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 1.128 by fanzago, Thu Oct 11 16:23:44 2007 UTC vs.
Revision 1.226 by ewv, Thu Jul 3 19:30:07 2008 UTC

#	Line 5 | Line 5 | from crab_util import *
5		from BlackWhiteListParser import BlackWhiteListParser
6		import common
7		import Scram
8	+	from LFNBaseName import *
9
10		import os, string, glob
11
12		class Cmssw(JobType):
13	<	def __init__(self, cfg_params, ncjobs):
13	>	def __init__(self, cfg_params, ncjobs,skip_blocks, isNew):
14		JobType.__init__(self, 'CMSSW')
15		common.logger.debug(3,'CMSSW::__init__')
16	+	self.skip_blocks = skip_blocks
17	+
18	+	self.argsList = []
19
20		self._params = {}
21		self.cfg_params = cfg_params
18	–
22		# init BlackWhiteListParser
23		self.blackWhiteListParser = BlackWhiteListParser(cfg_params)
24
25	<	try:
23	<	self.MaxTarBallSize = float(self.cfg_params['EDG.maxtarballsize'])
24	<	except KeyError:
25	<	self.MaxTarBallSize = 9.5
25	>	self.MaxTarBallSize = float(self.cfg_params.get('EDG.maxtarballsize',9.5))
26
27		# number of jobs requested to be created, limit obj splitting
28		self.ncjobs = ncjobs
29
30		log = common.logger
31	<
31	>
32		self.scram = Scram.Scram(cfg_params)
33		self.additional_inbox_files = []
34		self.scriptExe = ''
35		self.executable = ''
36		self.executable_arch = self.scram.getArch()
37		self.tgz_name = 'default.tgz'
38	–	self.additional_tgz_name = 'additional.tgz'
38		self.scriptName = 'CMSSW.sh'
39	<	self.pset = ''      #scrip use case Da
40	<	self.datasetPath = '' #scrip use case Da
39	>	self.pset = ''
40	>	self.datasetPath = ''
41
42		# set FJR file name
43		self.fjrFileName = 'crab_fjr.xml'
44
45		self.version = self.scram.getSWVersion()
46	<
47	<	#
48	<	# Try to block creation in case of arch/version mismatch
49	<	#
50	<
51	<	a = string.split(self.version, "_")
52	<
53	<	if int(a[1]) == 1 and (int(a[2]) < 5 and self.executable_arch.find('slc4') == 0):
54	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
55	<	raise CrabException(msg)
57	<	if int(a[1]) == 1 and (int(a[2]) >= 5 and self.executable_arch.find('slc3') == 0):
58	<	msg = "Error: CMS does not support %s with %s architecture"%(self.version, self.executable_arch)
46	>	version_array = self.version.split('_')
47	>	self.CMSSW_major = 0
48	>	self.CMSSW_minor = 0
49	>	self.CMSSW_patch = 0
50	>	try:
51	>	self.CMSSW_major = int(version_array[1])
52	>	self.CMSSW_minor = int(version_array[2])
53	>	self.CMSSW_patch = int(version_array[3])
54	>	except:
55	>	msg = "Cannot parse CMSSW version string: " + self.version + " for major and minor release number!"
56		raise CrabException(msg)
60	–
61	–	common.taskDB.setDict('codeVersion',self.version)
62	–	self.setParam_('application', self.version)
57
58		### collect Data cards
59
60	<	## get DBS mode
61	<	try:
68	<	self.use_dbs_1 = int(self.cfg_params['CMSSW.use_dbs_1'])
69	<	except KeyError:
70	<	self.use_dbs_1 = 0
71	<
72	<	try:
73	<	tmp =  cfg_params['CMSSW.datasetpath']
74	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
75	<	if string.lower(tmp)=='none':
76	<	self.datasetPath = None
77	<	self.selectNoInput = 1
78	<	else:
79	<	self.datasetPath = tmp
80	<	self.selectNoInput = 0
81	<	except KeyError:
82	<	msg = "Error: datasetpath not defined "
60	>	if not cfg_params.has_key('CMSSW.datasetpath'):
61	>	msg = "Error: datasetpath not defined "
62		raise CrabException(msg)
63
64	<	# ML monitoring
65	<	# split dataset path style: /PreProdR3Minbias/SIM/GEN-SIM
66	<	if not self.datasetPath:
67	<	self.setParam_('dataset', 'None')
68	<	self.setParam_('owner', 'None')
64	>	### Temporary: added to remove input file control in the case of PU
65	>	self.dataset_pu = cfg_params.get('CMSSW.dataset_pu', None)
66	>
67	>	tmp =  cfg_params['CMSSW.datasetpath']
68	>	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
69	>	if string.lower(tmp)=='none':
70	>	self.datasetPath = None
71	>	self.selectNoInput = 1
72		else:
73	<	try:
74	<	datasetpath_split = self.datasetPath.split("/")
93	<	# standard style
94	<	self.setParam_('datasetFull', self.datasetPath)
95	<	if self.use_dbs_1 == 1 :
96	<	self.setParam_('dataset', datasetpath_split[1])
97	<	self.setParam_('owner', datasetpath_split[-1])
98	<	else:
99	<	self.setParam_('dataset', datasetpath_split[1])
100	<	self.setParam_('owner', datasetpath_split[2])
101	<	except:
102	<	self.setParam_('dataset', self.datasetPath)
103	<	self.setParam_('owner', self.datasetPath)
104	<
105	<	self.setTaskid_()
106	<	self.setParam_('taskId', self.cfg_params['taskId'])
73	>	self.datasetPath = tmp
74	>	self.selectNoInput = 0
75
76		self.dataTiers = []
77	<
77	>	self.debugWrap = ''
78	>	self.debug_wrapper = cfg_params.get('USER.debug_wrapper',False)
79	>	if self.debug_wrapper: self.debugWrap='--debug'
80		## now the application
81	<	try:
82	<	self.executable = cfg_params['CMSSW.executable']
113	<	self.setParam_('exe', self.executable)
114	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
115	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
116	<	log.debug(3,msg)
117	<	except KeyError:
118	<	self.executable = 'cmsRun'
119	<	self.setParam_('exe', self.executable)
120	<	msg = "User executable not defined. Use cmsRun"
121	<	log.debug(3,msg)
122	<	pass
81	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
82	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
83
84	<	try:
125	<	self.pset = cfg_params['CMSSW.pset']
126	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
127	<	if self.pset.lower() != 'none' :
128	<	if (not os.path.exists(self.pset)):
129	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
130	<	else:
131	<	self.pset = None
132	<	except KeyError:
84	>	if not cfg_params.has_key('CMSSW.pset'):
85		raise CrabException("PSet file missing. Cannot run cmsRun ")
86	+	self.pset = cfg_params['CMSSW.pset']
87	+	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
88	+	if self.pset.lower() != 'none' :
89	+	if (not os.path.exists(self.pset)):
90	+	raise CrabException("User defined PSet file "+self.pset+" does not exist")
91	+	else:
92	+	self.pset = None
93
94		# output files
95		## stuff which must be returned always via sandbox
#	Line 140 | Line 99 | class Cmssw(JobType):
99		self.output_file_sandbox.append(self.fjrFileName)
100
101		# other output files to be returned via sandbox or copied to SE
102	<	try:
103	<	self.output_file = []
104	<	tmp = cfg_params['CMSSW.output_file']
105	<	if tmp != '':
106	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
107	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
108	<	for tmp in tmpOutFiles:
109	<	tmp=string.strip(tmp)
151	<	self.output_file.append(tmp)
152	<	pass
153	<	else:
154	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
155	<	pass
156	<	pass
157	<	except KeyError:
158	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
159	<	pass
102	>	outfileflag = False
103	>	self.output_file = []
104	>	tmp = cfg_params.get('CMSSW.output_file',None)
105	>	if tmp :
106	>	self.output_file = [x.strip() for x in tmp.split(',')]
107	>	outfileflag = True #output found
108	>	#else:
109	>	#    log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available\n")
110
111		# script_exe file as additional file in inputSandbox
112	<	try:
113	<	self.scriptExe = cfg_params['USER.script_exe']
114	<	if self.scriptExe != '':
115	<	if not os.path.isfile(self.scriptExe):
116	<	msg ="ERROR. file "+self.scriptExe+" not found"
117	<	raise CrabException(msg)
168	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
169	<	except KeyError:
170	<	self.scriptExe = ''
112	>	self.scriptExe = cfg_params.get('USER.script_exe',None)
113	>	if self.scriptExe :
114	>	if not os.path.isfile(self.scriptExe):
115	>	msg ="ERROR. file "+self.scriptExe+" not found"
116	>	raise CrabException(msg)
117	>	self.additional_inbox_files.append(string.strip(self.scriptExe))
118
172	–	#CarlosDaniele
119		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
120	<	msg ="Error. script_exe  not defined"
121	<	raise CrabException(msg)
120	>	msg ="Error. script_exe  not defined"
121	>	raise CrabException(msg)
122	>
123	>	# use parent files...
124	>	self.useParent = self.cfg_params.get('CMSSW.use_parent',False)
125
126		## additional input files
127	<	try:
127	>	if cfg_params.has_key('USER.additional_input_files'):
128		tmpAddFiles = string.split(cfg_params['USER.additional_input_files'],',')
129		for tmp in tmpAddFiles:
130		tmp = string.strip(tmp)
#	Line 192 | Line 141 | class Cmssw(JobType):
141		if not os.path.exists(file):
142		raise CrabException("Additional input file not found: "+file)
143		pass
195	–	# fname = string.split(file, '/')[-1]
196	–	# storedFile = common.work_space.pathForTgz()+'share/'+fname
197	–	# shutil.copyfile(file, storedFile)
144		self.additional_inbox_files.append(string.strip(file))
145		pass
146		pass
147		common.logger.debug(5,"Additional input files: "+str(self.additional_inbox_files))
148	<	except KeyError:
203	<	pass
204	<
205	<	# files per job
206	<	try:
207	<	if (cfg_params['CMSSW.files_per_jobs']):
208	<	raise CrabException("files_per_jobs no longer supported.  Quitting.")
209	<	except KeyError:
210	<	pass
148	>	pass
149
150		## Events per job
151	<	try:
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	<	except KeyError:
154	>	else:
155		self.eventsPerJob = -1
156		self.selectEventsPerJob = 0
157	<
157	>
158		## number of jobs
159	<	try:
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	<	except KeyError:
162	>	else:
163		self.theNumberOfJobs = 0
164		self.selectNumberOfJobs = 0
165
166	<	try:
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	<	except KeyError:
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: #CarlosDaniele
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)
#	Line 241 | Line 183 | class Cmssw(JobType):
183		msg = 'Must specify  number_of_jobs.'
