[ViewVC] Diff of: cvsroot/COMP/CRAB/python/cms

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.62 by slacapra, Wed Jan 17 13:37:25 2007 UTC vs.
Revision 1.231 by spiga, Mon Aug 25 16:13:05 2008 UTC

#	Line 2 \| Line 2 \| from JobType import JobType
2		from crab_logger import Logger
3		from crab_exceptions import *
4		from crab_util import *
5	<	import math
5	>	from BlackWhiteListParser import SEBlackWhiteListParser
6		import common
7	–	import PsetManipulator
8	–
9	–	import DBSInfo
10	–	import DataDiscovery
11	–	import DataLocation
7		import Scram
8	+	from LFNBaseName import *
9
10	<	import glob, os, string, re, shutil
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
21	–	# Marco.
20		self._params = {}
21		self.cfg_params = cfg_params
22	+	# init BlackWhiteListParser
23	+	self.blackWhiteListParser = SEBlackWhiteListParser(cfg_params)
24	+
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)
31	–	scramArea = ''
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.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	<	common.taskDB.setDict('codeVersion',self.version)
47	<	self.setParam_('application', self.version)
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)
57
58		### collect Data cards
59	<	try:
60	<	tmp = cfg_params['CMSSW.datasetpath']
61	<	log.debug(6, "CMSSW::CMSSW(): datasetPath = "+tmp)
51	<	if string.lower(tmp)=='none':
52	<	self.datasetPath = None
53	<	self.selectNoInput = 1
54	<	else:
55	<	self.datasetPath = tmp
56	<	self.selectNoInput = 0
57	<	except KeyError:
58	<	msg = "Error: datasetpath not defined "
59	>
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
63	<	if not self.datasetPath:
64	<	self.setParam_('dataset', 'None')
65	<	self.setParam_('owner', 'None')
66	<	else:
67	<	datasetpath_split = self.datasetPath.split("/")
68	<	self.setParam_('dataset', datasetpath_split[1])
69	<	self.setParam_('owner', datasetpath_split[-1])
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	<	self.setTaskid_()
68	<	self.setParam_('taskId', self.cfg_params['taskId'])
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	>	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']
79	<	self.setParam_('exe', self.executable)
80	<	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
81	<	msg = "Default executable cmsRun overridden. Switch to " + self.executable
82	<	log.debug(3,msg)
83	<	except KeyError:
84	<	self.executable = 'cmsRun'
85	<	self.setParam_('exe', self.executable)
86	<	msg = "User executable not defined. Use cmsRun"
87	<	log.debug(3,msg)
88	<	pass
81	>	self.executable = cfg_params.get('CMSSW.executable','cmsRun')
82	>	log.debug(6, "CMSSW::CMSSW(): executable = "+self.executable)
83
84	<	try:
91	<	self.pset = cfg_params['CMSSW.pset']
92	<	log.debug(6, "Cmssw::Cmssw(): PSet file = "+self.pset)
93	<	if self.pset.lower() != 'none' :
94	<	if (not os.path.exists(self.pset)):
95	<	raise CrabException("User defined PSet file "+self.pset+" does not exist")
96	<	else:
97	<	self.pset = None
98	<	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 106 \| 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	<
105	<
106	<	tmp = cfg_params['CMSSW.output_file']
107	<	if tmp != '':
108	<	tmpOutFiles = string.split(cfg_params['CMSSW.output_file'],',')
109	<	log.debug(7, 'cmssw::cmssw(): output files '+str(tmpOutFiles))
117	<	for tmp in tmpOutFiles:
118	<	tmp=string.strip(tmp)
119	<	self.output_file.append(tmp)
120	<	pass
121	<	else:
122	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
123	<	pass
124	<	pass
125	<	except KeyError:
126	<	log.message("No output file defined: only stdout/err and the CRAB Framework Job Report will be available")
127	<	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 ="WARNING. file "+self.scriptExe+" not found"
117	<	raise CrabException(msg)
118	<	self.additional_inbox_files.append(string.strip(self.scriptExe))
137	<	except KeyError:
138	<	self.scriptExe = ''
139	<	#CarlosDaniele
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	>
119		if self.datasetPath == None and self.pset == None and self.scriptExe == '' :
120	<	msg ="WARNING. 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)
131		dirname = ''
132		if not tmp[0]=="/": dirname = "."
133	<	shutil.copyfile(tmp, common.work_space.shareDir()+tmp)
134	<	files = common.work_space.pathForTgz()+'share/' + tmp
135	<	if not os.path.exists(files):
136	<	raise CrabException("Additional input file not found: "+files)
133	>	files = []
134	>	if string.find(tmp,"*")>-1:
135	>	files = glob.glob(os.path.join(dirname, tmp))
136	>	if len(files)==0:
137	>	raise CrabException("No additional input file found with this pattern: "+tmp)
138	>	else:
139	>	files.append(tmp)
140	>	for file in files:
141	>	if not os.path.exists(file):
142	>	raise CrabException("Additional input file not found: "+file)
143		pass
144	<	self.additional_inbox_files.append(string.strip(files))
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:
161	<	pass
162	<
163	<	# files per job
164	<	try:
165	<	if (cfg_params['CMSSW.files_per_jobs']):
166	<	raise CrabException("files_per_jobs no longer supported. Quitting.")
167	<	except KeyError:
168	<	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 199 \| 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")
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	>	## 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
209	<	try:
210	<	self.sourceSeedVtx = int(cfg_params['CMSSW.vtx_seed'])
211	<	except KeyError:
212	<	self.sourceSeedVtx = None
213	<	common.logger.debug(5,"No vertex seed given")
214	<	try:
215	<	self.firstRun = int(cfg_params['CMSSW.first_run'])
216	<	except KeyError:
217	<	self.firstRun = None
218	<	common.logger.debug(5,"No first run given")
219	<	if self.pset != None: #CarlosDaniele
220	<	self.PsetEdit = PsetManipulator.PsetManipulator(self.pset) #Daniele Pset
209	>	self.firstRun = cfg_params.get('CMSSW.first_run',None)
210	>
211	>	# Copy/return
212	>	self.copy_data = int(cfg_params.get('USER.copy_data',0))
213	>	self.return_data = int(cfg_params.get('USER.return_data',0))
214
215		#DBSDLS-start
216	<	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
216	>	## Initialize the variables that are extracted from DBS/DLS and needed in other places of the code
217		self.maxEvents=0 # max events available ( --> check the requested nb. of evts in Creator.py)
218		self.DBSPaths={} # all dbs paths requested ( --> input to the site local discovery script)
219		self.jobDestination=[] # Site destination(s) for each job (list of lists)
#	Line 229 \| Line 222 \| class Cmssw(JobType):
222		blockSites = {}
223		if self.datasetPath:
224		blockSites = self.DataDiscoveryAndLocation(cfg_params)
225	<	#DBSDLS-end
225	>	#DBSDLS-end
226
234	–	self.tgzNameWithPath = self.getTarBall(self.executable)
235	–
227		## Select Splitting
228	<	if self.selectNoInput:
229	<	if self.pset == None: #CarlosDaniele
228	>	if self.selectNoInput:
229	>	if self.pset == None:
230		self.jobSplittingForScript()
231		else:
232		self.jobSplittingNoInput()
233	<	else:
233	>	else:
234		self.jobSplittingByBlocks(blockSites)
235
236	<	# modify Pset
237	<	if self.pset != None: #CarlosDaniele
238	<	try:
239	<	if (self.datasetPath): # standard job
240	<	# allow to processa a fraction of events in a file
241	<	self.PsetEdit.inputModule("INPUT")
242	<	self.PsetEdit.maxEvent("INPUTMAXEVENTS")
243	<	self.PsetEdit.skipEvent("INPUTSKIPEVENTS")
244	<	else: # pythia like job
245	<	self.PsetEdit.maxEvent(self.eventsPerJob)
246	<	if (self.firstRun):
247	<	self.PsetEdit.pythiaFirstRun("INPUTFIRSTRUN") #First Run
248	<	if (self.sourceSeed) :
249	<	self.PsetEdit.pythiaSeed("INPUT")
250	<	if (self.sourceSeedVtx) :
251	<	self.PsetEdit.pythiaSeedVtx("INPUTVTX")
252	<	# add FrameworkJobReport to parameter-set
253	<	self.PsetEdit.addCrabFJR(self.fjrFileName)
254	<	self.PsetEdit.psetWriter(self.configFilename())
255	<	except:
256	<	msg='Error while manipuliating ParameterSet: exiting...'
257	<	raise CrabException(msg)
236	>	# modify Pset only the first time
237	>	if isNew:
238	>	if self.pset != None:
239	>	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)
273	>	## Prepare inputSandbox TarBall (only the first time)
274	>	self.tgzNameWithPath = self.getTarBall(self.executable)
275
276		def DataDiscoveryAndLocation(self, cfg_params):
277
278	+	import DataDiscovery
279	+	import DataLocation
280		common.logger.debug(10,"CMSSW::DataDiscoveryAndLocation()")
281
282		datasetPath=self.datasetPath
283
274	–	## TODO
275	–	dataTiersList = ""
276	–	dataTiers = dataTiersList.split(',')
277	–
284		## Contact the DBS
285	<	common.logger.message("Contacting DBS...")
