CMS 3D CMS Logo

ConfigBuilder.py
Go to the documentation of this file.
1 #! /usr/bin/env python
2 
3 from __future__ import print_function
4 __version__ = "$Revision: 1.19 $"
5 __source__ = "$Source: /local/reps/CMSSW/CMSSW/Configuration/Applications/python/ConfigBuilder.py,v $"
6 
7 import FWCore.ParameterSet.Config as cms
8 from FWCore.ParameterSet.Modules import _Module
9 
10 import six
11 # The following import is provided for backward compatibility reasons.
12 # The function used to be defined in this file.
13 from FWCore.ParameterSet.MassReplace import massReplaceInputTag as MassReplaceInputTag
14 
15 import sys
16 import re
17 import collections
18 from subprocess import Popen,PIPE
19 import FWCore.ParameterSet.DictTypes as DictTypes
20 class Options:
21  pass
22 
23 # the canonical defaults
24 defaultOptions = Options()
25 defaultOptions.datamix = 'DataOnSim'
26 defaultOptions.isMC=False
27 defaultOptions.isData=True
28 defaultOptions.step=''
29 defaultOptions.pileup='NoPileUp'
30 defaultOptions.pileup_input = None
31 defaultOptions.pileup_dasoption = ''
32 defaultOptions.geometry = 'SimDB'
33 defaultOptions.geometryExtendedOptions = ['ExtendedGFlash','Extended','NoCastor']
34 defaultOptions.magField = ''
35 defaultOptions.conditions = None
36 defaultOptions.scenarioOptions=['pp','cosmics','nocoll','HeavyIons']
37 defaultOptions.harvesting= 'AtRunEnd'
38 defaultOptions.gflash = False
39 defaultOptions.number = -1
40 defaultOptions.number_out = None
41 defaultOptions.arguments = ""
42 defaultOptions.name = "NO NAME GIVEN"
43 defaultOptions.evt_type = ""
44 defaultOptions.filein = ""
45 defaultOptions.dasquery=""
46 defaultOptions.dasoption=""
47 defaultOptions.secondfilein = ""
48 defaultOptions.customisation_file = []
49 defaultOptions.customisation_file_unsch = []
50 defaultOptions.customise_commands = ""
51 defaultOptions.inline_custom=False
52 defaultOptions.particleTable = 'pythiapdt'
53 defaultOptions.particleTableList = ['pythiapdt','pdt']
54 defaultOptions.dirin = ''
55 defaultOptions.dirout = ''
56 defaultOptions.filetype = 'EDM'
57 defaultOptions.fileout = 'output.root'
58 defaultOptions.filtername = ''
59 defaultOptions.lazy_download = False
60 defaultOptions.custom_conditions = ''
61 defaultOptions.hltProcess = ''
62 defaultOptions.eventcontent = None
63 defaultOptions.datatier = None
64 defaultOptions.inlineEventContent = True
65 defaultOptions.inlineObjets =''
66 defaultOptions.hideGen=False
67 from Configuration.StandardSequences.VtxSmeared import VtxSmearedDefaultKey,VtxSmearedHIDefaultKey
68 defaultOptions.beamspot=None
69 defaultOptions.outputDefinition =''
70 defaultOptions.inputCommands = None
71 defaultOptions.outputCommands = None
72 defaultOptions.inputEventContent = ''
73 defaultOptions.dropDescendant = False
74 defaultOptions.relval = None
75 defaultOptions.profile = None
76 defaultOptions.isRepacked = False
77 defaultOptions.restoreRNDSeeds = False
78 defaultOptions.donotDropOnInput = ''
79 defaultOptions.python_filename =''
80 defaultOptions.io=None
81 defaultOptions.lumiToProcess=None
82 defaultOptions.fast=False
83 defaultOptions.runsAndWeightsForMC = None
84 defaultOptions.runsScenarioForMC = None
85 defaultOptions.runUnscheduled = False
86 defaultOptions.timeoutOutput = False
87 defaultOptions.nThreads = '1'
88 defaultOptions.nStreams = '0'
89 defaultOptions.nConcurrentLumis = '1'
90 
91 # some helper routines
92 def dumpPython(process,name):
93  theObject = getattr(process,name)
94  if isinstance(theObject,cms.Path) or isinstance(theObject,cms.EndPath) or isinstance(theObject,cms.Sequence):
95  return "process."+name+" = " + theObject.dumpPython("process")
96  elif isinstance(theObject,_Module) or isinstance(theObject,cms.ESProducer):
97  return "process."+name+" = " + theObject.dumpPython()+"\n"
98  else:
99  return "process."+name+" = " + theObject.dumpPython()+"\n"
100 def filesFromList(fileName,s=None):
101  import os
102  import FWCore.ParameterSet.Config as cms
103  prim=[]
104  sec=[]
105  for line in open(fileName,'r'):
106  if line.count(".root")>=2:
107  #two files solution...