184		raise CrabException(msg)
185
186	<	## source seed for pythia
187	<	try:
188	<	self.sourceSeed = int(cfg_params['CMSSW.pythia_seed'])
189	<	except KeyError:
190	<	self.sourceSeed = None
191	<	common.logger.debug(5,"No seed given")
192	<
193	<	try:
194	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
195	<	except KeyError:
196	<	self.sourceSeedVtx = None
197	<	common.logger.debug(5,"No vertex seed given")
198	<
199	<	try:
200	<	self.sourceSeedG4 = int(cfg_params['CMSSW.g4_seed'])
201	<	except KeyError:
202	<	self.sourceSeedG4 = None
203	<	common.logger.debug(5,"No g4 sim hits seed given")
204	<
205	<	try:
206	<	self.sourceSeedMix = int(cfg_params['CMSSW.mix_seed'])
207	<	except KeyError:
208	<	self.sourceSeedMix = None
209	<	common.logger.debug(5,"No mix seed given")
210	<
211	<	try:
212	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
213	<	except KeyError:
214	<	self.firstRun = None
215	<	common.logger.debug(5,"No first run given")
216	<	if self.pset != None: #CarlosDaniele
217	<	ver = string.split(self.version,"_")
218	<	if (int(ver[1])>=1 and int(ver[2])>=5):
219	<	import PsetManipulator150 as pp
220	<	else:
221	<	import PsetManipulator as pp
222	<	PsetEdit = pp.PsetManipulator(self.pset) #Daniele Pset
186	>	## New method of dealing with seeds
187	>	self.incrementSeeds = []
188	>	self.preserveSeeds = []
189	>	if cfg_params.has_key('CMSSW.preserve_seeds'):
190	>	tmpList = cfg_params['CMSSW.preserve_seeds'].split(',')
191	>	for tmp in tmpList:
192	>	tmp.strip()
193	>	self.preserveSeeds.append(tmp)
194	>	if cfg_params.has_key('CMSSW.increment_seeds'):
195	>	tmpList = cfg_params['CMSSW.increment_seeds'].split(',')
196	>	for tmp in tmpList:
197	>	tmp.strip()
198	>	self.incrementSeeds.append(tmp)
199	>
200	>	## Old method of dealing with seeds
201	>	## FUTURE: This is for old CMSSW and old CRAB. Can throw exceptions after a couple of CRAB releases and then
202	>	## remove
203	>	self.sourceSeed = cfg_params.get('CMSSW.pythia_seed',None)
204	>	if self.sourceSeed:
205	>	print "pythia_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
206	>	self.incrementSeeds.append('sourceSeed')
207	>	self.incrementSeeds.append('theSource')
208	>
209	>	self.sourceSeedVtx = cfg_params.get('CMSSW.vtx_seed',None)
210	>	if self.sourceSeedVtx:
211	>	print "vtx_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
212	>	self.incrementSeeds.append('VtxSmeared')
213	>
214	>	self.sourceSeedG4 = cfg_params.get('CMSSW.g4_seed',None)
215	>	if self.sourceSeedG4:
216	>	print "g4_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
217	>	self.incrementSeeds.append('g4SimHits')
218	>
219	>	self.sourceSeedMix = cfg_params.get('CMSSW.mix_seed',None)
220	>	if self.sourceSeedMix:
221	>	print "mix_seed is a deprecated parameter. Use preserve_seeds or increment_seeds in the future.\n","Added to increment_seeds."
222	>	self.incrementSeeds.append('mix')
223	>
224	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
225	>
226	>
227	>	# Copy/return
228	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
229	>	self.return_data = int(cfg_params.get('USER.return_data',0))
230
231		#DBSDLS-start
232	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
232	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
233		self.maxEvents=0  # max events available   ( --> check the requested nb. of evts in Creator.py)
234		self.DBSPaths={}  # all dbs paths requested ( --> input to the site local discovery script)
235		self.jobDestination=[]  # Site destination(s) for each job (list of lists)
#	Line 289 | Line 238 | class Cmssw(JobType):
238		blockSites = {}
239		if self.datasetPath:
240		blockSites = self.DataDiscoveryAndLocation(cfg_params)
241	<	#DBSDLS-end
241	>	#DBSDLS-end
242
294	–	self.tgzNameWithPath = self.getTarBall(self.executable)
295	–
243		## Select Splitting
244	<	if self.selectNoInput:
245	<	if self.pset == None: #CarlosDaniele
244	>	if self.selectNoInput:
245	>	if self.pset == None:
246		self.jobSplittingForScript()
247		else:
248		self.jobSplittingNoInput()
249		else:
250		self.jobSplittingByBlocks(blockSites)
251
252	<	# modify Pset
253	<	if self.pset != None: #CarlosDaniele
254	<	try:
255	<	if (self.datasetPath): # standard job
256	<	# allow to processa a fraction of events in a file
257	<	PsetEdit.inputModule("INPUT")
258	<	PsetEdit.maxEvent("INPUTMAXEVENTS")
259	<	PsetEdit.skipEvent("INPUTSKIPEVENTS")
260	<	else:  # pythia like job
252	>	# modify Pset only the first time
253	>	if isNew:
254	>	if self.pset != None:
255	>	import PsetManipulator as pp
256	>	PsetEdit = pp.PsetManipulator(self.pset)
257	>	try:
258	>	# Add FrameworkJobReport to parameter-set, set max events.
259	>	# Reset later for data jobs by writeCFG which does all modifications
260	>	PsetEdit.addCrabFJR(self.fjrFileName) # FUTURE: Job report addition not needed by CMSSW>1.5
261		PsetEdit.maxEvent(self.eventsPerJob)
262	<	if (self.firstRun):
263	<	PsetEdit.pythiaFirstRun("INPUTFIRSTRUN")  #First Run
264	<	if (self.sourceSeed) :
265	<	PsetEdit.pythiaSeed("INPUT")
266	<	if (self.sourceSeedVtx) :
267	<	PsetEdit.vtxSeed("INPUTVTX")
268	<	if (self.sourceSeedG4) :
269	<	PsetEdit.g4Seed("INPUTG4")
270	<	if (self.sourceSeedMix) :
271	<	PsetEdit.mixSeed("INPUTMIX")
272	<	# add FrameworkJobReport to parameter-set
273	<	PsetEdit.addCrabFJR(self.fjrFileName)
274	<	PsetEdit.psetWriter(self.configFilename())
275	<	except:
276	<	msg='Error while manipuliating ParameterSet: exiting...'
277	<	raise CrabException(msg)
262	>	PsetEdit.psetWriter(self.configFilename())
263	>	## If present, add TFileService to output files
264	>	if not int(cfg_params.get('CMSSW.skip_TFileService_output',0)):
265	>	tfsOutput = PsetEdit.getTFileService()
266	>	if tfsOutput:
267	>	if tfsOutput in self.output_file:
268	>	common.logger.debug(5,"Output from TFileService "+tfsOutput+" already in output files")
269	>	else:
270	>	outfileflag = True #output found
271	>	self.output_file.append(tfsOutput)
272	>	common.logger.message("Adding "+tfsOutput+" to output files (from TFileService)")
273	>	pass
274	>	pass
275	>	## If present and requested, add PoolOutputModule to output files
276	>	if int(cfg_params.get('CMSSW.get_edm_output',0)):
277	>	edmOutput = PsetEdit.getPoolOutputModule()
278	>	if edmOutput:
279	>	if edmOutput in self.output_file:
280	>	common.logger.debug(5,"Output from PoolOutputModule "+edmOutput+" already in output files")
281	>	else:
282	>	self.output_file.append(edmOutput)
283	>	common.logger.message("Adding "+edmOutput+" to output files (from PoolOutputModule)")
284	>	pass
285	>	pass
286	>	except CrabException:
287	>	msg='Error while manipulating ParameterSet: exiting...'
288	>	raise CrabException(msg)
289	>	## Prepare inputSandbox TarBall (only the first time)
290	>	self.tgzNameWithPath = self.getTarBall(self.executable)
291
292		def DataDiscoveryAndLocation(self, cfg_params):
293
294		import DataDiscovery
335	–	import DataDiscovery_DBS2
295		import DataLocation
296		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
297
#	Line 341 | Line 300 | class Cmssw(JobType):
300		## Contact the DBS
301		common.logger.message("Contacting Data Discovery Services ...")