285	>	common.logger.message("Contacting Data Discovery Services ...")
286		try:
287	<	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, dataTiers, cfg_params)
287	>	self.pubdata=DataDiscovery.DataDiscovery(datasetPath, cfg_params,self.skip_blocks)
288		self.pubdata.fetchDBSInfo()
289
290		except DataDiscovery.NotExistingDatasetError, ex :
291		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
292		raise CrabException(msg)
287	–
293		except DataDiscovery.NoDataTierinProvenanceError, ex :
294		msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
295		raise CrabException(msg)
296		except DataDiscovery.DataDiscoveryError, ex:
297	<	msg = 'ERROR ***: failed Data Discovery in DBS %s'%ex.getErrorMessage()
297	>	msg = 'ERROR ***: failed Data Discovery in DBS : %s'%ex.getErrorMessage()
298		raise CrabException(msg)
299
295	–	## get list of all required data in the form of dbs paths (dbs path = /dataset/datatier/owner)
296	–	## self.DBSPaths=self.pubdata.getDBSPaths()
297	–	common.logger.message("Required data are :"+self.datasetPath)
298	–
300		self.filesbyblock=self.pubdata.getFiles()
301		self.eventsbyblock=self.pubdata.getEventsPerBlock()
302		self.eventsbyfile=self.pubdata.getEventsPerFile()
303	<	# print str(self.filesbyblock)
303	<	# print 'self.eventsbyfile',len(self.eventsbyfile)
304	<	# print str(self.eventsbyfile)
303	>	self.parentFiles=self.pubdata.getParent()
304
305		## get max number of events
306	<	self.maxEvents=self.pubdata.getMaxEvents() ## self.maxEvents used in Creator.py
308	<	common.logger.message("The number of available events is %s\n"%self.maxEvents)
306	>	self.maxEvents=self.pubdata.getMaxEvents()
307
310	–	common.logger.message("Contacting DLS...")
308		## Contact the DLS and build a list of sites hosting the fileblocks
309		try:
310		dataloc=DataLocation.DataLocation(self.filesbyblock.keys(),cfg_params)
#	Line 315 \| Line 312 \| class Cmssw(JobType):
312		except DataLocation.DataLocationError , ex:
313		msg = 'ERROR ***: failed Data Location in DLS \n %s '%ex.getErrorMessage()
314		raise CrabException(msg)
315	<
315	>
316
317		sites = dataloc.getSites()
318		allSites = []
319		listSites = sites.values()
320	<	for list in listSites:
321	<	for oneSite in list:
320	>	for listSite in listSites:
321	>	for oneSite in listSite:
322		allSites.append(oneSite)
323		allSites = self.uniquelist(allSites)
324
325	<	common.logger.message("Sites ("+str(len(allSites))+") hosting part/all of dataset: "+str(allSites))
326	<	common.logger.debug(6, "List of Sites: "+str(allSites))
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	<
329	>
330		def jobSplittingByBlocks(self, blockSites):
331		"""
332		Perform job splitting. Jobs run over an integer number of files
#	Line 378 \| Line 376 \| class Cmssw(JobType):
376		totalNumberOfJobs = 999999999
377		else :
378		totalNumberOfJobs = self.ncjobs
381	–
379
380		blocks = blockSites.keys()
381		blockCount = 0
#	Line 388 \| Line 385 \| class Cmssw(JobType):
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	<
396	>	if block not in jobsOfBlock.keys() :
397	>	jobsOfBlock[block] = []
398
399	<	numEventsInBlock = self.eventsbyblock[block]
400	<	common.logger.debug(5,'Events in Block File '+str(numEventsInBlock))
401	<
402	<	files = self.filesbyblock[block]
403	<	numFilesInBlock = len(files)
404	<	if (numFilesInBlock <= 0):
405	<	continue
406	<	fileCount = 0
407	<
408	<	# ---- New block => New job ---- #
409	<	parString = "\\{"
410	<	# counter for number of events in files currently worked on
411	<	filesEventCount = 0
412	<	# flag if next while loop should touch new file
413	<	newFile = 1
414	<	# job event counter
415	<	jobSkipEventCount = 0
416	<
417	<	# ---- Iterate over the files in the block until we've met the requested ---- #
418	<	# ---- total # of events or we've gone over all the files in this block ---- #
419	<	while ( (eventsRemaining > 0) and (fileCount < numFilesInBlock) and (jobCount < totalNumberOfJobs) ):
420	<	file = files[fileCount]
421	<	if newFile :
422	<	try:
423	<	numEventsInFile = self.eventsbyfile[file]
424	<	common.logger.debug(6, "File "+str(file)+" has "+str(numEventsInFile)+" events")
425	<	# increase filesEventCount
426	<	filesEventCount += numEventsInFile
427	<	# Add file to current job
428	<	parString += '\\\"' + file + '\\\"\,'
429	<	newFile = 0
430	<	except KeyError:
431	<	common.logger.message("File "+str(file)+" has unknown number of events: skipping")
432	<
433	<
434	<	# if less events in file remain than eventsPerJobRequested
435	<	if ( filesEventCount - jobSkipEventCount < eventsPerJobRequested ) :
436	<	# if last file in block
437	<	if ( fileCount == numFilesInBlock-1 ) :
438	<	# end job using last file, use remaining events in block
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	<	fullString += '\\}'
479	<	list_of_lists.append([fullString,str(-1),str(jobSkipEventCount)])
480	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(filesEventCount - jobSkipEventCount)+" events (last file in block).")
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 + filesEventCount - jobSkipEventCount
490	<	eventsRemaining = eventsRemaining - filesEventCount + jobSkipEventCount
489	>	totalEventCount = totalEventCount + eventsPerJobRequested
490	>	eventsRemaining = eventsRemaining - eventsPerJobRequested
491		jobSkipEventCount = 0
492		# reset file
493	<	parString = "\\{"
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	<	# go to next file
502	<	newFile = 1
503	<	fileCount += 1
504	<	# if events in file equal to eventsPerJobRequested
505	<	elif ( filesEventCount - jobSkipEventCount == eventsPerJobRequested ) :
506	<	# close job and touch new file
507	<	fullString = parString[:-2]
508	<	fullString += '\\}'
509	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
510	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
511	<	self.jobDestination.append(blockSites[block])
512	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
513	<	# reset counter
514	<	jobCount = jobCount + 1
515	<	totalEventCount = totalEventCount + eventsPerJobRequested
516	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
517	<	jobSkipEventCount = 0
518	<	# reset file
519	<	parString = "\\{"
520	<	filesEventCount = 0
521	<	newFile = 1
522	<	fileCount += 1
523	<
524	<	# if more events in file remain than eventsPerJobRequested
525	<	else :
481	<	# close job but don't touch new file
482	<	fullString = parString[:-2]
483	<	fullString += '\\}'
484	<	list_of_lists.append([fullString,str(eventsPerJobRequested),str(jobSkipEventCount)])
485	<	common.logger.debug(3,"Job "+str(jobCount+1)+" can run over "+str(eventsPerJobRequested)+" events.")
