${args[1]}; export FirstRun\n'
918	–	txt += 'echo "FirstRun: <$FirstRun>"\n'
719
720	<	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
720	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
721	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
722	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
723	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
724
725	<
726	<	if self.pset != None:
924	<	# FUTURE: Can simply for 2_1_x and higher
925	<	txt += '\n'
926	<	if self.debug_wrapper==True:
927	<	txt += 'echo "***** cat ' + psetName + ' *********"\n'
928	<	txt += 'cat ' + psetName + '\n'
929	<	txt += 'echo "****** end ' + psetName + ' ********"\n'
930	<	txt += '\n'
931	<	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
932	<	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
933	<	else:
934	<	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
935	<	txt += 'echo "PSETHASH = $PSETHASH" \n'
725	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
726	>	else:
727		txt += '\n'
728	+	if self.AdditionalArgs: txt += 'export AdditionalArgs=%s\n'%(self.AdditionalArgs)
729	+	if int(self.NumEvents) != 0: txt += 'export MaxEvents=%s\n'%str(self.NumEvents)
730		return txt
731
732		def wsUntarSoftware(self, nj=0):
#	Line 946 | Line 739 | class Cmssw(JobType):
739
740		if os.path.isfile(self.tgzNameWithPath):
741		txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
742	<	txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
743	<	if  self.debug_wrapper:
742	>	txt += 'tar zxvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
743	>	if  self.debug_wrapper==1 :
744		txt += 'ls -Al \n'
745		txt += 'untar_status=$? \n'
746		txt += 'if [ $untar_status -ne 0 ]; then \n'
#	Line 1001 | Line 794 | class Cmssw(JobType):
794		txt += 'fi\n'
795		txt += '\n'
796
797	+	if self.pset != None:
798	+	# FUTURE: Drop support for .cfg when possible
799	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
800	+	psetName = 'pset.py'
801	+	else:
802	+	psetName = 'pset.cfg'
803	+	# FUTURE: Can simply for 2_1_x and higher
804	+	txt += '\n'
805	+	if self.debug_wrapper == 1:
806	+	txt += 'echo "***** cat ' + psetName + ' *********"\n'
807	+	txt += 'cat ' + psetName + '\n'
808	+	txt += 'echo "****** end ' + psetName + ' ********"\n'
809	+	txt += '\n'
810	+	txt += 'echo "***********************" \n'
811	+	txt += 'which edmConfigHash \n'
812	+	txt += 'echo "***********************" \n'
813	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
814	+	txt += 'edmConfigHash ' + psetName + ' \n'
815	+	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
816	+	else:
817	+	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
818	+	txt += 'echo "PSETHASH = $PSETHASH" \n'
819	+	#### FEDE temporary fix for noEdm files #####
820	+	txt += 'if [ -z "$PSETHASH" ]; then \n'
821	+	txt += '   export PSETHASH=null\n'
822	+	txt += 'fi \n'
823	+	#############################################
824	+	txt += '\n'
825		return txt
826
827
#	Line 1012 | Line 833 | class Cmssw(JobType):
833
834		def executableArgs(self):
835		# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
836	<	if self.scriptExe:#CarlosDaniele
837	<	return   self.scriptExe + " $NJob"
836	>	if self.scriptExe:
837	>	return self.scriptExe + " $NJob $AdditionalArgs"
838		else:
839		ex_args = ""
840	<	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
841	<	# Framework job report
1021	<	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1022	<	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1023	<	# Type of config file
840	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
841	>	# Type of config file depends on CMSSW version
842		if self.CMSSW_major >= 2 :
843		ex_args += " -p pset.py"
844		else:
#	Line 1034 | Line 852 | class Cmssw(JobType):
852		inp_box = []
853		if os.path.isfile(self.tgzNameWithPath):
854		inp_box.append(self.tgzNameWithPath)
855	<	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