302		try:
303	<
345	<	if self.use_dbs_1 == 1 :
346	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params)
347	<	else :
348	<	self.pubdata=DataDiscovery_DBS2.DataDiscovery_DBS2(datasetPath, cfg_params)
303	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
304		self.pubdata.fetchDBSInfo()
305
306		except DataDiscovery.NotExistingDatasetError, ex :
#	Line 357 | Line 312 | class Cmssw(JobType):
312		except DataDiscovery.DataDiscoveryError, ex:
313		msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
314		raise CrabException(msg)
360	–	except DataDiscovery_DBS2.NotExistingDatasetError_DBS2, ex :
361	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
362	–	raise CrabException(msg)
363	–	except DataDiscovery_DBS2.NoDataTierinProvenanceError_DBS2, ex :
364	–	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
365	–	raise CrabException(msg)
366	–	except DataDiscovery_DBS2.DataDiscoveryError_DBS2, ex:
367	–	msg = 'ERROR ***: failed Data Discovery in DBS :  %s'%ex.getErrorMessage()
368	–	raise CrabException(msg)
315
316		self.filesbyblock=self.pubdata.getFiles()
317		self.eventsbyblock=self.pubdata.getEventsPerBlock()
318		self.eventsbyfile=self.pubdata.getEventsPerFile()
319	+	self.parentFiles=self.pubdata.getParent()
320
321		## get max number of events
322	<	self.maxEvents=self.pubdata.getMaxEvents() ##  self.maxEvents used in Creator.py
322	>	self.maxEvents=self.pubdata.getMaxEvents()
323
324		## Contact the DLS and build a list of sites hosting the fileblocks
325		try:
#	Line 381 | Line 328 | class Cmssw(JobType):
328		except DataLocation.DataLocationError , ex:
329		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
330		raise CrabException(msg)
331	<
331	>
332
333		sites = dataloc.getSites()
334		allSites = []
#	Line 395 | Line 342 | class Cmssw(JobType):
342		common.logger.message("Requested dataset: " + datasetPath + " has " + str(self.maxEvents) + " events in " + str(len(self.filesbyblock.keys())) + " blocks.\n")
343
344		return sites
345	<
345	>
346		def jobSplittingByBlocks(self, blockSites):
347		"""
348		Perform job splitting. Jobs run over an integer number of files
#	Line 445 | Line 392 | class Cmssw(JobType):
392		totalNumberOfJobs = 999999999
393		else :
394		totalNumberOfJobs = self.ncjobs
448	–
395
396		blocks = blockSites.keys()
397		blockCount = 0
#	Line 465 | Line 411 | class Cmssw(JobType):
411		blockCount += 1
412		if block not in jobsOfBlock.keys() :
413		jobsOfBlock[block] = []
414	<
414	>
415		if self.eventsbyblock.has_key(block) :
416		numEventsInBlock = self.eventsbyblock[block]
417		common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
418	<
418	>
419		files = self.filesbyblock[block]
420		numFilesInBlock = len(files)
421		if (numFilesInBlock <= 0):
#	Line 477 | Line 423 | class Cmssw(JobType):
423		fileCount = 0
424
425		# ---- New block => New job ---- #
426	<	parString = "\\{"
426	>	parString = ""
427		# counter for number of events in files currently worked on
428		filesEventCount = 0
429		# flag if next while loop should touch new file
430		newFile = 1
431		# job event counter
432		jobSkipEventCount = 0
433	<
433	>
434		# ---- Iterate over the files in the block until we've met the requested ---- #
435		# ---- total # of events or we've gone over all the files in this block  ---- #
436	+	pString=''
437		while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
438		file = files[fileCount]
439	+	if self.useParent:
440	+	parent = self.parentFiles[file]
441	+	for f in parent :
442	+	pString += '\\\"' + f + '\\\"\,'
443	+	common.logger.debug(6, "File "+str(file)+" has the following parents: "+str(parent))
444	+	common.logger.write("File "+str(file)+" has the following parents: "+str(parent))
445		if newFile :
446		try:
447		numEventsInFile = self.eventsbyfile[file]
#	Line 500 | Line 453 | class Cmssw(JobType):
453		newFile = 0
454		except KeyError:
455		common.logger.message("File "+str(file)+" has unknown number of events: skipping")
503	–
456
457	+	eventsPerJobRequested = min(eventsPerJobRequested, eventsRemaining)
458		# if less events in file remain than eventsPerJobRequested
459	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
459	>	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested):
460		# if last file in block
461		if ( fileCount == numFilesInBlock-1 ) :
462		# end job using last file, use remaining events in block
463		# close job and touch new file
464		fullString = parString[:-2]
465	<	fullString += '\\}'
466	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
465	>	if self.useParent:
466	>	fullParentString = pString[:-2]
467	>	list_of_lists.append([fullString,fullParentString,str(-1),str(jobSkipEventCount)])
468	>	else:
469	>	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
470		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
471		self.jobDestination.append(blockSites[block])
472		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 522 | Line 478 | class Cmssw(JobType):
478		eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
479		jobSkipEventCount = 0
480		# reset file
481	<	parString = "\\{"
481	>	pString = ""
482	>	parString = ""
483		filesEventCount = 0
484		newFile = 1
485		fileCount += 1
#	Line 534 | Line 491 | class Cmssw(JobType):
491		elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
492		# close job and touch new file
493		fullString = parString[:-2]
494	<	fullString += '\\}'
495	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
494	>	if self.useParent:
495	>	fullParentString = pString[:-2]
496	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
497	>	else:
498	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
499		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
500		self.jobDestination.append(blockSites[block])
501		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 546 | Line 506 | class Cmssw(JobType):
506		eventsRemaining = eventsRemaining - eventsPerJobRequested
507		jobSkipEventCount = 0
508		# reset file
509	<	parString = "\\{"
509	>	pString = ""
510	>	parString = ""
511		filesEventCount = 0
512		newFile = 1
513		fileCount += 1
514	<
514	>
515		# if more events in file remain than eventsPerJobRequested
516		else :
517		# close job but don't touch new file
518		fullString = parString[:-2]
519	<	fullString += '\\}'
520	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
519	>	if self.useParent:
520	>	fullParentString = pString[:-2]
521	>	list_of_lists.append([fullString,fullParentString,str(eventsPerJobRequested),str(jobSkipEventCount)])
522	>	else:
523	>	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
524		common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
525		self.jobDestination.append(blockSites[block])
526		common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
#	Line 570 | Line 534 | class Cmssw(JobType):
534		jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
535		# remove all but the last file
536		filesEventCount = self.eventsbyfile[file]
537	<	parString = "\\{"
538	<	parString += '\\\"' + file + '\\\"\,'
537	>	if self.useParent:
538	>	for f in parent : pString += '\\\"' + f + '\\\"\,'
539	>	parString = '\\\"' + file + '\\\"\,'
540		pass # END if
541		pass # END while (iterate over files in the block)
542		pass # END while (iterate over blocks in the dataset)
#	Line 579 | Line 544 | class Cmssw(JobType):
544		if (eventsRemaining > 0 and jobCount < totalNumberOfJobs ):
545		common.logger.message("Could not run on all requested events because some blocks not hosted at allowed sites.")
546		common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
547	<
547	>
548		# screen output
549		screenOutput = "List of jobs and available destination sites:\n\n"
550
#	Line 591 | Line 556 | class Cmssw(JobType):
556		for block in blocks:
557		if block in jobsOfBlock.keys() :
558		blockCounter += 1
559	<	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
559	>	screenOutput += "Block %5i: jobs %20s: sites: %s\n" % (blockCounter,spanRanges(jobsOfBlock[block]),
560	>	','.join(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)))
561		if len(self.blackWhiteListParser.checkWhiteList(self.blackWhiteListParser.checkBlackList(blockSites[block],block),block)) == 0:
562	<	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
562	>	noSiteBlock.append( spanRanges(jobsOfBlock[block]) )
563		bloskNoSite.append( blockCounter )
564	<
564	>
565		common.logger.message(screenOutput)
566		if len(noSiteBlock) > 0 and len(bloskNoSite) > 0:
567		msg = 'WARNING: No sites are hosting any part of data for block:\n                '
#	Line 611 | Line 577 | class Cmssw(JobType):
577		for range_jobs in noSiteBlock:
578		msg += str(range_jobs) + virgola
579		msg += '\n               will not be submitted and this block of data can not be analyzed!\n'
580	+	if self.cfg_params.has_key('EDG.se_white_list'):
581	+	msg += 'WARNING: SE White List: '+self.cfg_params['EDG.se_white_list']+'\n'
582	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
583	+	msg += 'Please check if the dataset is available at this site!)\n'
584	+	if self.cfg_params.has_key('EDG.ce_white_list'):
585	+	msg += 'WARNING: CE White List: '+self.cfg_params['EDG.ce_white_list']+'\n'
586	+	msg += '(Hint: By whitelisting you force the job to run at this particular site(s).\n'
587	+	msg += 'Please check if the dataset is available at this site!)\n'
588	+
589		common.logger.message(msg)
590
591		self.list_of_args = list_of_lists
#	Line 621 | Line 596 | class Cmssw(JobType):
596		Perform job splitting based on number of event per job
597		"""
598		common.logger.debug(5,'Splitting per events')
599	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
600	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
601	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
599	>
600	>	if (self.selectEventsPerJob):
601	>	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
602	>	if (self.selectNumberOfJobs):
603	>	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
604	>	if (self.selectTotalNumberEvents):
605	>	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
606
607		if (self.total_number_of_events < 0):
608		msg='Cannot split jobs per Events with "-1" as total number of events'
#	Line 632 | Line 611 | class Cmssw(JobType):
611		if (self.selectEventsPerJob):
612		if (self.selectTotalNumberEvents):
613		self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
614	<	elif(self.selectNumberOfJobs) :
614	>	elif(self.selectNumberOfJobs) :
615		self.total_number_of_jobs =self.theNumberOfJobs
616	<	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
616	>	self.total_number_of_events =int(self.theNumberOfJobs*self.eventsPerJob)
617
618		elif (self.selectNumberOfJobs) :
619		self.total_number_of_jobs = self.theNumberOfJobs
620		self.eventsPerJob = int(self.total_number_of_events/self.total_number_of_jobs)
621	<
621	>
622		common.logger.debug(5,'N jobs  '+str(self.total_number_of_jobs))
623
624		# is there any remainder?