486	<	self.jobDestination.append(blockSites[block])
487	<	common.logger.debug(5,"Job "+str(jobCount+1)+" Destination: "+str(self.jobDestination[jobCount]))
488	<	# increase counter
489	<	jobCount = jobCount + 1
490	<	totalEventCount = totalEventCount + eventsPerJobRequested
491	<	eventsRemaining = eventsRemaining - eventsPerJobRequested
492	<	# calculate skip events for last file
493	<	# use filesEventCount (contains several files), jobSkipEventCount and eventsPerJobRequest
494	<	jobSkipEventCount = eventsPerJobRequested - (filesEventCount - jobSkipEventCount - self.eventsbyfile[file])
495	<	# remove all but the last file
496	<	filesEventCount = self.eventsbyfile[file]
497	<	parString = "\\{"
498	<	parString += '\\\"' + file + '\\\"\,'
499	<	pass # END if
500	<	pass # END while (iterate over files in the block)
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("\n"+str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
531	<
530	>	common.logger.message(str(jobCount)+" job(s) can run on "+str(totalEventCount)+" events.\n")
531	>
532	>	# screen output
533	>	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
#	Line 512 \| Line 580 \| class Cmssw(JobType):
580		Perform job splitting based on number of event per job
581		"""
582		common.logger.debug(5,'Splitting per events')
583	<	common.logger.message('Required '+str(self.eventsPerJob)+' events per job ')
584	<	common.logger.message('Required '+str(self.theNumberOfJobs)+' jobs in total ')
585	<	common.logger.message('Required '+str(self.total_number_of_events)+' events in total ')
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	<	self.total_number_of_jobs = int(self.total_number_of_events/self.eventsPerJob)
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)
#	Line 541 \| Line 618 \| class Cmssw(JobType):
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(["Any"])
622	<	self.jobDestination.append([""]) #must be empty to write correctly the xml
546	<	args=''
621	>	self.jobDestination.append([""]) #must be empty to write correctly the xml
622	>	args=[]
623		if (self.firstRun):
624	<	## pythia first run
625	<	#self.list_of_args.append([(str(self.firstRun)+str(i))])
626	<	args=args+(str(self.firstRun)+str(i))
551	<	else:
552	<	## no first run
553	<	#self.list_of_args.append([str(i)])
554	<	args=args+str(i)
555	<	if (self.sourceSeed):
556	<	if (self.sourceSeedVtx):
557	<	## pythia + vtx random seed
558	<	#self.list_of_args.append([
559	<	# str(self.sourceSeed)+str(i),
560	<	# str(self.sourceSeedVtx)+str(i)
561	<	# ])
562	<	args=args+str(',')+str(self.sourceSeed)+str(i)+str(',')+str(self.sourceSeedVtx)+str(i)
563	<	else:
564	<	## only pythia random seed
565	<	#self.list_of_args.append([(str(self.sourceSeed)+str(i))])
566	<	args=args +str(',')+str(self.sourceSeed)+str(i)
567	<	else:
568	<	## no random seed
569	<	if str(args)=='': args=args+(str(self.firstRun)+str(i))
570	<	arguments=args.split(',')
571	<	if len(arguments)==3:self.list_of_args.append([str(arguments[0]),str(arguments[1]),str(arguments[2])])
572	<	elif len(arguments)==2:self.list_of_args.append([str(arguments[0]),str(arguments[1])])
573	<	else :self.list_of_args.append([str(arguments[0])])
574	<
575	<	# print self.list_of_args
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):#CarlosDaniele
631	>	def jobSplittingForScript(self):
632		"""
633		Perform job splitting based on number of job
634		"""
#	Line 593 \| Line 644 \| class Cmssw(JobType):
644		# argument is seed number.$i
645		self.list_of_args = []
646		for i in range(self.total_number_of_jobs):
596	–	## Since there is no input, any site is good
597	–	# self.jobDestination.append(["Any"])
647		self.jobDestination.append([""])
599	–	## no random seed
648		self.list_of_args.append([str(i)])
649		return
650
651	<	def split(self, jobParams):
652	<
605	<	common.jobDB.load()
606	<	#### Fabio
651	>	def split(self, jobParams,firstJobID):
652	>
653		njobs = self.total_number_of_jobs
654		arglist = self.list_of_args
655		# create the empty structure
656		for i in range(njobs):
657		jobParams.append("")
612	–
613	–	for job in range(njobs):
614	–	jobParams[job] = arglist[job]
615	–	# print str(arglist[job])
616	–	# print jobParams[job]
617	–	common.jobDB.setArguments(job, jobParams[job])
618	–	common.logger.debug(5,"Job "+str(job)+" Destination: "+str(self.jobDestination[job]))
619	–	common.jobDB.setDestination(job, self.jobDestination[job])
658
659	<	common.jobDB.save()
659	>	listID=[]
660	>	listField=[]
661	>	for id in range(njobs):
662	>	job = id + int(firstJobID)
663	>	jobParams[id] = arglist[id]
664	>	listID.append(job+1)
665	>	job_ToSave ={}
666	>	concString = ' '
667	>	argu=''
668	>	if len(jobParams[id]):
669	>	argu += concString.join(jobParams[id] )
670	>	job_ToSave['arguments']= str(job+1)+' '+argu
671	>	job_ToSave['dlsDestination']= self.jobDestination[id]
672	>	listField.append(job_ToSave)
673	>	msg="Job "+str(job)+" Arguments: "+str(job+1)+" "+argu+"\n" \
674	>	+" Destination: "+str(self.jobDestination[id])
675	>	common.logger.debug(5,msg)
676	>	common._db.updateJob_(listID,listField)
677	>	self.argsList = (len(jobParams[0])+1)
678	>
679		return
680	<
624	<	def getJobTypeArguments(self, nj, sched):
625	<	result = ''
626	<	for i in common.jobDB.arguments(nj):
627	<	result=result+str(i)+" "
628	<	return result
629	<
680	>
681		def numberOfJobs(self):
631	–	# Fabio
682		return self.total_number_of_jobs
683
684		def getTarBall(self, exe):
685		"""
686		Return the TarBall with lib and exe
687		"""
638	–
639	–	# if it exist, just return it
640	–	#
641	–	# Marco. Let's start to use relative path for Boss XML files
642	–	#
688		self.tgzNameWithPath = common.work_space.pathForTgz()+'share/'+self.tgz_name
689		if os.path.exists(self.tgzNameWithPath):
690		return self.tgzNameWithPath
#	Line 653 \| Line 698 \| class Cmssw(JobType):
698
699		# First of all declare the user Scram area
700		swArea = self.scram.getSWArea_()
656	–	#print "swArea = ", swArea
657	–	swVersion = self.scram.getSWVersion()
658	–	#print "swVersion = ", swVersion
701		swReleaseTop = self.scram.getReleaseTop_()
702	<	#print "swReleaseTop = ", swReleaseTop
661	<
702	>
703		## check if working area is release top
704		if swReleaseTop == '' or swArea == swReleaseTop:
705	+	common.logger.debug(3,"swArea = "+swArea+" swReleaseTop ="+swReleaseTop)
706		return
707
708		import tarfile
#	Line 671 \| Line 713 \| class Cmssw(JobType):
713		exeWithPath = self.scram.findFile_(executable)
714		if ( not exeWithPath ):
715		raise CrabException('User executable '+executable+' not found')
716	<
716	>
717		## then check if it's private or not
718		if exeWithPath.find(swReleaseTop) == -1:
719		# the exe is private, so we must ship
720		common.logger.debug(5,"Exe "+exeWithPath+" to be tarred")
721		path = swArea+'/'
722	<	exe = string.replace(exeWithPath, path,'')
723	<	tar.add(path+exe,exe)
722	>	# distinguish case when script is in user project area or given by full path somewhere else
723	>	if exeWithPath.find(path) >= 0 :
724	>	exe = string.replace(exeWithPath, path,'')
725	>	tar.add(path+exe,exe)
726	>	else :
727	>	tar.add(exeWithPath,os.path.basename(executable))
728		pass
729		else:
730		# the exe is from release, we'll find it on WN
731		pass
732	<
732	>
733		## Now get the libraries: only those in local working area
734		libDir = 'lib'
735		lib = swArea+'/' +libDir
736		common.logger.debug(5,"lib "+lib+" to be tarred")
737		if os.path.exists(lib):
738		tar.add(lib,libDir)
739	<
739	>
740		## Now check if module dir is present
741		moduleDir = 'module'
742		module = swArea + '/' + moduleDir
#	Line 698 \| Line 744 \| class Cmssw(JobType):
744		tar.add(module,moduleDir)
745
746		## Now check if any data dir(s) is present
747	<	swAreaLen=len(swArea)
748	<	for root, dirs, files in os.walk(swArea):
749	<	if "data" in dirs:
750	<	common.logger.debug(5,"data "+root+"/data"+" to be tarred")
751	<	tar.add(root+"/data",root[swAreaLen:]+"/data")
752	<
753	<	## Add ProdAgent dir to tar
754	<	paDir = 'ProdAgentApi'
755	<	pa = os.environ['CRABDIR'] + '/' + 'ProdAgentApi'
756	<	if os.path.isdir(pa):
757	<	tar.add(pa,paDir)
758	<
747	>	self.dataExist = False
748	>	todo_list = [(i, i) for i in os.listdir(swArea+"/src")]
749	>	while len(todo_list):
750	>	entry, name = todo_list.pop()
751	>	if name.startswith('crab_0_') or name.startswith('.') or name == 'CVS':
752	>	continue
753	>	if os.path.isdir(swArea+"/src/"+entry):
754	>	entryPath = entry + '/'
755	>	todo_list += [(entryPath + i, i) for i in os.listdir(swArea+"/src/"+entry)]
756	>	if name == 'data':
757	>	self.dataExist=True
758	>	common.logger.debug(5,"data "+entry+" to be tarred")
759	>	tar.add(swArea+"/src/"+entry,"src/"+entry)
760	>	pass
761	>	pass
762	>
763	>	### CMSSW ParameterSet
764	>	if not self.pset is None:
765	>	cfg_file = common.work_space.jobDir()+self.configFilename()
766	>	tar.add(cfg_file,self.configFilename())
767	>	common.logger.debug(5,"File added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
768	>
769	>
770	>	## Add ProdCommon dir to tar
771	>	prodcommonDir = './'
772	>	prodcommonPath = os.environ['CRABDIR'] + '/' + 'external/'
773	>	neededStuff = ['ProdCommon/__init__.py','ProdCommon/FwkJobRep', 'ProdCommon/CMSConfigTools','ProdCommon/Core','ProdCommon/MCPayloads', 'IMProv']
774	>	for file in neededStuff:
775	>	tar.add(prodcommonPath+file,prodcommonDir+file)
776		common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
777	<	tar.close()
778	<	except :
779	<	raise CrabException('Could not create tar-ball')
717	<
718	<	## create tar-ball with ML stuff
719	<	self.MLtgzfile = common.work_space.pathForTgz()+'share/MLfiles.tgz'
720	<	try:
721	<	tar = tarfile.open(self.MLtgzfile, "w:gz")
777	>
778	>	##### ML stuff
779	>	ML_file_list=['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py']
780		path=os.environ['CRABDIR'] + '/python/'
781	<	for file in ['report.py', 'DashboardAPI.py', 'Logger.py', 'ProcInfo.py', 'apmon.py', 'parseCrabFjr.py']:
781	>	for file in ML_file_list:
782	>	tar.add(path+file,file)
783	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
784	>
785	>	##### Utils
786	>	Utils_file_list=['parseCrabFjr.py','writeCfg.py', 'fillCrabFjr.py']
787	>	for file in Utils_file_list:
788		tar.add(path+file,file)
789	<	common.logger.debug(5,"Files added to "+self.MLtgzfile+" : "+str(tar.getnames()))
789	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
790	>
791	>	##### AdditionalFiles
792	>	for file in self.additional_inbox_files:
793	>	tar.add(file,string.split(file,'/')[-1])
794	>	common.logger.debug(5,"Files added to "+self.tgzNameWithPath+" : "+str(tar.getnames()))
795	>
796		tar.close()
797	<	except :
798	<	raise CrabException('Could not create ML files tar-ball')
799	<
800	<	return
801	<
802	<	def wsSetupEnvironment(self, nj):
797	>	except IOError:
798	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
799	>	except tarfile.TarError:
800	>	raise CrabException('Could not create tar-ball '+self.tgzNameWithPath)
801	>
802	>	## check for tarball size
803	>	tarballinfo = os.stat(self.tgzNameWithPath)
804	>	if ( tarballinfo.st_size > self.MaxTarBallSize10241024 ) :
805	>	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.')