856	<	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
855	>	if os.path.isfile(self.argsFile):
856	>	inp_box.append(self.argsFile)
857	>	inp_box.append(common.work_space.jobDir() + self.scriptName)
858		return inp_box
859
860		def outputSandbox(self, nj):
#	Line 1059 | Line 878 | class Cmssw(JobType):
878		txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
879		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
880		txt += 'echo ">>> current directory content:"\n'
881	<	if self.debug_wrapper:
881	>	if self.debug_wrapper==1:
882		txt += 'ls -Al\n'
883		txt += '\n'
884
#	Line 1085 | Line 904 | class Cmssw(JobType):
904		txt += 'fi\n'
905		file_list = []
906		for fileWithSuffix in (self.output_file):
907	<	file_list.append(numberFile(fileWithSuffix, '$NJob'))
907	>	file_list.append(numberFile('$SOFTWARE_DIR/'+fileWithSuffix, '$NJob'))
908
909	<	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
909	>	txt += 'file_list="'+string.join(file_list,',')+'"\n'
910		txt += '\n'
911		txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
912		txt += 'echo ">>> current directory content:"\n'
913	<	if self.debug_wrapper:
913	>	if self.debug_wrapper==1:
914		txt += 'ls -Al\n'
915		txt += '\n'
916		txt += 'cd $RUNTIME_AREA\n'
#	Line 1114 | Line 933 | class Cmssw(JobType):
933
934		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
935		if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
936	<	req += ' && other.GlueCEStateStatus == "Production" '
936	>	## 25-Jun-2009 SL: patch to use Cream enabled WMS
937	>	if ( self.cfg_params.get('GRID.use_cream',None) ):
938	>	req += ' && (other.GlueCEStateStatus == "Production" || other.GlueCEStateStatus == "Special")'
939	>	else:
940	>	req += ' && other.GlueCEStateStatus == "Production" '
941
942		return req
943
#	Line 1122 | Line 945 | class Cmssw(JobType):
945		""" return the config filename """
946		# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
947		if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
948	<	return self.name()+'.py'
948	>	return self.name()+'.py'
949		else:
950	<	return self.name()+'.cfg'
950	>	return self.name()+'.cfg'
951
952		def wsSetupCMSOSGEnvironment_(self):
953		"""
#	Line 1184 | Line 1007 | class Cmssw(JobType):
1007		txt += '    echo "==> setup cms environment ok"\n'
1008		return txt
1009
1010	<	def modifyReport(self, nj):
1010	>	def wsModifyReport(self, nj):
1011		"""
1012		insert the part of the script that modifies the FrameworkJob Report
1013		"""
1014	<	txt = '\n#Written by cms_cmssw::modifyReport\n'
1015	<	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1016	<	if (publish_data == 1):
1017	<	processedDataset = self.cfg_params['USER.publish_data_name']
1018	<	if (self.primaryDataset == 'null'):
1019	<	self.primaryDataset = processedDataset
1020	<	if (common.scheduler.name().upper() == "CAF" or common.scheduler.name().upper() == "LSF"):
1021	<	### FEDE FOR NEW LFN ###
1022	<	LFNBaseName = LFNBase(self.primaryDataset, processedDataset, LocalUser=True)
1023	<	self.user = getUserName(LocalUser=True)
1201	<	########################
1202	<	else :
1203	<	### FEDE FOR NEW LFN ###
1204	<	LFNBaseName = LFNBase(self.primaryDataset, processedDataset)
1205	<	self.user = getUserName()
1206	<	########################
1207	<
1208	<	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1209	<	### FEDE FOR NEW LFN ###
1210	<	#txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1211	<	txt += '    FOR_LFN=%s/${PSETHASH}/\n'%(LFNBaseName)
1212	<	########################
1014	>
1015	>	txt = ''
1016	>	#publish_data = int(self.cfg_params.get('USER.publish_data',0))
1017	>	if (self.copy_data == 1):
1018	>	txt = '\n#Written by cms_cmssw::wsModifyReport\n'
1019	>	#publish_data = int(self.cfg_params.get('USER.publish_data',0))
1020	>
1021	>
1022	>	txt += 'if [ $StageOutExitStatus -eq 0 ]; then\n'
1023	>	txt += '    FOR_LFN=$LFNBaseName\n'
1024		txt += 'else\n'