#	Line 655 | Line 634 | class Cmssw(JobType):
634		self.list_of_args = []
635		for i in range(self.total_number_of_jobs):
636		## Since there is no input, any site is good
637	<	# self.jobDestination.append(["Any"])
659	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
637	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
638		args=[]
639		if (self.firstRun):
640	<	## pythia first run
663	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
640	>	## pythia first run
641		args.append(str(self.firstRun)+str(i))
665	–	else:
666	–	## no first run
667	–	#self.list_of_args.append([str(i)])
668	–	args.append(str(i))
669	–	if (self.sourceSeed):
670	–	args.append(str(self.sourceSeed)+str(i))
671	–	if (self.sourceSeedVtx):
672	–	## + vtx random seed
673	–	args.append(str(self.sourceSeedVtx)+str(i))
674	–	if (self.sourceSeedG4):
675	–	## + G4 random seed
676	–	args.append(str(self.sourceSeedG4)+str(i))
677	–	if (self.sourceSeedMix):
678	–	## + Mix random seed
679	–	args.append(str(self.sourceSeedMix)+str(i))
680	–	pass
681	–	pass
642		self.list_of_args.append(args)
683	–	pass
684	–
685	–	# print self.list_of_args
643
644		return
645
646
647	<	def jobSplittingForScript(self):#CarlosDaniele
647	>	def jobSplittingForScript(self):
648		"""
649		Perform job splitting based on number of job
650		"""
#	Line 703 | Line 660 | class Cmssw(JobType):
660		# argument is seed number.$i
661		self.list_of_args = []
662		for i in range(self.total_number_of_jobs):
706	–	## Since there is no input, any site is good
707	–	# self.jobDestination.append(["Any"])
663		self.jobDestination.append([""])
709	–	## no random seed
664		self.list_of_args.append([str(i)])
665		return
666
667	<	def split(self, jobParams):
668	<
715	<	common.jobDB.load()
716	<	#### Fabio
667	>	def split(self, jobParams,firstJobID):
668	>
669		njobs = self.total_number_of_jobs
670		arglist = self.list_of_args
671		# create the empty structure
672		for i in range(njobs):
673		jobParams.append("")
722	–
723	–	for job in range(njobs):
724	–	jobParams[job] = arglist[job]
725	–	# print str(arglist[job])
726	–	# print jobParams[job]
727	–	common.jobDB.setArguments(job, jobParams[job])
728	–	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
729	–	common.jobDB.setDestination(job, self.jobDestination[job])
674
675	<	common.jobDB.save()
675	>	listID=[]
676	>	listField=[]
677	>	for id in range(njobs):
678	>	job = id + int(firstJobID)
679	>	jobParams[id] = arglist[id]
680	>	listID.append(job+1)
681	>	job_ToSave ={}
682	>	concString = ' '
683	>	argu=''
684	>	if len(jobParams[id]):
685	>	argu +=   concString.join(jobParams[id] )
686	>	job_ToSave['arguments']= str(job+1)+' '+argu
687	>	job_ToSave['dlsDestination']= self.jobDestination[id]
688	>	listField.append(job_ToSave)
689	>	msg="Job "+str(job)+" Arguments:   "+str(job+1)+" "+argu+"\n"  \
690	>	+"                     Destination: "+str(self.jobDestination[id])
691	>	common.logger.debug(5,msg)
692	>	common._db.updateJob_(listID,listField)
693	>	self.argsList = (len(jobParams[0])+1)
694	>
695		return
696	<
734	<	def getJobTypeArguments(self, nj, sched):
735	<	result = ''
736	<	for i in common.jobDB.arguments(nj):
737	<	result=result+str(i)+" "
738	<	return result
739	<
696	>
697		def numberOfJobs(self):
741	–	# Fabio
698		return self.total_number_of_jobs
699
700		def getTarBall(self, exe):
701		"""
702		Return the TarBall with lib and exe
703		"""
748	–
749	–	# if it exist, just return it
750	–	#
751	–	# Marco. Let's start to use relative path for Boss XML files
752	–	#
704		self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
705		if os.path.exists(self.tgzNameWithPath):
706		return self.tgzNameWithPath
#	Line 763 | Line 714 | class Cmssw(JobType):
714
715		# First of all declare the user Scram area
716		swArea = self.scram.getSWArea_()
766	–	#print "swArea = ", swArea
767	–	# swVersion = self.scram.getSWVersion()
768	–	# print "swVersion = ", swVersion
717		swReleaseTop = self.scram.getReleaseTop_()
718	<	#print "swReleaseTop = ", swReleaseTop
771	<
718	>
719		## check if working area is release top
720		if swReleaseTop == '' or swArea == swReleaseTop:
721	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
722		return
723
724		import tarfile
#	Line 781 | Line 729 | class Cmssw(JobType):
729		exeWithPath = self.scram.findFile_(executable)
730		if ( not exeWithPath ):
731		raise CrabException('User executable '+executable+' not found')
732	<
732	>
733		## then check if it's private or not
734		if exeWithPath.find(swReleaseTop) == -1:
735		# the exe is private, so we must ship
#	Line 790 | Line 738 | class Cmssw(JobType):
738		# distinguish case when script is in user project area or given by full path somewhere else
739		if exeWithPath.find(path) >= 0 :
740		exe = string.replace(exeWithPath, path,'')
741	<	tar.add(path+exe,os.path.basename(executable))
741	>	tar.add(path+exe,exe)
742		else :
743		tar.add(exeWithPath,os.path.basename(executable))
744		pass
745		else:
746		# the exe is from release, we'll find it on WN
747		pass
748	<
748	>
749		## Now get the libraries: only those in local working area
750		libDir = 'lib'
751		lib = swArea+'/' +libDir
752		common.logger.debug(5,"lib "+lib+" to be tarred")
753		if os.path.exists(lib):
754		tar.add(lib,libDir)
755	<
755	>
756		## Now check if module dir is present
757		moduleDir = 'module'
758		module = swArea + '/' + moduleDir
#	Line 812 | Line 760 | class Cmssw(JobType):
760		tar.add(module,moduleDir)
761
762		## Now check if any data dir(s) is present
763	<	swAreaLen=len(swArea)
764	<	for root, dirs, files in os.walk(swArea):
765	<	if "data" in dirs:
766	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
767	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
768	<
769	<	## Add ProdAgent dir to tar
770	<	paDir = 'ProdAgentApi'
771	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
772	<	if os.path.isdir(pa):
773	<	tar.add(pa,paDir)
774	<
775	<	### FEDE FOR DBS PUBLICATION
776	<	## Add PRODCOMMON dir to tar
777	<	prodcommonDir = 'ProdCommon'
778	<	prodcommonPath = os.environ['CRABDIR'] + '/' + 'ProdCommon'
779	<	if os.path.isdir(prodcommonPath):
780	<	tar.add(prodcommonPath,prodcommonDir)
781	<	#############################
782	<
763	>	self.dataExist = False
764	>	todo_list = [(i, i) for i in  os.listdir(swArea+"/src")]
765	>	while len(todo_list):
766	>	entry, name = todo_list.pop()
767	>	if name.startswith('crab_0_') or  name.startswith('.') or name == 'CVS':
768	>	continue
769	>	if os.path.isdir(swArea+"/src/"+entry):
770	>	entryPath = entry + '/'
771	>	todo_list += [(entryPath + i, i) for i in  os.listdir(swArea+"/src/"+entry)]
772	>	if name == 'data':
773	>	self.dataExist=True
774	>	common.logger.debug(5,"data "+entry+" to be tarred")
775	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
776	>	pass
777	>	pass
778	>
779	>	### CMSSW ParameterSet
780	>	if not self.pset is None:
781	>	cfg_file = common.work_space.jobDir()+self.configFilename()
782	>	tar.add(cfg_file,self.configFilename())
783	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
784	>
785	>
786	>	## Add ProdCommon dir to tar
787	>	prodcommonDir = './'
788	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
789	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv']
790	>	for file in neededStuff:
791	>	tar.add(prodcommonPath+file,prodcommonDir+file)
792	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
793	>
794	>	##### ML stuff
795	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
796	>	path=os.environ['CRABDIR'] + '/python/'
797	>	for file in ML_file_list:
798	>	tar.add(path+file,file)
799		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
800	+
801	+	##### Utils
802	+	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py']
803	+	for file in Utils_file_list:
804	+	tar.add(path+file,file)
805	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
806	+
807	+	##### AdditionalFiles
808	+	for file in self.additional_inbox_files:
809	+	tar.add(file,string.split(file,'/')[-1])
810	+	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
811	+
812		tar.close()
813	<	except :
814	<	raise CrabException('Could not create tar-ball')
813	>	except IOError:
814	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
815	>	except tarfile.TarError:
816	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
817
818		## check for tarball size
819		tarballinfo = os.stat(self.tgzNameWithPath)
#	Line 843 | Line 821 | class Cmssw(JobType):
821		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.')
822
823		## create tar-ball with ML stuff
846	–	self.MLtgzfile =  common.work_space.pathForTgz()+'share/MLfiles.tgz'
847	–	try:
848	–	tar = tarfile.open(self.MLtgzfile, "w:gz")
849	–	path=os.environ['CRABDIR'] + '/python/'
850	–	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
851	–	tar.add(path+file,file)
852	–	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
853	–	tar.close()
854	–	except :
855	–	raise CrabException('Could not create ML files tar-ball')
856	–
857	–	return
858	–
859	–	def additionalInputFileTgz(self):
860	–	"""
861	–	Put all additional files into a tar ball and return its name
862	–	"""
863	–	import tarfile
864	–	tarName=  common.work_space.pathForTgz()+'share/'+self.additional_tgz_name
865	–	tar = tarfile.open(tarName, "w:gz")
866	–	for file in self.additional_inbox_files:
867	–	tar.add(file,string.split(file,'/')[-1])
868	–	common.logger.debug(5,"Files added to "+self.additional_tgz_name+" : "+str(tar.getnames()))
869	–	tar.close()
870	–	return tarName
824
825	<	def wsSetupEnvironment(self, nj):
825	>	def wsSetupEnvironment(self, nj=0):
826		"""
827		Returns part of a job script which prepares
828		the execution environment for the job 'nj'.