806	>
807	>	## create tar-ball with ML stuff
808	>
809	>	def wsSetupEnvironment(self, nj=0):
810		"""
811		Returns part of a job script which prepares
812		the execution environment for the job 'nj'.
813		"""
814	+	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
815	+	psetName = 'pset.py'
816	+	else:
817	+	psetName = 'pset.cfg'
818		# Prepare JobType-independent part
819	<	txt = ''
820	<
821	<	## OLI_Daniele at this level middleware already known
741	<
742	<	txt += 'if [ $middleware == LCG ]; then \n'
819	>	txt = '\n#Written by cms_cmssw::wsSetupEnvironment\n'
820	>	txt += 'echo ">>> setup environment"\n'
821	>	txt += 'if [ $middleware == LCG ]; then \n'
822		txt += self.wsSetupCMSLCGEnvironment_()
823		txt += 'elif [ $middleware == OSG ]; then\n'
824		txt += ' WORKING_DIR=`/bin/mktemp -d $OSG_WN_TMP/cms_XXXXXXXXXXXX`\n'
825	<	txt += ' echo "Created working directory: $WORKING_DIR"\n'
826	<	txt += ' if [ ! -d $WORKING_DIR ] ;then\n'
827	<	txt += ' echo "SET_CMS_ENV 10016 ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
828	<	txt += ' echo "JOB_EXIT_STATUS = 10016"\n'
750	<	txt += ' echo "JobExitCode=10016" \| tee -a $RUNTIME_AREA/$repo\n'
751	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
752	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
753	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
754	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
755	<	txt += ' exit 1\n'
825	>	txt += ' if [ ! $? == 0 ] ;then\n'
826	>	txt += ' echo "ERROR ==> OSG $WORKING_DIR could not be created on WN `hostname`"\n'
827	>	txt += ' job_exit_code=10016\n'
828	>	txt += ' func_exit\n'
829		txt += ' fi\n'
830	+	txt += ' echo ">>> Created working directory: $WORKING_DIR"\n'
831		txt += '\n'
832		txt += ' echo "Change to working directory: $WORKING_DIR"\n'
833		txt += ' cd $WORKING_DIR\n'
834	<	txt += self.wsSetupCMSOSGEnvironment_()
834	>	txt += ' echo ">>> current directory (WORKING_DIR): $WORKING_DIR"\n'
835	>	txt += self.wsSetupCMSOSGEnvironment_()
836		txt += 'fi\n'
837
838		# Prepare JobType-specific part
839		scram = self.scram.commandName()
840		txt += '\n\n'
841	<	txt += 'echo "### SPECIFIC JOB SETUP ENVIRONMENT ###"\n'
841	>	txt += 'echo ">>> specific cmssw setup environment:"\n'
842	>	txt += 'echo "CMSSW_VERSION = '+self.version+'"\n'
843		txt += scram+' project CMSSW '+self.version+'\n'
844		txt += 'status=$?\n'
845		txt += 'if [ $status != 0 ] ; then\n'
846	<	txt += ' echo "SET_EXE_ENV 10034 ==>ERROR CMSSW '+self.version+' not found on `hostname`" \n'
847	<	txt += ' echo "JOB_EXIT_STATUS = 10034"\n'
848	<	txt += ' echo "JobExitCode=10034" \| tee -a $RUNTIME_AREA/$repo\n'
773	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
774	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
775	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
776	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
777	<	## OLI_Daniele
778	<	txt += ' if [ $middleware == OSG ]; then \n'
779	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
780	<	txt += ' cd $RUNTIME_AREA\n'
781	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
782	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
783	<	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'
784	<	txt += ' echo "JOB_EXIT_STATUS = 10018"\n'
785	<	txt += ' echo "JobExitCode=10018" \| tee -a $RUNTIME_AREA/$repo\n'
786	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
787	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
788	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
789	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
790	<	txt += ' fi\n'
791	<	txt += ' fi \n'
792	<	txt += ' exit 1 \n'
846	>	txt += ' echo "ERROR ==> CMSSW '+self.version+' not found on `hostname`" \n'
847	>	txt += ' job_exit_code=10034\n'
848	>	txt += ' func_exit\n'
849		txt += 'fi \n'
794	–	txt += 'echo "CMSSW_VERSION = '+self.version+'"\n'
850		txt += 'cd '+self.version+'\n'
851	<	### needed grep for bug in scramv1 ###
852	<	txt += scram+' runtime -sh\n'
851	>	txt += 'SOFTWARE_DIR=`pwd`\n'
852	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
853		txt += 'eval `'+scram+' runtime -sh \| grep -v SCRAMRT_LSB_JOBNAME`\n'
854	<	txt += 'echo $PATH\n'
855	<
854	>	txt += 'if [ $? != 0 ] ; then\n'
855	>	txt += ' echo "ERROR ==> Problem with the command: "\n'
856	>	txt += ' echo "eval \`'+scram+' runtime -sh \| grep -v SCRAMRT_LSB_JOBNAME \` at `hostname`"\n'
857	>	txt += ' job_exit_code=10034\n'
858	>	txt += ' func_exit\n'
859	>	txt += 'fi \n'
860		# Handle the arguments:
861		txt += "\n"
862		txt += "## number of arguments (first argument always jobnumber)\n"
863		txt += "\n"
864	<	# txt += "narg=$#\n"
806	<	txt += "if [ $nargs -lt 2 ]\n"
864	>	txt += "if [ $nargs -lt "+str(self.argsList)+" ]\n"
865		txt += "then\n"
866	<	txt += " echo 'SET_EXE_ENV 1 ==> ERROR Too few arguments' +$nargs+ \n"
867	<	txt += ' echo "JOB_EXIT_STATUS = 50113"\n'
868	<	txt += ' echo "JobExitCode=50113" \| tee -a $RUNTIME_AREA/$repo\n'
811	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
812	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
813	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
814	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
815	<	## OLI_Daniele
816	<	txt += ' if [ $middleware == OSG ]; then \n'
817	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
818	<	txt += ' cd $RUNTIME_AREA\n'
819	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
820	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
821	<	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'
822	<	txt += ' echo "JOB_EXIT_STATUS = 50114"\n'
823	<	txt += ' echo "JobExitCode=50114" \| tee -a $RUNTIME_AREA/$repo\n'
824	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
825	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
826	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
827	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
828	<	txt += ' fi\n'
829	<	txt += ' fi \n'
830	<	txt += " exit 1\n"
866	>	txt += " echo 'ERROR ==> Too few arguments' +$nargs+ \n"
867	>	txt += ' job_exit_code=50113\n'
868	>	txt += " func_exit\n"
869		txt += "fi\n"
870		txt += "\n"
871
872		# Prepare job-specific part
873		job = common.job_list[nj]
874	<	if self.pset != None: #CarlosDaniele
874	>	if (self.datasetPath):
875	>	txt += '\n'
876	>	txt += 'DatasetPath='+self.datasetPath+'\n'
877	>
878	>	datasetpath_split = self.datasetPath.split("/")
879	>	### 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'
884	>	txt += 'ApplicationFamily=cmsRun\n'
885	>
886	>	else:
887	>	txt += 'DatasetPath=MCDataTier\n'
888	>	### FEDE FOR NEW LFN ###
889	>	self.primaryDataset = 'null'
890	>	########################
891	>	txt += 'PrimaryDataset=null\n'
892	>	txt += 'DataTier=null\n'
893	>	txt += 'ApplicationFamily=MCDataTier\n'
894	>	if self.pset != None:
895		pset = os.path.basename(job.configFilename())
896		txt += '\n'
897	+	txt += 'cp $RUNTIME_AREA/'+pset+' .\n'
898		if (self.datasetPath): # standard job
899	<	#txt += 'InputFiles=$2\n'
900	<	txt += 'InputFiles=${args[1]}\n'
901	<	txt += 'MaxEvents=${args[2]}\n'
902	<	txt += 'SkipEvents=${args[3]}\n'
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:<$InputFiles>"\n'
908	<	txt += 'sed "s#{\'INPUT\'}#$InputFiles#" $RUNTIME_AREA/'+pset+' > pset_tmp_1.cfg\n'