1025		txt += '    FOR_LFN=/copy_problems/ \n'
1215	–	txt += '    SE=""\n'
1216	–	txt += '    SE_PATH=""\n'
1026		txt += 'fi\n'
1027
1028		txt += 'echo ">>> Modify Job Report:" \n'
1029		txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1221	–	txt += 'ProcessedDataset='+processedDataset+'\n'
1222	–	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1030		txt += 'echo "SE = $SE"\n'
1031		txt += 'echo "SE_PATH = $SE_PATH"\n'
1032		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1033		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1034	<	### FEDE FOR NEW LFN ###
1035	<	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' + self.user + '-$ProcessedDataset-$PSETHASH $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1036	<	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier ' + self.user + '-$ProcessedDataset-$PSETHASH $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1037	<	########################
1034	>
1035	>
1036	>	args = 'fjr $RUNTIME_AREA/crab_fjr_$NJob.xml n_job $NJob for_lfn $FOR_LFN PrimaryDataset $PrimaryDataset  ApplicationFamily $ApplicationFamily ApplicationName $executable cmssw_version $CMSSW_VERSION psethash $PSETHASH se_name $SE se_path $SE_PATH file_list $file_list'
1037	>	if (self.publish_data == 1):
1038	>	#processedDataset = self.cfg_params['USER.publish_data_name']
1039	>	txt += 'ProcessedDataset='+self.processedDataset+'\n'
1040	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1041	>	args += ' UserProcessedDataset $USER-$ProcessedDataset-$PSETHASH'
1042	>
1043	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'"\n'
1044	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py '+str(args)+'\n'
1045		txt += 'modifyReport_result=$?\n'
1046		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1047		txt += '    modifyReport_result=70500\n'
#	Line 1248 | Line 1062 | class Cmssw(JobType):
1062		txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1063		txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1064		txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1065	<	if self.debug_wrapper :
1065	>	if self.debug_wrapper==1 :
1066		txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1067		txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1068		txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
#	Line 1264 | Line 1078 | class Cmssw(JobType):
1078		txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1079		txt += '    fi\n'
1080		#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1081	<
1082	<	if (self.datasetPath and not (self.dataset_pu or self.useParent) :
1081	>	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1082	>	txt += '        echo ">>> Executable succeded  $executable_exit_status"\n'
1083	>	## This cannot more work given the changes on the Job argumentsJob
1084	>	"""
1085	>	if (self.datasetPath and not (self.dataset_pu or self.useParent==1)) :
1086		# VERIFY PROCESSED DATA
1087	<	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1088	<	txt += '      echo ">>> Verify list of processed files:"\n'
1089	<	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1090	<	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1091	<	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1092	<	txt += '      mv tmp.txt input-files.txt\n'
1093	<	txt += '      echo "cat input-files.txt"\n'
1094	<	txt += '      echo "----------------------"\n'
1095	<	txt += '      cat input-files.txt\n'
1096	<	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1097	<	txt += '      mv tmp.txt processed-files.txt\n'
1098	<	txt += '      echo "----------------------"\n'
1099	<	txt += '      echo "cat processed-files.txt"\n'
1100	<	txt += '      echo "----------------------"\n'
1101	<	txt += '      cat processed-files.txt\n'
1102	<	txt += '      echo "----------------------"\n'
1103	<	txt += '      diff -q input-files.txt processed-files.txt\n'
1104	<	txt += '      fileverify_status=$?\n'