829		"""
830	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
831	+	psetName = 'pset.py'
832	+	else:
833	+	psetName = 'pset.cfg'
834		# Prepare JobType-independent part
835	<	txt = ''
836	<
837	<	## OLI_Daniele at this level  middleware already known
881	<
882	<	txt += 'if [ $middleware == LCG ]; then \n'
883	<	txt += '    echo "### First set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
884	<	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
885	<	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
835	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
836	>	txt += 'echo ">>> setup environment"\n'
837	>	txt += 'if [ $middleware == LCG ]; then \n'
838		txt += self.wsSetupCMSLCGEnvironment_()
839		txt += 'elif [ $middleware == OSG ]; then\n'
840		txt += '    WORKING_DIR=`/bin/mktemp  -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
841	<	txt += '    echo "Created working directory: $WORKING_DIR"\n'
842	<	txt += '    if [ ! -d $WORKING_DIR ] ;then\n'
843	<	txt += '        echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
844	<	txt += '    echo "JOB_EXIT_STATUS = 10016"\n'
893	<	txt += '    echo "JobExitCode=10016" | tee -a $RUNTIME_AREA/$repo\n'
894	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
895	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
896	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
897	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
898	<	txt += '        exit 1\n'
841	>	txt += '    if [ ! $? == 0 ] ;then\n'
842	>	txt += '        echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
843	>	txt += '        job_exit_code=10016\n'
844	>	txt += '        func_exit\n'
845		txt += '    fi\n'
846	+	txt += '    echo ">>> Created working directory: $WORKING_DIR"\n'
847		txt += '\n'
848		txt += '    echo "Change to working directory: $WORKING_DIR"\n'
849		txt += '    cd $WORKING_DIR\n'
850	<	txt += self.wsSetupCMSOSGEnvironment_()
851	<	txt += '    echo "### Set SCRAM ARCH to ' + self.executable_arch + ' ###"\n'
905	<	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
850	>	txt += '    echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
851	>	txt += self.wsSetupCMSOSGEnvironment_()
852		txt += 'fi\n'
853
854		# Prepare JobType-specific part
855		scram = self.scram.commandName()
856		txt += '\n\n'
857	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
857	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
858	>	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
859		txt += scram+' project CMSSW '+self.version+'\n'
860		txt += 'status=$?\n'
861		txt += 'if [ $status != 0 ] ; then\n'
862	<	txt += '   echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
863	<	txt += '   echo "JOB_EXIT_STATUS = 10034"\n'
864	<	txt += '   echo "JobExitCode=10034" | tee -a $RUNTIME_AREA/$repo\n'
918	<	txt += '   dumpStatus $RUNTIME_AREA/$repo\n'
919	<	txt += '   rm -f $RUNTIME_AREA/$repo \n'
920	<	txt += '   echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
921	<	txt += '   echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
922	<	## OLI_Daniele
923	<	txt += '    if [ $middleware == OSG ]; then \n'
924	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
925	<	txt += '        cd $RUNTIME_AREA\n'
926	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
927	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
928	<	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'
929	<	txt += '            echo "JOB_EXIT_STATUS = 10018"\n'
930	<	txt += '            echo "JobExitCode=10018" | tee -a $RUNTIME_AREA/$repo\n'
931	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
932	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
933	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
934	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
935	<	txt += '        fi\n'
936	<	txt += '    fi \n'
937	<	txt += '   exit 1 \n'
862	>	txt += '    echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
863	>	txt += '    job_exit_code=10034\n'
864	>	txt += '    func_exit\n'
865		txt += 'fi \n'
939	–	txt += 'echo "CMSSW_VERSION =  '+self.version+'"\n'
866		txt += 'cd '+self.version+'\n'
941	–	########## FEDE FOR DBS2 ######################
867		txt += 'SOFTWARE_DIR=`pwd`\n'
868	<	txt += 'echo SOFTWARE_DIR=$SOFTWARE_DIR \n'
944	<	###############################################
945	<	### needed grep for bug in scramv1 ###
946	<	txt += scram+' runtime -sh\n'
868	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
869		txt += 'eval `'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME`\n'
870	<	txt += 'echo $PATH\n'
871	<
870	>	txt += 'if [ $? != 0 ] ; then\n'
871	>	txt += '    echo "ERROR ==> Problem with the command: "\n'
872	>	txt += '    echo "eval \`'+scram+' runtime -sh | grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
873	>	txt += '    job_exit_code=10034\n'
874	>	txt += '    func_exit\n'
875	>	txt += 'fi \n'
876		# Handle the arguments:
877		txt += "\n"
878		txt += "## number of arguments (first argument always jobnumber)\n"
879		txt += "\n"
880	<	#        txt += "narg=$#\n"
955	<	txt += "if [ $nargs -lt 2 ]\n"
880	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
881		txt += "then\n"
882	<	txt += "    echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
883	<	txt += '    echo "JOB_EXIT_STATUS = 50113"\n'
884	<	txt += '    echo "JobExitCode=50113" | tee -a $RUNTIME_AREA/$repo\n'
960	<	txt += '    dumpStatus $RUNTIME_AREA/$repo\n'
961	<	txt += '    rm -f $RUNTIME_AREA/$repo \n'
962	<	txt += '    echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
963	<	txt += '    echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
964	<	## OLI_Daniele
965	<	txt += '    if [ $middleware == OSG ]; then \n'
966	<	txt += '        echo "Remove working directory: $WORKING_DIR"\n'
967	<	txt += '        cd $RUNTIME_AREA\n'
968	<	txt += '        /bin/rm -rf $WORKING_DIR\n'
969	<	txt += '        if [ -d $WORKING_DIR ] ;then\n'
970	<	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'
971	<	txt += '            echo "JOB_EXIT_STATUS = 50114"\n'
972	<	txt += '            echo "JobExitCode=50114" | tee -a $RUNTIME_AREA/$repo\n'
973	<	txt += '            dumpStatus $RUNTIME_AREA/$repo\n'
974	<	txt += '            rm -f $RUNTIME_AREA/$repo \n'
975	<	txt += '            echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
976	<	txt += '            echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
977	<	txt += '        fi\n'
978	<	txt += '    fi \n'
979	<	txt += "    exit 1\n"
882	>	txt += "    echo 'ERROR ==> Too few arguments' +$nargs+ \n"
883	>	txt += '    job_exit_code=50113\n'
884	>	txt += "    func_exit\n"
885		txt += "fi\n"
886		txt += "\n"
887
888		# Prepare job-specific part
889		job = common.job_list[nj]
890	<	### FEDE FOR DBS OUTPUT PUBLICATION
986	<	if (self.datasetPath):
890	>	if (self.datasetPath):
891		txt += '\n'
892		txt += 'DatasetPath='+self.datasetPath+'\n'
893
894		datasetpath_split = self.datasetPath.split("/")
895	<
895	>
896		txt += 'PrimaryDataset='+datasetpath_split[1]+'\n'
897		txt += 'DataTier='+datasetpath_split[2]+'\n'
994	–	#txt += 'ProcessedDataset='+datasetpath_split[3]+'\n'
898		txt += 'ApplicationFamily=cmsRun\n'
899
900		else:
901		txt += 'DatasetPath=MCDataTier\n'
902		txt += 'PrimaryDataset=null\n'
903		txt += 'DataTier=null\n'
1001	–	#txt += 'ProcessedDataset=null\n'
904		txt += 'ApplicationFamily=MCDataTier\n'
905	<	if self.pset != None: #CarlosDaniele
905	>	if self.pset != None:
906		pset = os.path.basename(job.configFilename())
907		txt += '\n'
908		txt += 'cp  $RUNTIME_AREA/'+pset+' .\n'
909		if (self.datasetPath): # standard job
910	<	#txt += 'InputFiles=$2\n'
911	<	txt += 'InputFiles=${args[1]}\n'
912	<	txt += 'MaxEvents=${args[2]}\n'
913	<	txt += 'SkipEvents=${args[3]}\n'
910	>	txt += 'InputFiles=${args[1]}; export InputFiles\n'
911	>	if (self.useParent):
912	>	txt += 'ParentFiles=${args[2]}; export ParentFiles\n'
913	>	txt += 'MaxEvents=${args[3]}; export MaxEvents\n'
914	>	txt += 'SkipEvents=${args[4]}; export SkipEvents\n'
915	>	else:
916	>	txt += 'MaxEvents=${args[2]}; export MaxEvents\n'
917	>	txt += 'SkipEvents=${args[3]}; export SkipEvents\n'
918		txt += 'echo "Inputfiles:<$InputFiles>"\n'
919	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
919	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
920		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
1015	–	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
921		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
1017	–	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
922		else:  # pythia like job
923	<	seedIndex=1
923	>	txt += 'PreserveSeeds='  + ','.join(self.preserveSeeds)  + '; export PreserveSeeds\n'
924	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
925	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
926	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
927		if (self.firstRun):
928	<	txt += 'FirstRun=${args['+str(seedIndex)+']}\n'
928	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
929		txt += 'echo "FirstRun: <$FirstRun>"\n'
1023	–	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1024	–	seedIndex=seedIndex+1
930
931	<	if (self.sourceSeed):
1027	<	txt += 'Seed=${args['+str(seedIndex)+']}\n'
1028	<	txt += 'sed "s#\<INPUT\>#$Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1029	<	seedIndex=seedIndex+1
1030	<	## the following seeds are not always present
1031	<	if (self.sourceSeedVtx):
1032	<	txt += 'VtxSeed=${args['+str(seedIndex)+']}\n'
1033	<	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
1034	<	txt += 'sed "s#\<INPUTVTX\>#$VtxSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1035	<	seedIndex += 1
1036	<	if (self.sourceSeedG4):
1037	<	txt += 'G4Seed=${args['+str(seedIndex)+']}\n'
1038	<	txt += 'echo "G4Seed: <$G4Seed>"\n'
1039	<	txt += 'sed "s#\<INPUTG4\>#$G4Seed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1040	<	seedIndex += 1
1041	<	if (self.sourceSeedMix):
1042	<	txt += 'mixSeed=${args['+str(seedIndex)+']}\n'
1043	<	txt += 'echo "MixSeed: <$mixSeed>"\n'
1044	<	txt += 'sed "s#\<INPUTMIX\>#$mixSeed#" '+pset+' > tmp && mv -f tmp '+pset+'\n'
1045	<	seedIndex += 1
1046	<	pass
1047	<	pass
1048	<	txt += 'mv -f '+pset+' pset.cfg\n'
931	>	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
932
1050	–	if len(self.additional_inbox_files) > 0:
1051	–	txt += 'if [ -e $RUNTIME_AREA/'+self.additional_tgz_name+' ] ; then\n'
1052	–	txt += '  tar xzvf $RUNTIME_AREA/'+self.additional_tgz_name+'\n'
1053	–	txt += 'fi\n'
1054	–	pass
933
934	<	if self.pset != None: #CarlosDaniele
935	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
1058	<
1059	<	txt += '\n'
1060	<	txt += 'echo "***** cat pset.cfg *********"\n'
1061	<	txt += 'cat pset.cfg\n'
1062	<	txt += 'echo "****** end pset.cfg ********"\n'
934	>	if self.pset != None:
935	>	# FUTURE: Can simply for 2_1_x and higher
936		txt += '\n'
937	<	### FEDE FOR DBS OUTPUT PUBLICATION
938	<	txt += 'PSETHASH=`EdmConfigHash < pset.cfg` \n'
937	>	if self.debug_wrapper==True:
938	>	txt += 'echo "***** cat ' + psetName + ' *********"\n'
939	>	txt += 'cat ' + psetName + '\n'
940	>	txt += 'echo "****** end ' + psetName + ' ********"\n'
941	>	txt += '\n'
942	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
943	>	txt += 'PSETHASH=`edmConfigHash ' + psetName + '` \n'
944	>	else:
945	>	txt += 'PSETHASH=`edmConfigHash < ' + psetName + '` \n'
946		txt += 'echo "PSETHASH = $PSETHASH" \n'
1067	–	##############
947		txt += '\n'
1069	–	# txt += 'echo "***** cat pset1.cfg *********"\n'
1070	–	# txt += 'cat pset1.cfg\n'
1071	–	# txt += 'echo "****** end pset1.cfg ********"\n'
948		return txt
949
950	<	def wsBuildExe(self, nj=0):
950	>	def wsUntarSoftware(self, nj=0):
951		"""
952		Put in the script the commands to build an executable
953		or a library.