908	>	if (self.useParent): txt += 'echo "ParentFiles:<$ParentFiles>"\n'
909		txt += 'echo "MaxEvents:<$MaxEvents>"\n'
847	–	txt += 'sed "s#INPUTMAXEVENTS#$MaxEvents#" pset_tmp_1.cfg > pset_tmp_2.cfg\n'
910		txt += 'echo "SkipEvents:<$SkipEvents>"\n'
849	–	txt += 'sed "s#INPUTSKIPEVENTS#$SkipEvents#" pset_tmp_2.cfg > pset.cfg\n'
911		else: # pythia like job
912	<	if (self.sourceSeed):
913	<	txt += 'FirstRun=${args[1]}\n'
912	>	txt += 'PreserveSeeds=' + ','.join(self.preserveSeeds) + '; export PreserveSeeds\n'
913	>	txt += 'IncrementSeeds=' + ','.join(self.incrementSeeds) + '; export IncrementSeeds\n'
914	>	txt += 'echo "PreserveSeeds: <$PreserveSeeds>"\n'
915	>	txt += 'echo "IncrementSeeds:<$IncrementSeeds>"\n'
916	>	if (self.firstRun):
917	>	txt += 'FirstRun=${args[1]}; export FirstRun\n'
918		txt += 'echo "FirstRun: <$FirstRun>"\n'
854	–	txt += 'sed "s#\<INPUTFIRSTRUN\>#$FirstRun#" $RUNTIME_AREA/'+pset+' > tmp_1.cfg\n'
855	–	else:
856	–	txt += '# Copy untouched pset\n'
857	–	txt += 'cp $RUNTIME_AREA/'+pset+' tmp_1.cfg\n'
858	–	if (self.sourceSeed):
859	–	# txt += 'Seed=$2\n'
860	–	txt += 'Seed=${args[2]}\n'
861	–	txt += 'echo "Seed: <$Seed>"\n'
862	–	txt += 'sed "s#\<INPUT\>#$Seed#" tmp_1.cfg > tmp_2.cfg\n'
863	–	if (self.sourceSeedVtx):
864	–	# txt += 'VtxSeed=$3\n'
865	–	txt += 'VtxSeed=${args[3]}\n'
866	–	txt += 'echo "VtxSeed: <$VtxSeed>"\n'
867	–	txt += 'sed "s#INPUTVTX#$VtxSeed#" tmp_2.cfg > pset.cfg\n'
868	–	else:
869	–	txt += 'mv tmp_2.cfg pset.cfg\n'
870	–	else:
871	–	txt += 'mv tmp_1.cfg pset.cfg\n'
872	–	# txt += '# Copy untouched pset\n'
873	–	# txt += 'cp $RUNTIME_AREA/'+pset+' pset.cfg\n'
919
920	+	txt += 'mv -f ' + pset + ' ' + psetName + '\n'
921
922	<	if len(self.additional_inbox_files) > 0:
923	<	for file in self.additional_inbox_files:
924	<	relFile = file.split("/")[-1]
879	<	txt += 'if [ -e $RUNTIME_AREA/'+relFile+' ] ; then\n'
880	<	txt += ' cp $RUNTIME_AREA/'+relFile+' .\n'
881	<	txt += ' chmod +x '+relFile+'\n'
882	<	txt += 'fi\n'
883	<	pass
884	<
885	<	if self.pset != None: #CarlosDaniele
886	<	txt += 'echo "### END JOB SETUP ENVIRONMENT ###"\n\n'
887	<
922	>
923	>	if self.pset != None:
924	>	# FUTURE: Can simply for 2_1_x and higher
925		txt += '\n'
926	<	txt += 'echo "*** cat pset.cfg *******"\n'
927	<	txt += 'cat pset.cfg\n'
928	<	txt += 'echo "**** end pset.cfg ******"\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'
936		txt += '\n'
893	–	# txt += 'echo "*** cat pset1.cfg *******"\n'
894	–	# txt += 'cat pset1.cfg\n'
895	–	# txt += 'echo "**** end pset1.cfg ******"\n'
937		return txt
938
939	<	def wsBuildExe(self, nj):
939	>	def wsUntarSoftware(self, nj=0):
940		"""
941		Put in the script the commands to build an executable
942		or a library.
943		"""
944
945	<	txt = ""
945	>	txt = '\n#Written by cms_cmssw::wsUntarSoftware\n'
946
947		if os.path.isfile(self.tgzNameWithPath):
948	<	txt += 'echo "tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'"\n'
948	>	txt += 'echo ">>> tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+' :" \n'
949		txt += 'tar xzvf $RUNTIME_AREA/'+os.path.basename(self.tgzNameWithPath)+'\n'
950	+	if self.debug_wrapper:
951	+	txt += 'ls -Al \n'
952		txt += 'untar_status=$? \n'
953		txt += 'if [ $untar_status -ne 0 ]; then \n'
954	<	txt += ' echo "SET_EXE 1 ==> ERROR Untarring .tgz file failed"\n'
955	<	txt += ' echo "JOB_EXIT_STATUS = $untar_status" \n'
956	<	txt += ' echo "JobExitCode=$untar_status" \| tee -a $RUNTIME_AREA/$repo\n'
914	<	txt += ' if [ $middleware == OSG ]; then \n'
915	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
916	<	txt += ' cd $RUNTIME_AREA\n'
917	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
918	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
919	<	txt += ' echo "SET_EXE 50999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after Untarring .tgz file failed"\n'
920	<	txt += ' echo "JOB_EXIT_STATUS = 50999"\n'
921	<	txt += ' echo "JobExitCode=50999" \| tee -a $RUNTIME_AREA/$repo\n'
922	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
923	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
924	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
925	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
926	<	txt += ' fi\n'
927	<	txt += ' fi \n'
928	<	txt += ' \n'
929	<	txt += ' exit 1 \n'
954	>	txt += ' echo "ERROR ==> Untarring .tgz file failed"\n'
955	>	txt += ' job_exit_code=$untar_status\n'
956	>	txt += ' func_exit\n'
957		txt += 'else \n'
958		txt += ' echo "Successful untar" \n'
959		txt += 'fi \n'
960		txt += '\n'
961	<	txt += 'echo "Include ProdAgentApi in PYTHONPATH"\n'
961	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
962		txt += 'if [ -z "$PYTHONPATH" ]; then\n'
963	<	txt += ' export PYTHONPATH=ProdAgentApi\n'
963	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/\n'
964		txt += 'else\n'
965	<	txt += ' export PYTHONPATH=ProdAgentApi:${PYTHONPATH}\n'
965	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
966	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
967		txt += 'fi\n'
968		txt += '\n'
969
970		pass
971	<
971	>
972		return txt
973
974	<	def modifySteeringCards(self, nj):
974	>	def wsBuildExe(self, nj=0):
975		"""
976	<	modify the card provided by the user,
977	<	writing a new card into share dir
976	>	Put in the script the commands to build an executable
977	>	or a library.
978		"""
979	<
979	>
980	>	txt = '\n#Written by cms_cmssw::wsBuildExe\n'
981	>	txt += 'echo ">>> moving CMSSW software directories in `pwd`" \n'
982	>
983	>	txt += 'rm -r lib/ module/ \n'
984	>	txt += 'mv $RUNTIME_AREA/lib/ . \n'
985	>	txt += 'mv $RUNTIME_AREA/module/ . \n'
986	>	if self.dataExist == True:
987	>	txt += 'rm -r src/ \n'
988	>	txt += 'mv $RUNTIME_AREA/src/ . \n'
989	>	if len(self.additional_inbox_files)>0:
990	>	for file in self.additional_inbox_files:
991	>	txt += 'mv $RUNTIME_AREA/'+os.path.basename(file)+' . \n'
992	>	# txt += 'mv $RUNTIME_AREA/ProdCommon/ . \n'
993	>	# txt += 'mv $RUNTIME_AREA/IMProv/ . \n'
994	>
995	>	txt += 'echo ">>> Include $RUNTIME_AREA in PYTHONPATH:"\n'
996	>	txt += 'if [ -z "$PYTHONPATH" ]; then\n'
997	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/\n'
998	>	txt += 'else\n'
999	>	txt += ' export PYTHONPATH=$RUNTIME_AREA/:${PYTHONPATH}\n'
1000	>	txt += 'echo "PYTHONPATH=$PYTHONPATH"\n'
1001	>	txt += 'fi\n'
1002	>	txt += '\n'
1003	>
1004	>	return txt
1005	>
1006	>
1007		def executableName(self):
1008	<	if self.pset == None: #CarlosDaniele
1008	>	if self.scriptExe:
1009		return "sh "
1010		else:
1011		return self.executable
1012
1013		def executableArgs(self):
1014	<	if self.pset == None:#CarlosDaniele
1014	>	# FUTURE: This function tests the CMSSW version. Can be simplified as we drop support for old versions
1015	>	if self.scriptExe:#CarlosDaniele
1016		return self.scriptExe + " $NJob"
1017	<	else:
1018	<	return " -p pset.cfg"
1017	>	else:
1018	>	ex_args = ""
1019	>	# FUTURE: This tests the CMSSW version. Can remove code as versions deprecated
1020	>	# 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
1024	>	if self.CMSSW_major >= 2 :
1025	>	ex_args += " -p pset.py"
1026	>	else:
1027	>	ex_args += " -p pset.cfg"
1028	>	return ex_args
1029
1030		def inputSandbox(self, nj):
1031		"""