1105	<	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1106	<	txt += '         executable_exit_status=30001\n'
1107	<	txt += '         echo "ERROR ==> not all input files processed"\n'
1108	<	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1109	<	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1110	<	txt += '      fi\n'
1111	<	txt += '    fi\n'
1295	<	txt += '\n'
1087	>	txt += '        echo ">>> Verify list of processed files:"\n'
1088	>	txt += '        echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1089	>	txt += '        python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1090	>	txt += '        cat input-files.txt  | sort | uniq > tmp.txt\n'
1091	>	txt += '        mv tmp.txt input-files.txt\n'
1092	>	txt += '        echo "cat input-files.txt"\n'
1093	>	txt += '        echo "----------------------"\n'
1094	>	txt += '        cat input-files.txt\n'
1095	>	txt += '        cat processed-files.txt | sort | uniq > tmp.txt\n'
1096	>	txt += '        mv tmp.txt processed-files.txt\n'
1097	>	txt += '        echo "----------------------"\n'
1098	>	txt += '        echo "cat processed-files.txt"\n'
1099	>	txt += '        echo "----------------------"\n'
1100	>	txt += '        cat processed-files.txt\n'
1101	>	txt += '        echo "----------------------"\n'
1102	>	txt += '        diff -qbB input-files.txt processed-files.txt\n'
1103	>	txt += '        fileverify_status=$?\n'
1104	>	txt += '        if [ $fileverify_status -ne 0 ]; then\n'
1105	>	txt += '            executable_exit_status=30001\n'
1106	>	txt += '            echo "ERROR ==> not all input files processed"\n'
1107	>	txt += '            echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1108	>	txt += '            echo "      ==> diff input-files.txt processed-files.txt"\n'
1109	>	txt += '        fi\n'
1110	>	"""
1111	>	txt += '    fi\n'
1112		txt += 'else\n'
1113		txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1114		txt += 'fi\n'
1115		txt += '\n'
1116	+	txt += 'if [ $executable_exit_status -ne 0 ] && [ $executable_exit_status -ne 50115 ] && [ $executable_exit_status -ne 50117 ] && [ $executable_exit_status -ne 30001 ];then\n'
1117	+	txt += '    echo ">>> Executable failed  $executable_exit_status"\n'
1118	+	txt += '    echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1119	+	txt += '    echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1120	+	txt += '    job_exit_code=$executable_exit_status\n'
1121	+	txt += '    func_exit\n'
1122	+	txt += 'fi\n\n'
1123		txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1124		txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1125		txt += 'job_exit_code=$executable_exit_status\n'
#	Line 1309 | Line 1132 | class Cmssw(JobType):
1132		def getParams(self):
1133		return self._params
1134
1135	<	def uniquelist(self, old):
1313	<	"""
1314	<	remove duplicates from a list
1315	<	"""
1316	<	nd={}
1317	<	for e in old:
1318	<	nd[e]=0
1319	<	return nd.keys()
1320	<
1321	<	def outList(self):
1135	>	def outList(self,list=False):
1136		"""
1137		check the dimension of the output files
1138		"""
#	Line 1327 | Line 1141 | class Cmssw(JobType):
1141		listOutFiles = []
1142		stdout = 'CMSSW_$NJob.stdout'
1143		stderr = 'CMSSW_$NJob.stderr'
1144	+	if len(self.output_file) <= 0:
1145	+	msg ="WARNING: no output files name have been defined!!\n"
1146	+	msg+="\tno output files will be reported back/staged\n"
1147	+	common.logger.info(msg)
1148		if (self.return_data == 1):
1149		for file in (self.output_file+self.output_file_sandbox):
1150		listOutFiles.append(numberFile(file, '$NJob'))
#	Line 1340 | Line 1158 | class Cmssw(JobType):
1158		txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1159		txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1160		txt += 'export filesToCheck\n'
1161	+
1162	+	if list : return self.output_file
1163		return txt

Diff Legend

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