108  entries=line.replace("\n","").split()
109  if not entries[0] in prim:
110  prim.append(entries[0])
111  if not entries[1] in sec:
112  sec.append(entries[1])
113  elif (line.find(".root")!=-1):
114  entry=line.replace("\n","")
115  if not entry in prim:
116  prim.append(entry)
117  if s:
118  if not hasattr(s,"fileNames"):
119  s.fileNames=cms.untracked.vstring(prim)
120  else:
121  s.fileNames.extend(prim)
122  if len(sec)!=0:
123  if not hasattr(s,"secondaryFileNames"):
124  s.secondaryFileNames=cms.untracked.vstring(sec)
125  else:
126  s.secondaryFileNames.extend(sec)
127  print("found files: ",prim)
128  if len(prim)==0:
129  raise Exception("There are not files in input from the file list")
130  if len(sec)!=0:
131  print("found parent files:",sec)
132  return (prim,sec)
133 
134 def filesFromDASQuery(query,option="",s=None):
135  import os,time
136  import FWCore.ParameterSet.Config as cms
137  prim=[]
138  sec=[]
139  print("the query is",query)
140  eC=5
141  count=0
142  while eC!=0 and count<3:
143  if count!=0:
144  print('Sleeping, then retrying DAS')
145  time.sleep(100)
146  p = Popen('dasgoclient %s --query "%s"'%(option,query), stdout=PIPE,shell=True)
147  pipe=p.stdout.read()
148  tupleP = os.waitpid(p.pid, 0)
149  eC=tupleP[1]
150  count=count+1
151  if eC==0:
152  print("DAS succeeded after",count,"attempts",eC)
153  else:
154  print("DAS failed 3 times- I give up")
155  for line in pipe.split('\n'):
156  if line.count(".root")>=2:
157  #two files solution...