954		"""
955
956	<	txt = ""
956	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
957
958		if os.path.isfile(self.tgzNameWithPath):
959	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
959	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
960		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
961	+	if  self.debug_wrapper:
962	+	txt += 'ls -Al \n'
963		txt += 'untar_status=$? \n'
964		txt += 'if [ $untar_status -ne 0 ]; then \n'
965	<	txt += '   echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
966	<	txt += '   echo "JOB_EXIT_STATUS = $untar_status" \n'
967	<	txt += '   echo "JobExitCode=$untar_status" | tee -a $RUNTIME_AREA/$repo\n'
1090	<	txt += '   if [ $middleware == OSG ]; then \n'
1091	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1092	<	txt += '       cd $RUNTIME_AREA\n'
1093	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1094	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1095	<	txt += '           echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
1096	<	txt += '           echo "JOB_EXIT_STATUS = 50999"\n'
1097	<	txt += '           echo "JobExitCode=50999" | tee -a $RUNTIME_AREA/$repo\n'
1098	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1099	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1100	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1101	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1102	<	txt += '       fi\n'
1103	<	txt += '   fi \n'
1104	<	txt += '   \n'
1105	<	txt += '   exit 1 \n'
965	>	txt += '   echo "ERROR ==> Untarring .tgz file failed"\n'
966	>	txt += '   job_exit_code=$untar_status\n'
967	>	txt += '   func_exit\n'
968		txt += 'else \n'
969		txt += '   echo "Successful untar" \n'
970		txt += 'fi \n'
971		txt += '\n'
972	<	txt += 'echo "Include ProdAgentApi and PRODCOMMON in PYTHONPATH"\n'
972	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
973		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
974	<	#### FEDE FOR DBS OUTPUT PUBLICATION
1113	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon\n'
1114	<	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon\n'
1115	<	#txt += '   export PYTHONPATH=ProdAgentApi\n'
974	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
975		txt += 'else\n'
976	<	txt += '   export PYTHONPATH=$SOFTWARE_DIR/ProdAgentApi:$SOFTWARE_DIR/ProdCommon:${PYTHONPATH}\n'
1118	<	#txt += '   export PYTHONPATH=`pwd`/ProdAgentApi:`pwd`/ProdCommon:${PYTHONPATH}\n'
1119	<	#txt += '   export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
976	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
977		txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1121	–	###################
978		txt += 'fi\n'
979		txt += '\n'
980
981		pass
982	<
982	>
983		return txt
984
985	<	def modifySteeringCards(self, nj):
985	>	def wsBuildExe(self, nj=0):
986		"""
987	<	modify the card provided by the user,
988	<	writing a new card into share dir
987	>	Put in the script the commands to build an executable
988	>	or a library.
989		"""
990	<
990	>
991	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
992	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
993	>
994	>	txt += 'rm -r lib/ module/ \n'
995	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
996	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
997	>	if self.dataExist == True:
998	>	txt += 'rm -r src/ \n'
999	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
1000	>	if len(self.additional_inbox_files)>0:
1001	>	for file in self.additional_inbox_files:
1002	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
1003	>	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
1004	>	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
1005	>
1006	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
1007	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
1008	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/\n'
1009	>	txt += 'else\n'
1010	>	txt += '   export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1011	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1012	>	txt += 'fi\n'
1013	>	txt += '\n'
1014	>
1015	>	return txt
1016	>
1017	>
1018		def executableName(self):
1019	<	if self.scriptExe: #CarlosDaniele
1019	>	if self.scriptExe:
1020		return "sh "
1021		else:
1022		return self.executable
1023
1024		def executableArgs(self):
1025	+	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1026		if self.scriptExe:#CarlosDaniele
1027		return   self.scriptExe + " $NJob"
1028		else:
1029	<	# if >= CMSSW_1_5_X, add -e
1030	<	version_array = self.scram.getSWVersion().split('_')
1031	<	major = 0
1032	<	minor = 0
1033	<	try:
1034	<	major = int(version_array[1])
1035	<	minor = int(version_array[2])
1036	<	except:
1153	<	msg = "Cannot parse CMSSW version string: " + "_".join(version_array) + " for major and minor release number!"
1154	<	raise CrabException(msg)
1155	<	if major >= 1 and minor >= 5 :
1156	<	return " -e -p pset.cfg"
1029	>	ex_args = ""
1030	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1031	>	# Framework job report
1032	>	if (self.CMSSW_major >= 1 and self.CMSSW_minor >= 5) or (self.CMSSW_major >= 2):
1033	>	ex_args += " -j $RUNTIME_AREA/crab_fjr_$NJob.xml"
1034	>	# Type of config file
1035	>	if self.CMSSW_major >= 2 :
1036	>	ex_args += " -p pset.py"
1037		else:
1038	<	return " -p pset.cfg"
1038	>	ex_args += " -p pset.cfg"
1039	>	return ex_args
1040
1041		def inputSandbox(self, nj):
1042		"""
1043		Returns a list of filenames to be put in JDL input sandbox.
1044		"""
1045		inp_box = []
1165	–	# # dict added to delete duplicate from input sandbox file list
1166	–	# seen = {}
1167	–	## code
1046		if os.path.isfile(self.tgzNameWithPath):
1047		inp_box.append(self.tgzNameWithPath)
1048	<	if os.path.isfile(self.MLtgzfile):
1049	<	inp_box.append(self.MLtgzfile)
1172	<	## config
1173	<	if not self.pset is None:
1174	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
1175	<	## additional input files
1176	<	tgz = self.additionalInputFileTgz()
1177	<	inp_box.append(tgz)
1048	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1049	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1050		return inp_box
1051
1052		def outputSandbox(self, nj):
#	Line 1185 | Line 1057 | class Cmssw(JobType):
1057
1058		## User Declared output files
1059		for out in (self.output_file+self.output_file_sandbox):
1060	<	n_out = nj + 1
1061	<	out_box.append(self.numberFile_(out,str(n_out)))
1060	>	n_out = nj + 1
1061	>	out_box.append(numberFile(out,str(n_out)))
1062		return out_box
1063
1192	–	def prepareSteeringCards(self):
1193	–	"""
1194	–	Make initial modifications of the user's steering card file.
1195	–	"""
1196	–	return
1064
1065		def wsRenameOutput(self, nj):
1066		"""
1067		Returns part of a job script which renames the produced files.
1068		"""
1069
1070	<	txt = '\n'
1071	<	txt += '# directory content\n'
1072	<	txt += 'ls \n'
1073	<
1074	<	txt += 'output_exit_status=0\n'
1075	<
1076	<	for fileWithSuffix in (self.output_file_sandbox):
1210	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1211	<	txt += '\n'
1212	<	txt += '# check output file\n'
1213	<	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1214	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1215	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1216	<	txt += 'else\n'
1217	<	txt += '    exit_status=60302\n'
1218	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1219	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1220	<	txt += '    if [ $middleware == OSG ]; then \n'
1221	<	txt += '        echo "prepare dummy output file"\n'
1222	<	txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1223	<	txt += '    fi \n'
1224	<	txt += 'fi\n'
1225	<
1070	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1071	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1072	>	txt += 'echo ">>> current directory content:"\n'
1073	>	if self.debug_wrapper:
1074	>	txt += 'ls -Al\n'
1075	>	txt += '\n'
1076	>
1077		for fileWithSuffix in (self.output_file):
1078	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1078	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1079		txt += '\n'
1080		txt += '# check output file\n'
1081		txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1082	<	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA\n'
1083	<	txt += '    cp $RUNTIME_AREA/'+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1082	>	if (self.copy_data == 1):  # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1083	>	txt += '    mv '+fileWithSuffix+' '+output_file_num+'\n'
1084	>	txt += '    ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1085	>	else:
1086	>	txt += '    mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1087	>	txt += '    ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1088		txt += 'else\n'
1089	<	txt += '    exit_status=60302\n'
1090	<	txt += '    echo "ERROR: Problem with output file '+fileWithSuffix+'"\n'
1091	<	txt += '    echo "JOB_EXIT_STATUS = $exit_status"\n'
1237	<	txt += '    output_exit_status=$exit_status\n'
1238	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1089	>	txt += '    job_exit_code=60302\n'
1090	>	txt += '    echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1091	>	if common.scheduler.name().upper() == 'CONDOR_G':
1092		txt += '    if [ $middleware == OSG ]; then \n'
1093		txt += '        echo "prepare dummy output file"\n'
1094		txt += '        echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
#	Line 1243 | Line 1096 | class Cmssw(JobType):
1096		txt += 'fi\n'
1097		file_list = []
1098		for fileWithSuffix in (self.output_file):
1099	<	file_list.append(self.numberFile_(fileWithSuffix, '$NJob'))
1100	<
1099	>	file_list.append(numberFile(fileWithSuffix, '$NJob'))
1100	>
1101		txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1102	+	txt += '\n'
1103	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1104	+	txt += 'echo ">>> current directory content:"\n'
1105	+	if self.debug_wrapper:
1106	+	txt += 'ls -Al\n'
1107	+	txt += '\n'
1108		txt += 'cd $RUNTIME_AREA\n'
1109	+	txt += 'echo ">>> current directory (RUNTIME_AREA):  $RUNTIME_AREA"\n'
1110		return txt
1111
1252	–	def numberFile_(self, file, txt):
1253	–	"""
1254	–	append _'txt' before last extension of a file
1255	–	"""
1256	–	p = string.split(file,".")
1257	–	# take away last extension
1258	–	name = p[0]
1259	–	for x in p[1:-1]:
1260	–	name=name+"."+x
1261	–	# add "_txt"
1262	–	if len(p)>1:
1263	–	ext = p[len(p)-1]
1264	–	result = name + '_' + txt + "." + ext
1265	–	else:
1266	–	result = name + '_' + txt
1267	–
1268	–	return result
1269	–
1112		def getRequirements(self, nj=[]):
1113		"""
1114	<	return job requirements to add to jdl files
1114	>	return job requirements to add to jdl files
1115		"""
1116		req = ''
1117		if self.version:
1118		req='Member("VO-cms-' + \
1119		self.version + \
1120		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1121	<	## SL add requirement for OS version only if SL4
1280	<	#reSL4 = re.compile( r'slc4' )
1281	<	if self.executable_arch: # and reSL4.search(self.executable_arch):
1121	>	if self.executable_arch:
1122		req+=' && Member("VO-cms-' + \
1123		self.executable_arch + \
1124		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1125
1126		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1127	+	if common.scheduler.name() == "glitecoll":
1128	+	req += ' && other.GlueCEStateStatus == "Production" '
1129
1130		return req
1131
1132		def configFilename(self):
1133		""" return the config filename """
1134	<	return self.name()+'.cfg'
1134	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1135	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1136	>	return self.name()+'.py'
1137	>	else:
1138	>	return self.name()+'.cfg'
1139
1294	–	### OLI_DANIELE
1140		def wsSetupCMSOSGEnvironment_(self):
1141		"""
1142		Returns part of a job script which is prepares
1143		the execution environment and which is common for all CMS jobs.