1032		Returns a list of filenames to be put in JDL input sandbox.
1033		"""
1034		inp_box = []
969	–	# # dict added to delete duplicate from input sandbox file list
970	–	# seen = {}
971	–	## code
1035		if os.path.isfile(self.tgzNameWithPath):
1036		inp_box.append(self.tgzNameWithPath)
1037	<	if os.path.isfile(self.MLtgzfile):
1038	<	inp_box.append(self.MLtgzfile)
976	<	## config
977	<	if not self.pset is None: #CarlosDaniele
978	<	inp_box.append(common.work_space.pathForTgz() + 'job/' + self.configFilename())
979	<	## additional input files
980	<	#for file in self.additional_inbox_files:
981	<	# inp_box.append(common.work_space.cwdDir()+file)
1037	>	wrapper = os.path.basename(str(common._db.queryTask('scriptName')))
1038	>	inp_box.append(common.work_space.pathForTgz() +'job/'+ wrapper)
1039		return inp_box
1040
1041		def outputSandbox(self, nj):
#	Line 989 \| Line 1046 \| class Cmssw(JobType):
1046
1047		## User Declared output files
1048		for out in (self.output_file+self.output_file_sandbox):
1049	<	n_out = nj + 1
1050	<	out_box.append(self.numberFile_(out,str(n_out)))
1049	>	n_out = nj + 1
1050	>	out_box.append(numberFile(out,str(n_out)))
1051		return out_box
1052
996	–	def prepareSteeringCards(self):
997	–	"""
998	–	Make initial modifications of the user's steering card file.
999	–	"""
1000	–	return
1053
1054		def wsRenameOutput(self, nj):
1055		"""
1056		Returns part of a job script which renames the produced files.
1057		"""
1058
1059	<	txt = '\n'
1060	<	txt += '# directory content\n'
1061	<	txt += 'ls \n'
1059	>	txt = '\n#Written by cms_cmssw::wsRenameOutput\n'
1060	>	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1061	>	txt += 'echo ">>> current directory content:"\n'
1062	>	if self.debug_wrapper:
1063	>	txt += 'ls -Al\n'
1064	>	txt += '\n'
1065
1066	<	for fileWithSuffix in (self.output_file+self.output_file_sandbox):
1067	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1066	>	for fileWithSuffix in (self.output_file):
1067	>	output_file_num = numberFile(fileWithSuffix, '$NJob')
1068		txt += '\n'
1069		txt += '# check output file\n'
1070	<	txt += 'ls '+fileWithSuffix+'\n'
1071	<	txt += 'ls_result=$?\n'
1072	<	txt += 'if [ $ls_result -ne 0 ] ; then\n'
1073	<	txt += ' echo "ERROR: Problem with output file"\n'
1074	<	if common.scheduler.boss_scheduler_name == 'condor_g':
1070	>	txt += 'if [ -e ./'+fileWithSuffix+' ] ; then\n'
1071	>	if (self.copy_data == 1): # For OSG nodes, file is in $WORKING_DIR, should not be moved to $RUNTIME_AREA
1072	>	txt += ' mv '+fileWithSuffix+' '+output_file_num+'\n'
1073	>	txt += ' ln -s `pwd`/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1074	>	else:
1075	>	txt += ' mv '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1076	>	txt += ' ln -s $RUNTIME_AREA/'+output_file_num+' $RUNTIME_AREA/'+fileWithSuffix+'\n'
1077	>	txt += 'else\n'
1078	>	txt += ' job_exit_code=60302\n'
1079	>	txt += ' echo "WARNING: Output file '+fileWithSuffix+' not found"\n'
1080	>	if common.scheduler.name().upper() == 'CONDOR_G':
1081		txt += ' if [ $middleware == OSG ]; then \n'
1082		txt += ' echo "prepare dummy output file"\n'
1083		txt += ' echo "Processing of job output failed" > $RUNTIME_AREA/'+output_file_num+'\n'
1084		txt += ' fi \n'
1024	–	txt += 'else\n'
1025	–	txt += ' cp '+fileWithSuffix+' $RUNTIME_AREA/'+output_file_num+'\n'
1085		txt += 'fi\n'
1086	<
1087	<	txt += 'cd $RUNTIME_AREA\n'
1088	<	txt += 'cd $RUNTIME_AREA\n'
1030	<	### OLI_DANIELE
1031	<	txt += 'if [ $middleware == OSG ]; then\n'
1032	<	txt += ' cd $RUNTIME_AREA\n'
1033	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
1034	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
1035	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
1036	<	txt += ' echo "SET_EXE 60999 ==> OSG $WORKING_DIR could not be deleted on WN `hostname` after cleanup of WN"\n'
1037	<	txt += ' echo "JOB_EXIT_STATUS = 60999"\n'
1038	<	txt += ' echo "JobExitCode=60999" \| tee -a $RUNTIME_AREA/$repo\n'
1039	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1040	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1041	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1042	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1043	<	txt += ' fi\n'
1044	<	txt += 'fi\n'
1045	<	txt += '\n'
1046	<
1047	<	file_list = ''
1048	<	## Add to filelist only files to be possibly copied to SE
1049	<	for fileWithSuffix in self.output_file:
1050	<	output_file_num = self.numberFile_(fileWithSuffix, '$NJob')
1051	<	file_list=file_list+output_file_num+' '
1052	<	file_list=file_list[:-1]
1053	<	txt += 'file_list="'+file_list+'"\n'
1086	>	file_list = []
1087	>	for fileWithSuffix in (self.output_file):
1088	>	file_list.append(numberFile(fileWithSuffix, '$NJob'))
1089
1090	+	txt += 'file_list="'+string.join(file_list,' ')+'"\n'
1091	+	txt += '\n'
1092	+	txt += 'echo ">>> current directory (SOFTWARE_DIR): $SOFTWARE_DIR" \n'
1093	+	txt += 'echo ">>> current directory content:"\n'
1094	+	if self.debug_wrapper:
1095	+	txt += 'ls -Al\n'
1096	+	txt += '\n'
1097	+	txt += 'cd $RUNTIME_AREA\n'
1098	+	txt += 'echo ">>> current directory (RUNTIME_AREA): $RUNTIME_AREA"\n'
1099		return txt
1100
1101	<	def numberFile_(self, file, txt):
1101	>	def getRequirements(self, nj=[]):
1102		"""
1103	<	append _'txt' before last extension of a file
1060	<	"""
1061	<	p = string.split(file,".")
1062	<	# take away last extension
1063	<	name = p[0]
1064	<	for x in p[1:-1]:
1065	<	name=name+"."+x
1066	<	# add "_txt"
1067	<	if len(p)>1:
1068	<	ext = p[len(p)-1]
1069	<	result = name + '_' + txt + "." + ext
1070	<	else:
1071	<	result = name + '_' + txt
1072	<
1073	<	return result
1074	<
1075	<	def getRequirements(self):
1076	<	"""
1077	<	return job requirements to add to jdl files
1103	>	return job requirements to add to jdl files
1104		"""
1105		req = ''
1106		if self.version:
1107		req='Member("VO-cms-' + \
1108		self.version + \
1109		'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1110	+	if self.executable_arch:
1111	+	req+=' && Member("VO-cms-' + \
1112	+	self.executable_arch + \
1113	+	'", other.GlueHostApplicationSoftwareRunTimeEnvironment)'
1114
1115		req = req + ' && (other.GlueHostNetworkAdapterOutboundIP)'
1116	+	if ( common.scheduler.name() == "glitecoll" ) or ( common.scheduler.name() == "glite"):
1117	+	req += ' && other.GlueCEStateStatus == "Production" '
1118
1119		return req
1120
1121		def configFilename(self):
1122		""" return the config filename """
1123	<	return self.name()+'.cfg'
1123	>	# FUTURE: Can remove cfg mode for CMSSW >= 2_1_x
1124	>	if (self.CMSSW_major >= 2 and self.CMSSW_minor >= 1) or (self.CMSSW_major >= 3):
1125	>	return self.name()+'.py'
1126	>	else:
1127	>	return self.name()+'.cfg'
1128
1093	–	### OLI_DANIELE
1129		def wsSetupCMSOSGEnvironment_(self):
1130		"""
1131		Returns part of a job script which is prepares
1132		the execution environment and which is common for all CMS jobs.