1144		"""
1145	<	txt = '\n'
1146	<	txt += '   echo "### SETUP CMS OSG  ENVIRONMENT ###"\n'
1147	<	txt += '   if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1148	<	txt += '      # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1149	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1150	<	txt += '       source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1306	<	txt += '   elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1145	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1146	>	txt += '    echo ">>> setup CMS OSG environment:"\n'
1147	>	txt += '    echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1148	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1149	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1150	>	txt += '    if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1151		txt += '      # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1152	<	txt += '       export SCRAM_ARCH='+self.executable_arch+'\n'
1153	<	txt += '       source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1154	<	txt += '   else\n'
1155	<	txt += '       echo "SET_CMS_ENV 10020 ==> ERROR $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1156	<	txt += '       echo "JOB_EXIT_STATUS = 10020"\n'
1157	<	txt += '       echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1314	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1315	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1316	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1317	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1318	<	txt += '       exit 1\n'
1319	<	txt += '\n'
1320	<	txt += '       echo "Remove working directory: $WORKING_DIR"\n'
1321	<	txt += '       cd $RUNTIME_AREA\n'
1322	<	txt += '       /bin/rm -rf $WORKING_DIR\n'
1323	<	txt += '       if [ -d $WORKING_DIR ] ;then\n'
1324	<	txt += '           echo "SET_CMS_ENV 10017 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after $GRID3_APP_DIR/cmssoft/cmsset_default.sh and $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1325	<	txt += '           echo "JOB_EXIT_STATUS = 10017"\n'
1326	<	txt += '           echo "JobExitCode=10017" | tee -a $RUNTIME_AREA/$repo\n'
1327	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1328	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1329	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1330	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1331	<	txt += '       fi\n'
1332	<	txt += '\n'
1333	<	txt += '       exit 1\n'
1334	<	txt += '   fi\n'
1152	>	txt += '        source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1153	>	txt += '    else\n'
1154	>	txt += '        echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1155	>	txt += '        job_exit_code=10020\n'
1156	>	txt += '        func_exit\n'
1157	>	txt += '    fi\n'
1158		txt += '\n'
1159	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1160	<	txt += '   echo " END SETUP CMS OSG  ENVIRONMENT "\n'
1159	>	txt += '    echo "==> setup cms environment ok"\n'
1160	>	txt += '    echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1161
1162		return txt
1163	<
1341	<	### OLI_DANIELE
1163	>
1164		def wsSetupCMSLCGEnvironment_(self):
1165		"""
1166		Returns part of a job script which is prepares
1167		the execution environment and which is common for all CMS jobs.
1168		"""
1169	<	txt  = '   \n'
1170	<	txt += '   echo " ### SETUP CMS LCG  ENVIRONMENT ### "\n'
1171	<	txt += '   if [ ! $VO_CMS_SW_DIR ] ;then\n'
1172	<	txt += '       echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1173	<	txt += '       echo "JOB_EXIT_STATUS = 10031" \n'
1174	<	txt += '       echo "JobExitCode=10031" | tee -a $RUNTIME_AREA/$repo\n'
1175	<	txt += '       dumpStatus $RUNTIME_AREA/$repo\n'
1176	<	txt += '       rm -f $RUNTIME_AREA/$repo \n'
1177	<	txt += '       echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1178	<	txt += '       echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1179	<	txt += '       exit 1\n'
1180	<	txt += '   else\n'
1181	<	txt += '       echo "Sourcing environment... "\n'
1182	<	txt += '       if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1183	<	txt += '           echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1184	<	txt += '           echo "JOB_EXIT_STATUS = 10020"\n'
1185	<	txt += '           echo "JobExitCode=10020" | tee -a $RUNTIME_AREA/$repo\n'
1186	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1187	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1188	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1189	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1190	<	txt += '           exit 1\n'
1191	<	txt += '       fi\n'
1192	<	txt += '       echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1193	<	txt += '       source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1194	<	txt += '       result=$?\n'
1195	<	txt += '       if [ $result -ne 0 ]; then\n'
1374	<	txt += '           echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1375	<	txt += '           echo "JOB_EXIT_STATUS = 10032"\n'
1376	<	txt += '           echo "JobExitCode=10032" | tee -a $RUNTIME_AREA/$repo\n'
1377	<	txt += '           dumpStatus $RUNTIME_AREA/$repo\n'
1378	<	txt += '           rm -f $RUNTIME_AREA/$repo \n'
1379	<	txt += '           echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1380	<	txt += '           echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1381	<	txt += '           exit 1\n'
1382	<	txt += '       fi\n'
1383	<	txt += '   fi\n'
1384	<	txt += '   \n'
1385	<	txt += '   echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1386	<	txt += '   echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1169	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1170	>	txt += '    echo ">>> setup CMS LCG environment:"\n'
1171	>	txt += '    echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1172	>	txt += '    export SCRAM_ARCH='+self.executable_arch+'\n'
1173	>	txt += '    export BUILD_ARCH='+self.executable_arch+'\n'
1174	>	txt += '    if [ ! $VO_CMS_SW_DIR ] ;then\n'
1175	>	txt += '        echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1176	>	txt += '        job_exit_code=10031\n'
1177	>	txt += '        func_exit\n'
1178	>	txt += '    else\n'
1179	>	txt += '        echo "Sourcing environment... "\n'
1180	>	txt += '        if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1181	>	txt += '            echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1182	>	txt += '            job_exit_code=10020\n'
1183	>	txt += '            func_exit\n'
1184	>	txt += '        fi\n'
1185	>	txt += '        echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1186	>	txt += '        source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1187	>	txt += '        result=$?\n'
1188	>	txt += '        if [ $result -ne 0 ]; then\n'
1189	>	txt += '            echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1190	>	txt += '            job_exit_code=10032\n'
1191	>	txt += '            func_exit\n'
1192	>	txt += '        fi\n'
1193	>	txt += '    fi\n'
1194	>	txt += '    \n'
1195	>	txt += '    echo "==> setup cms environment ok"\n'
1196		return txt
1197
1389	–	### FEDE FOR DBS OUTPUT PUBLICATION
1198		def modifyReport(self, nj):
1199		"""
1200	<	insert the part of the script that modifies the FrameworkJob Report
1200	>	insert the part of the script that modifies the FrameworkJob Report
1201		"""
1202	+	txt = '\n#Written by cms_cmssw::modifyReport\n'
1203	+	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1204	+	if (publish_data == 1):
1205	+	processedDataset = self.cfg_params['USER.publish_data_name']
1206	+	### FEDE  for publication with LSF and CAF schedulers ####
1207	+	print "common.scheduler.name().upper() = ", common.scheduler.name().upper()
1208	+	if (common.scheduler.name().upper() == "CAF" or common.scheduler.name().upper() == "LSF"):
1209	+	print "chiamo LFNBaseName con localUser = true"
1210	+	LFNBaseName = LFNBase(processedDataset, LocalUser=True)
1211	+	else :
1212	+	LFNBaseName = LFNBase(processedDataset)
1213	+	####
1214
1215	<	txt = ''
1216	<	try:
1217	<	publish_data = int(self.cfg_params['USER.publish_data'])
1218	<	except KeyError:
1219	<	publish_data = 0
1220	<	if (publish_data == 1):
1221	<	txt += 'echo "Modify Job Report" \n'
1402	<	#txt += 'chmod a+x $RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1403	<	################ FEDE FOR DBS2 #############################################
1404	<	txt += 'chmod a+x $SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py\n'
1405	<	#############################################################################
1406	<	#try:
1407	<	#    publish_data = int(self.cfg_params['USER.publish_data'])
1408	<	#except KeyError:
1409	<	#    publish_data = 0
1410	<
1411	<	txt += 'if [ -z "$SE" ]; then\n'
1412	<	txt += '    SE="" \n'
1413	<	txt += 'fi \n'
1414	<	txt += 'if [ -z "$SE_PATH" ]; then\n'
1415	<	txt += '    SE_PATH="" \n'
1416	<	txt += 'fi \n'
1417	<	txt += 'echo "SE = $SE"\n'
1418	<	txt += 'echo "SE_PATH = $SE_PATH"\n'
1215	>	txt += 'if [ $copy_exit_status -eq 0 ]; then\n'
1216	>	txt += '    FOR_LFN=%s_${PSETHASH}/\n'%(LFNBaseName)
1217	>	txt += 'else\n'
1218	>	txt += '    FOR_LFN=/copy_problems/ \n'
1219	>	txt += '    SE=""\n'
1220	>	txt += '    SE_PATH=""\n'
1221	>	txt += 'fi\n'
1222
1223	<	#if (publish_data == 1):
1224	<	#processedDataset = self.cfg_params['USER.processed_datasetname']
1422	<	processedDataset = self.cfg_params['USER.publish_data_name']
1223	>	txt += 'echo ">>> Modify Job Report:" \n'
1224	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1225		txt += 'ProcessedDataset='+processedDataset+'\n'
1424	–	#### LFN=/store/user/<user>/processedDataset_PSETHASH
1425	–	txt += 'if [ "$SE_PATH" == "" ]; then\n'
1426	–	#### FEDE: added slash in LFN ##############
1427	–	txt += '    FOR_LFN=/copy_problems/ \n'
1428	–	txt += 'else \n'
1429	–	txt += '    tmp=`echo $SE_PATH | awk -F \'store\' \'{print$2}\'` \n'
1430	–	#####  FEDE TO BE CHANGED, BECAUSE STORE IS HARDCODED!!!! ########
1431	–	txt += '    FOR_LFN=/store$tmp \n'
1432	–	txt += 'fi \n'
1226		txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1227	+	txt += 'echo "SE = $SE"\n'
1228	+	txt += 'echo "SE_PATH = $SE_PATH"\n'
1229		txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1230		txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1231	<	#txt += 'echo "$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1232	<	txt += 'echo "$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1438	<	txt += '$SOFTWARE_DIR/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1439	<	#txt += '$RUNTIME_AREA/'+self.version+'/ProdAgentApi/FwkJobRep/ModifyJobReport.py crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1440	<
1231	>	txt += 'echo "$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH"\n'
1232	>	txt += '$RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py $RUNTIME_AREA/crab_fjr_$NJob.xml $NJob $FOR_LFN $PrimaryDataset $DataTier $ProcessedDataset $ApplicationFamily $executable $CMSSW_VERSION $PSETHASH $SE $SE_PATH\n'
1233		txt += 'modifyReport_result=$?\n'
1442	–	txt += 'echo modifyReport_result = $modifyReport_result\n'
1234		txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1235	<	txt += '    exit_status=1\n'
1236	<	txt += '    echo "ERROR: Problem with ModifyJobReport"\n'
1235	>	txt += '    modifyReport_result=70500\n'
1236	>	txt += '    job_exit_code=$modifyReport_result\n'
1237	>	txt += '    echo "ModifyReportResult=$modifyReport_result" | tee -a $RUNTIME_AREA/$repo\n'
1238	>	txt += '    echo "WARNING: Problem with ModifyJobReport"\n'
1239		txt += 'else\n'
1240	<	txt += '    mv NewFrameworkJobReport.xml crab_fjr_$NJob.xml\n'
1240	>	txt += '    mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1241		txt += 'fi\n'
1449	–	else:
1450	–	txt += 'echo "no data publication required"\n'
1451	–	#txt += 'ProcessedDataset=no_data_to_publish \n'
1452	–	#### FEDE: added slash in LFN ##############
1453	–	#txt += 'FOR_LFN=/local/ \n'
1454	–	#txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1455	–	#txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1242		return txt
1243
1244	<	def cleanEnv(self):
1245	<	### OLI_DANIELE
1246	<	txt = ''
1247	<	txt += 'if [ $middleware == OSG ]; then\n'
1248	<	txt += '    cd $RUNTIME_AREA\n'
1249	<	txt += '    echo "Remove working directory: $WORKING_DIR"\n'
1250	<	txt += '    /bin/rm -rf $WORKING_DIR\n'
1251	<	txt += '    if [ -d $WORKING_DIR ] ;then\n'
1252	<	txt += '              echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1253	<	txt += '              echo "JOB_EXIT_STATUS = 60999"\n'
1254	<	txt += '              echo "JobExitCode=60999" | tee -a $RUNTIME_AREA/$repo\n'
1255	<	txt += '              dumpStatus $RUNTIME_AREA/$repo\n'
1256	<	txt += '        rm -f $RUNTIME_AREA/$repo \n'
1257	<	txt += '        echo "MonitorJobID=`echo $MonitorJobID`" | tee -a $RUNTIME_AREA/$repo \n'
1258	<	txt += '        echo "MonitorID=`echo $MonitorID`" | tee -a $RUNTIME_AREA/$repo\n'
1244	>	def wsParseFJR(self):
1245	>	"""
1246	>	Parse the FrameworkJobReport to obtain useful infos
1247	>	"""
1248	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1249	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1250	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1251	>	txt += '    if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1252	>	txt += '        cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1253	>	if self.debug_wrapper :
1254	>	txt += '        echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1255	>	txt += '        executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1256	>	txt += '        if [ $executable_exit_status -eq 50115 ];then\n'
1257	>	txt += '            echo ">>> crab_fjr.xml contents: "\n'
1258	>	txt += '            cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1259	>	txt += '            echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1260	>	txt += '        elif [ $executable_exit_status -eq -999 ];then\n'
1261	>	txt += '            echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1262	>	txt += '        else\n'
1263	>	txt += '            echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1264	>	txt += '        fi\n'
1265	>	txt += '    else\n'
1266	>	txt += '        echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1267		txt += '    fi\n'
1268	+	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1269	+
1270	+	if (self.datasetPath and not self.dataset_pu ):
1271	+	# VERIFY PROCESSED DATA
1272	+	txt += '    if [ $executable_exit_status -eq 0 ];then\n'
1273	+	txt += '      echo ">>> Verify list of processed files:"\n'
1274	+	txt += '      echo $InputFiles |tr -d \'\\\\\' |tr \',\' \'\\n\'|tr -d \'"\' > input-files.txt\n'
1275	+	txt += '      python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1276	+	txt += '      cat input-files.txt  | sort | uniq > tmp.txt\n'
1277	+	txt += '      mv tmp.txt input-files.txt\n'
1278	+	txt += '      echo "cat input-files.txt"\n'
1279	+	txt += '      echo "----------------------"\n'
1280	+	txt += '      cat input-files.txt\n'
1281	+	txt += '      cat processed-files.txt | sort | uniq > tmp.txt\n'
1282	+	txt += '      mv tmp.txt processed-files.txt\n'
1283	+	txt += '      echo "----------------------"\n'
1284	+	txt += '      echo "cat processed-files.txt"\n'
1285	+	txt += '      echo "----------------------"\n'
1286	+	txt += '      cat processed-files.txt\n'
1287	+	txt += '      echo "----------------------"\n'
1288	+	txt += '      diff -q input-files.txt processed-files.txt\n'
1289	+	txt += '      fileverify_status=$?\n'
1290	+	txt += '      if [ $fileverify_status -ne 0 ]; then\n'
1291	+	txt += '         executable_exit_status=30001\n'
1292	+	txt += '         echo "ERROR ==> not all input files processed"\n'
1293	+	txt += '         echo "      ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1294	+	txt += '         echo "      ==> diff input-files.txt processed-files.txt"\n'
1295	+	txt += '      fi\n'
1296	+	txt += '    fi\n'
1297	+	txt += '\n'
1298	+	txt += 'else\n'
1299	+	txt += '    echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1300		txt += 'fi\n'
1301		txt += '\n'
1302	+	txt += 'echo "ExeExitCode=$executable_exit_status" | tee -a $RUNTIME_AREA/$repo\n'
1303	+	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1304	+	txt += 'job_exit_code=$executable_exit_status\n'
1305	+
1306		return txt
1307
1308		def setParam_(self, param, value):
#	Line 1481 | Line 1311 | class Cmssw(JobType):
1311		def getParams(self):
1312		return self._params
1313
1484	–	def setTaskid_(self):
1485	–	self._taskId = self.cfg_params['taskId']
1486	–
1487	–	def getTaskid(self):
1488	–	return self._taskId
1489	–
1314		def uniquelist(self, old):
1315		"""
1316		remove duplicates from a list
#	Line 1496 | Line 1320 | class Cmssw(JobType):
1320		nd[e]=0
1321		return nd.keys()
1322
1323	<
1500	<	def checkOut(self, limit):
1323	>	def outList(self):
1324		"""
1325		check the dimension of the output files
1326		"""
1327	<	txt = 'echo "*****************************************"\n'
1328	<	txt += 'echo "** Starting output sandbox limit check **"\n'
1506	<	txt += 'echo "*****************************************"\n'
1507	<	allOutFiles = ""
1327	>	txt = ''
1328	>	txt += 'echo ">>> list of expected files on output sandbox"\n'
1329		listOutFiles = []
1330	<	for fileOut in (self.output_file+self.output_file_sandbox):
1331	<	if fileOut.find('crab_fjr') == -1:
1332	<	allOutFiles = allOutFiles + " " + self.numberFile_(fileOut, '$NJob')
1333	<	listOutFiles.append(self.numberFile_(fileOut, '$NJob'))
1334	<	txt += 'echo "OUTPUT files: '+str(allOutFiles)+'";\n'
1335	<	txt += 'ls -gGhrta;\n'
1336	<	txt += 'sum=0;\n'
1337	<	txt += 'for file in '+str(allOutFiles)+' ; do\n'
1338	<	txt += '    if [ -e $file ]; then\n'
1339	<	txt += '        tt=`ls -gGrta $file | awk \'{ print $3 }\'`\n'
1340	<	txt += '        sum=`expr $sum + $tt`\n'
1341	<	txt += '    else\n'
1342	<	txt += '        echo "WARNING: output file $file not found!"\n'
1343	<	txt += '    fi\n'
1344	<	txt += 'done\n'
1524	<	txt += 'echo "Total Output dimension: $sum";\n'
1525	<	txt += 'limit='+str(limit)+';\n'
1526	<	txt += 'echo "OUTPUT FILES LIMIT SET TO: $limit";\n'
1527	<	txt += 'if [ $limit -lt $sum ]; then\n'
1528	<	txt += '    echo "WARNING: output files have to big size - something will be lost;"\n'
1529	<	txt += '    echo "         checking the output file sizes..."\n'
1530	<	"""
1531	<	txt += '    dim=0;\n'
1532	<	txt += '    exclude=0;\n'
1533	<	txt += '    for files in '+str(allOutFiles)+' ; do\n'
1534	<	txt += '        sumTemp=0;\n'
1535	<	txt += '        for file2 in '+str(allOutFiles)+' ; do\n'
1536	<	txt += '            if [ $file != $file2 ]; then\n'
1537	<	txt += '                tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1538	<	txt += '                sumTemp=`expr $sumTemp + $tt`;\n'
1539	<	txt += '            fi\n'
1540	<	txt += '        done\n'
1541	<	txt += '        if [ $sumTemp -lt $limit ]; then\n'
1542	<	txt += '            if [ $dim -lt $sumTemp ]; then\n'
1543	<	txt += '                dim=$sumTemp;\n'
1544	<	txt += '                exclude=$file;\n'
1545	<	txt += '            fi\n'
1546	<	txt += '        fi\n'
1547	<	txt += '    done\n'
1548	<	txt += '    echo "Dimension calculated: $dim"; echo "File to exclude: $exclude";\n'
1549	<	"""
1550	<	txt += '    tot=0;\n'
1551	<	txt += '    for file2 in '+str(allOutFiles)+' ; do\n'
1552	<	txt += '        tt=`ls -gGrta $file2 | awk \'{ print $3 }\';`\n'
1553	<	txt += '        tot=`expr $tot + $tt`;\n'
1554	<	txt += '        if [ $limit -lt $tot ]; then\n'
1555	<	txt += '            tot=`expr $tot - $tt`;\n'
1556	<	txt += '            fileLast=$file;\n'
1557	<	txt += '            break;\n'
1558	<	txt += '        fi\n'
1559	<	txt += '    done\n'
1560	<	txt += '    echo "Dimension calculated: $tot"; echo "First file to exclude: $file";\n'
1561	<	txt += '    flag=0;\n'
1562	<	txt += '    for filess in '+str(allOutFiles)+' ; do\n'
1563	<	txt += '        if [ $fileLast = $filess ]; then\n'
1564	<	txt += '            flag=1;\n'
1565	<	txt += '        fi\n'
1566	<	txt += '        if [ $flag -eq 1 ]; then\n'
1567	<	txt += '            rm -f $filess;\n'
1568	<	txt += '        fi\n'
1569	<	txt += '    done\n'
1570	<	txt += '    ls -agGhrt;\n'
1571	<	txt += '    echo "WARNING: output files are too big in dimension: can not put in the output_sandbox.";\n'
1572	<	txt += '    echo "JOB_EXIT_STATUS = 70000";\n'
1573	<	txt += '    exit_status=70000;\n'
1574	<	txt += 'else'
1575	<	txt += '    echo "Total Output dimension $sum is fine.";\n'
1576	<	txt += 'fi\n'
1577	<	txt += 'echo "*****************************************"\n'
1578	<	txt += 'echo "*** Ending output sandbox limit check ***"\n'
1579	<	txt += 'echo "*****************************************"\n'
1330	>	stdout = 'CMSSW_$NJob.stdout'
1331	>	stderr = 'CMSSW_$NJob.stderr'
1332	>	if (self.return_data == 1):
1333	>	for file in (self.output_file+self.output_file_sandbox):
1334	>	listOutFiles.append(numberFile(file, '$NJob'))
1335	>	listOutFiles.append(stdout)
1336	>	listOutFiles.append(stderr)
1337	>	else:
1338	>	for file in (self.output_file_sandbox):
1339	>	listOutFiles.append(numberFile(file, '$NJob'))
1340	>	listOutFiles.append(stdout)
1341	>	listOutFiles.append(stderr)
1342	>	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1343	>	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1344	>	txt += 'export filesToCheck\n'
1345		return txt

Diff Legend

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