1133		"""
1134	<	txt = '\n'
1135	<	txt += ' echo "### SETUP CMS OSG ENVIRONMENT ###"\n'
1136	<	txt += ' if [ -f $GRID3_APP_DIR/cmssoft/cmsset_default.sh ] ;then\n'
1137	<	txt += ' # Use $GRID3_APP_DIR/cmssoft/cmsset_default.sh to setup cms software\n'
1138	<	txt += ' source $GRID3_APP_DIR/cmssoft/cmsset_default.sh '+self.version+'\n'
1139	<	txt += ' elif [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1134	>	txt = '\n#Written by cms_cmssw::wsSetupCMSOSGEnvironment_\n'
1135	>	txt += ' echo ">>> setup CMS OSG environment:"\n'
1136	>	txt += ' echo "set SCRAM ARCH to ' + self.executable_arch + '"\n'
1137	>	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1138	>	txt += ' echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1139	>	txt += ' if [ -f $OSG_APP/cmssoft/cms/cmsset_default.sh ] ;then\n'
1140		txt += ' # Use $OSG_APP/cmssoft/cms/cmsset_default.sh to setup cms software\n'
1141	<	txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1142	<	txt += ' else\n'
1143	<	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'
1144	<	txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1145	<	txt += ' echo "JobExitCode=10020" \| tee -a $RUNTIME_AREA/$repo\n'
1146	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1112	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1113	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1114	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1115	<	txt += ' exit 1\n'
1116	<	txt += '\n'
1117	<	txt += ' echo "Remove working directory: $WORKING_DIR"\n'
1118	<	txt += ' cd $RUNTIME_AREA\n'
1119	<	txt += ' /bin/rm -rf $WORKING_DIR\n'
1120	<	txt += ' if [ -d $WORKING_DIR ] ;then\n'
1121	<	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'
1122	<	txt += ' echo "JOB_EXIT_STATUS = 10017"\n'
1123	<	txt += ' echo "JobExitCode=10017" \| tee -a $RUNTIME_AREA/$repo\n'
1124	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1125	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1126	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1127	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1128	<	txt += ' fi\n'
1129	<	txt += '\n'
1130	<	txt += ' exit 1\n'
1131	<	txt += ' fi\n'
1141	>	txt += ' source $OSG_APP/cmssoft/cms/cmsset_default.sh '+self.version+'\n'
1142	>	txt += ' else\n'
1143	>	txt += ' echo "ERROR ==> $OSG_APP/cmssoft/cms/cmsset_default.sh file not found"\n'
1144	>	txt += ' job_exit_code=10020\n'
1145	>	txt += ' func_exit\n'
1146	>	txt += ' fi\n'
1147		txt += '\n'
1148	<	txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1149	<	txt += ' echo " END SETUP CMS OSG ENVIRONMENT "\n'
1148	>	txt += ' echo "==> setup cms environment ok"\n'
1149	>	txt += ' echo "SCRAM_ARCH = $SCRAM_ARCH"\n'
1150
1151		return txt
1152	<
1138	<	### OLI_DANIELE
1152	>
1153		def wsSetupCMSLCGEnvironment_(self):
1154		"""
1155		Returns part of a job script which is prepares
1156		the execution environment and which is common for all CMS jobs.
1157		"""
1158	<	txt = ' \n'
1159	<	txt += ' echo " ### SETUP CMS LCG ENVIRONMENT ### "\n'
1160	<	txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n'
1161	<	txt += ' echo "SET_CMS_ENV 10031 ==> ERROR CMS software dir not found on WN `hostname`"\n'
1162	<	txt += ' echo "JOB_EXIT_STATUS = 10031" \n'
1163	<	txt += ' echo "JobExitCode=10031" \| tee -a $RUNTIME_AREA/$repo\n'
1164	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1165	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1166	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1167	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1168	<	txt += ' exit 1\n'
1169	<	txt += ' else\n'
1170	<	txt += ' echo "Sourcing environment... "\n'
1171	<	txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1172	<	txt += ' echo "SET_CMS_ENV 10020 ==> ERROR cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1173	<	txt += ' echo "JOB_EXIT_STATUS = 10020"\n'
1174	<	txt += ' echo "JobExitCode=10020" \| 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 += ' fi\n'
1181	<	txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1182	<	txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1183	<	txt += ' result=$?\n'
1184	<	txt += ' if [ $result -ne 0 ]; then\n'
1185	<	txt += ' echo "SET_CMS_ENV 10032 ==> ERROR problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1186	<	txt += ' echo "JOB_EXIT_STATUS = 10032"\n'
1187	<	txt += ' echo "JobExitCode=10032" \| tee -a $RUNTIME_AREA/$repo\n'
1188	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1189	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1190	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1191	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1192	<	txt += ' exit 1\n'
1193	<	txt += ' fi\n'
1194	<	txt += ' fi\n'
1195	<	txt += ' \n'
1196	<	txt += ' string=`cat /etc/redhat-release`\n'
1197	<	txt += ' echo $string\n'
1198	<	txt += ' if [[ $string = alhalla ]]; then\n'
1199	<	txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1200	<	txt += ' elif [[ $string = Enterprise ]] \|\| [[ $string = cientific ]]; then\n'
1201	<	txt += ' export SCRAM_ARCH=slc3_ia32_gcc323\n'
1202	<	txt += ' echo "SCRAM_ARCH= $SCRAM_ARCH"\n'
1203	<	txt += ' else\n'
1204	<	txt += ' echo "SET_CMS_ENV 10033 ==> ERROR OS unknown, LCG environment not initialized"\n'
1205	<	txt += ' echo "JOB_EXIT_STATUS = 10033"\n'
1206	<	txt += ' echo "JobExitCode=10033" \| tee -a $RUNTIME_AREA/$repo\n'
1207	<	txt += ' dumpStatus $RUNTIME_AREA/$repo\n'
1208	<	txt += ' rm -f $RUNTIME_AREA/$repo \n'
1209	<	txt += ' echo "MonitorJobID=`echo $MonitorJobID`" \| tee -a $RUNTIME_AREA/$repo \n'
1210	<	txt += ' echo "MonitorID=`echo $MonitorID`" \| tee -a $RUNTIME_AREA/$repo\n'
1211	<	txt += ' exit 1\n'
1212	<	txt += ' fi\n'
1213	<	txt += ' echo "SET_CMS_ENV 0 ==> setup cms environment ok"\n'
1214	<	txt += ' echo "### END SETUP CMS LCG ENVIRONMENT ###"\n'
1158	>	txt = '\n#Written by cms_cmssw::wsSetupCMSLCGEnvironment_\n'
1159	>	txt += ' echo ">>> setup CMS LCG environment:"\n'
1160	>	txt += ' echo "set SCRAM ARCH and BUILD_ARCH to ' + self.executable_arch + ' ###"\n'
1161	>	txt += ' export SCRAM_ARCH='+self.executable_arch+'\n'
1162	>	txt += ' export BUILD_ARCH='+self.executable_arch+'\n'
1163	>	txt += ' if [ ! $VO_CMS_SW_DIR ] ;then\n'
1164	>	txt += ' echo "ERROR ==> CMS software dir not found on WN `hostname`"\n'
1165	>	txt += ' job_exit_code=10031\n'
1166	>	txt += ' func_exit\n'
1167	>	txt += ' else\n'
1168	>	txt += ' echo "Sourcing environment... "\n'
1169	>	txt += ' if [ ! -s $VO_CMS_SW_DIR/cmsset_default.sh ] ;then\n'
1170	>	txt += ' echo "ERROR ==> cmsset_default.sh file not found into dir $VO_CMS_SW_DIR"\n'
1171	>	txt += ' job_exit_code=10020\n'
1172	>	txt += ' func_exit\n'
1173	>	txt += ' fi\n'
1174	>	txt += ' echo "sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1175	>	txt += ' source $VO_CMS_SW_DIR/cmsset_default.sh\n'
1176	>	txt += ' result=$?\n'
1177	>	txt += ' if [ $result -ne 0 ]; then\n'
1178	>	txt += ' echo "ERROR ==> problem sourcing $VO_CMS_SW_DIR/cmsset_default.sh"\n'
1179	>	txt += ' job_exit_code=10032\n'
1180	>	txt += ' func_exit\n'
1181	>	txt += ' fi\n'
1182	>	txt += ' fi\n'
1183	>	txt += ' \n'
1184	>	txt += ' echo "==> setup cms environment ok"\n'
1185	>	return txt
1186	>
1187	>	def modifyReport(self, nj):
1188	>	"""
1189	>	insert the part of the script that modifies the FrameworkJob Report
1190	>	"""
1191	>	txt = '\n#Written by cms_cmssw::modifyReport\n'
1192	>	publish_data = int(self.cfg_params.get('USER.publish_data',0))
1193	>	if (publish_data == 1):
1194	>	processedDataset = self.cfg_params['USER.publish_data_name']
1195	>	if (self.primaryDataset == 'null'):
1196	>	self.primaryDataset = processedDataset
1197	>	if (common.scheduler.name().upper() == "CAF" or common.scheduler.name().upper() == "LSF"):
1198	>	### FEDE FOR NEW LFN ###
1199	>	LFNBaseName = LFNBase(self.primaryDataset, processedDataset, LocalUser=True)
1200	>	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	>	########################
1213	>	txt += 'else\n'
1214	>	txt += ' FOR_LFN=/copy_problems/ \n'
1215	>	txt += ' SE=""\n'
1216	>	txt += ' SE_PATH=""\n'
1217	>	txt += 'fi\n'
1218	>
1219	>	txt += 'echo ">>> Modify Job Report:" \n'
1220	>	txt += 'chmod a+x $RUNTIME_AREA/ProdCommon/FwkJobRep/ModifyJobReport.py\n'
1221	>	txt += 'ProcessedDataset='+processedDataset+'\n'
1222	>	txt += 'echo "ProcessedDataset = $ProcessedDataset"\n'
1223	>	txt += 'echo "SE = $SE"\n'
1224	>	txt += 'echo "SE_PATH = $SE_PATH"\n'
1225	>	txt += 'echo "FOR_LFN = $FOR_LFN" \n'
1226	>	txt += 'echo "CMSSW_VERSION = $CMSSW_VERSION"\n\n'
1227	>	### FEDE FOR NEW LFN ###
1228	>	txt += 'echo "$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 $SE_PATH"\n'
1229	>	txt += '$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 $SE_PATH\n'
1230	>	########################
1231	>	txt += 'modifyReport_result=$?\n'
1232	>	txt += 'if [ $modifyReport_result -ne 0 ]; then\n'
1233	>	txt += ' modifyReport_result=70500\n'
1234	>	txt += ' job_exit_code=$modifyReport_result\n'
1235	>	txt += ' echo "ModifyReportResult=$modifyReport_result" \| tee -a $RUNTIME_AREA/$repo\n'
1236	>	txt += ' echo "WARNING: Problem with ModifyJobReport"\n'
1237	>	txt += 'else\n'
1238	>	txt += ' mv NewFrameworkJobReport.xml $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1239	>	txt += 'fi\n'
1240	>	return txt
1241	>
1242	>	def wsParseFJR(self):
1243	>	"""
1244	>	Parse the FrameworkJobReport to obtain useful infos
1245	>	"""
1246	>	txt = '\n#Written by cms_cmssw::wsParseFJR\n'
1247	>	txt += 'echo ">>> Parse FrameworkJobReport crab_fjr.xml"\n'
1248	>	txt += 'if [ -s $RUNTIME_AREA/crab_fjr_$NJob.xml ]; then\n'
1249	>	txt += ' if [ -s $RUNTIME_AREA/parseCrabFjr.py ]; then\n'
1250	>	txt += ' cmd_out=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --dashboard $MonitorID,$MonitorJobID '+self.debugWrap+'`\n'
1251	>	if self.debug_wrapper :
1252	>	txt += ' echo "Result of parsing the FrameworkJobReport crab_fjr.xml: $cmd_out"\n'
1253	>	txt += ' executable_exit_status=`python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --exitcode`\n'
1254	>	txt += ' if [ $executable_exit_status -eq 50115 ];then\n'
1255	>	txt += ' echo ">>> crab_fjr.xml contents: "\n'
1256	>	txt += ' cat $RUNTIME_AREA/crab_fjr_$NJob.xml\n'
1257	>	txt += ' echo "Wrong FrameworkJobReport --> does not contain useful info. ExitStatus: $executable_exit_status"\n'
1258	>	txt += ' elif [ $executable_exit_status -eq -999 ];then\n'
1259	>	txt += ' echo "ExitStatus from FrameworkJobReport not available. not available. Using exit code of executable from command line."\n'
1260	>	txt += ' else\n'
1261	>	txt += ' echo "Extracted ExitStatus from FrameworkJobReport parsing output: $executable_exit_status"\n'
1262	>	txt += ' fi\n'
1263	>	txt += ' else\n'
1264	>	txt += ' echo "CRAB python script to parse CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1265	>	txt += ' fi\n'
1266	>	#### Patch to check input data reading for CMSSW16x Hopefully we-ll remove it asap
1267	>
1268	>	if (self.datasetPath and not (self.dataset_pu or self.useParent) :
1269	>	# VERIFY PROCESSED DATA
1270	>	txt += ' if [ $executable_exit_status -eq 0 ];then\n'
1271	>	txt += ' echo ">>> Verify list of processed files:"\n'
1272	>	txt += ' echo $InputFiles \|tr -d \'\\\\\' \|tr \',\' \'\\n\'\|tr -d \'"\' > input-files.txt\n'
1273	>	txt += ' python $RUNTIME_AREA/parseCrabFjr.py --input $RUNTIME_AREA/crab_fjr_$NJob.xml --lfn > processed-files.txt\n'
1274	>	txt += ' cat input-files.txt \| sort \| uniq > tmp.txt\n'
1275	>	txt += ' mv tmp.txt input-files.txt\n'
1276	>	txt += ' echo "cat input-files.txt"\n'
1277	>	txt += ' echo "----------------------"\n'
1278	>	txt += ' cat input-files.txt\n'
1279	>	txt += ' cat processed-files.txt \| sort \| uniq > tmp.txt\n'
1280	>	txt += ' mv tmp.txt processed-files.txt\n'
1281	>	txt += ' echo "----------------------"\n'
1282	>	txt += ' echo "cat processed-files.txt"\n'
1283	>	txt += ' echo "----------------------"\n'
1284	>	txt += ' cat processed-files.txt\n'
1285	>	txt += ' echo "----------------------"\n'
1286	>	txt += ' diff -q input-files.txt processed-files.txt\n'
1287	>	txt += ' fileverify_status=$?\n'
1288	>	txt += ' if [ $fileverify_status -ne 0 ]; then\n'
1289	>	txt += ' executable_exit_status=30001\n'
1290	>	txt += ' echo "ERROR ==> not all input files processed"\n'
1291	>	txt += ' echo " ==> list of processed files from crab_fjr.xml differs from list in pset.cfg"\n'
1292	>	txt += ' echo " ==> diff input-files.txt processed-files.txt"\n'
1293	>	txt += ' fi\n'
1294	>	txt += ' fi\n'
1295	>	txt += '\n'
1296	>	txt += 'else\n'
1297	>	txt += ' echo "CRAB FrameworkJobReport crab_fjr.xml is not available, using exit code of executable from command line."\n'
1298	>	txt += 'fi\n'
1299	>	txt += '\n'
1300	>	txt += 'echo "ExeExitCode=$executable_exit_status" \| tee -a $RUNTIME_AREA/$repo\n'
1301	>	txt += 'echo "EXECUTABLE_EXIT_STATUS = $executable_exit_status"\n'
1302	>	txt += 'job_exit_code=$executable_exit_status\n'
1303	>
1304		return txt
1305
1306		def setParam_(self, param, value):
#	Line 1206 \| Line 1309 \| class Cmssw(JobType):
1309		def getParams(self):
1310		return self._params
1311
1209	–	def setTaskid_(self):
1210	–	self._taskId = self.cfg_params['taskId']
1211	–
1212	–	def getTaskid(self):
1213	–	return self._taskId
1214	–
1215	–	#######################################################################
1312		def uniquelist(self, old):
1313		"""
1314		remove duplicates from a list
#	Line 1221 \| Line 1317 \| class Cmssw(JobType):
1317		for e in old:
1318		nd[e]=0
1319		return nd.keys()
1320	+
1321	+	def outList(self):
1322	+	"""
1323	+	check the dimension of the output files
1324	+	"""
1325	+	txt = ''
1326	+	txt += 'echo ">>> list of expected files on output sandbox"\n'
1327	+	listOutFiles = []
1328	+	stdout = 'CMSSW_$NJob.stdout'
1329	+	stderr = 'CMSSW_$NJob.stderr'
1330	+	if (self.return_data == 1):
1331	+	for file in (self.output_file+self.output_file_sandbox):
1332	+	listOutFiles.append(numberFile(file, '$NJob'))
1333	+	listOutFiles.append(stdout)
1334	+	listOutFiles.append(stderr)
1335	+	else:
1336	+	for file in (self.output_file_sandbox):
1337	+	listOutFiles.append(numberFile(file, '$NJob'))
1338	+	listOutFiles.append(stdout)
1339	+	listOutFiles.append(stderr)
1340	+	txt += 'echo "output files: '+string.join(listOutFiles,' ')+'"\n'
1341	+	txt += 'filesToCheck="'+string.join(listOutFiles,' ')+'"\n'
1342	+	txt += 'export filesToCheck\n'
1343	+	return txt

Diff Legend

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

Comparing COMP/CRAB/python/cms_cmssw.py (file contents): Revision 1.62 by slacapra, Wed Jan 17 13:37:25 2007 UTC vs. Revision 1.231 by spiga, Mon Aug 25 16:13:05 2008 UTC

Diff Legend

Comparing COMP/CRAB/python/cms_cmssw.py (file contents):
Revision 1.62 by slacapra, Wed Jan 17 13:37:25 2007 UTC vs.
Revision 1.231 by spiga, Mon Aug 25 16:13:05 2008 UTC