158  entries=line.replace("\n","").split()
159  if not entries[0] in prim:
160  prim.append(entries[0])
161  if not entries[1] in sec:
162  sec.append(entries[1])
163  elif (line.find(".root")!=-1):
164  entry=line.replace("\n","")
165  if not entry in prim:
166  prim.append(entry)
167  if s:
168  if not hasattr(s,"fileNames"):
169  s.fileNames=cms.untracked.vstring(prim)
170  else:
171  s.fileNames.extend(prim)
172  if len(sec)!=0:
173  if not hasattr(s,"secondaryFileNames"):
174  s.secondaryFileNames=cms.untracked.vstring(sec)
175  else:
176  s.secondaryFileNames.extend(sec)
177  print("found files: ",prim)
178  if len(sec)!=0:
179  print("found parent files:",sec)
180  return (prim,sec)
181 
182 def anyOf(listOfKeys,dict,opt=None):
183  for k in listOfKeys:
184  if k in dict:
185  toReturn=dict[k]
186  dict.pop(k)
187  return toReturn
188  if opt!=None:
189  return opt
190  else:
191  raise Exception("any of "+','.join(listOfKeys)+" are mandatory entries of --output options")
192 
194  """The main building routines """
195 
196  def __init__(self, options, process = None, with_output = False, with_input = False ):
197  """options taken from old cmsDriver and optparse """
198 
199  options.outfile_name = options.dirout+options.fileout
200 
201  self._options = options
202 
203  if self._options.isData and options.isMC:
204  raise Exception("ERROR: You may specify only --data or --mc, not both")
205  #if not self._options.conditions:
206  # raise Exception("ERROR: No conditions given!\nPlease specify conditions. E.g. via --conditions=IDEAL_30X::All")
207 
208  # check that MEtoEDMConverter (running in ENDJOB) and DQMIO don't run in the same job
209  if 'ENDJOB' in self._options.step:
210  if (hasattr(self._options,"outputDefinition") and \
211  self._options.outputDefinition != '' and \
212  any(anyOf(['t','tier','dataTier'],outdic) == 'DQMIO' for outdic in eval(self._options.outputDefinition))) or \
213  (hasattr(self._options,"datatier") and \
214  self._options.datatier and \
215  'DQMIO' in self._options.datatier):
216  print("removing ENDJOB from steps since not compatible with DQMIO dataTier")
217  self._options.step=self._options.step.replace(',ENDJOB','')
218 
219 
220 
221  # what steps are provided by this class?
222  stepList = [re.sub(r'^prepare_', '', methodName) for methodName in ConfigBuilder.__dict__ if methodName.startswith('prepare_')]
223  self.stepMap={}
224  self.stepKeys=[]
225  for step in self._options.step.split(","):
226  if step=='': continue
227  stepParts = step.split(":")
228  stepName = stepParts[0]
229  if stepName not in stepList and not stepName.startswith('re'):
230  raise ValueError("Step "+stepName+" unknown")
231  if len(stepParts)==1:
232  self.stepMap[stepName]=""
233  elif len(stepParts)==2:
234  self.stepMap[stepName]=stepParts[1].split('+')
235  elif len(stepParts)==3:
236  self.stepMap[stepName]=(stepParts[2].split('+'),stepParts[1])
237  else:
238  raise ValueError("Step definition "+step+" invalid")
239  self.stepKeys.append(stepName)
240 
241  #print "map of steps is:",self.stepMap
242 
243  self.with_output = with_output
244  self.process=process
245 
246  if hasattr(self._options,"no_output_flag") and self._options.no_output_flag:
247  self.with_output = False
248  self.with_input = with_input
249  self.imports = []
250  self.create_process()
251  self.define_Configs()
252  self.schedule = list()
253 
254  # we are doing three things here:
255  # creating a process to catch errors
256  # building the code to re-create the process
257 
259  # TODO: maybe a list of to be dumped objects would help as well
260  self.blacklist_paths = []
261  self.addedObjects = []
263 
269 
270  def profileOptions(self):
271  """
272  addIgProfService
273  Function to add the igprof profile service so that you can dump in the middle
274  of the run.
275  """
276  profileOpts = self._options.profile.split(':')
277  profilerStart = 1
278  profilerInterval = 100
279  profilerFormat = None
280  profilerJobFormat = None
281 
282  if len(profileOpts):
283  #type, given as first argument is unused here
284  profileOpts.pop(0)
285  if len(profileOpts):
286  startEvent = profileOpts.pop(0)
287  if not startEvent.isdigit():
288  raise Exception("%s is not a number" % startEvent)
289  profilerStart = int(startEvent)
290  if len(profileOpts):
291  eventInterval = profileOpts.pop(0)
292  if not eventInterval.isdigit():
293  raise Exception("%s is not a number" % eventInterval)
294  profilerInterval = int(eventInterval)
295  if len(profileOpts):
296  profilerFormat = profileOpts.pop(0)
297 
298 
299  if not profilerFormat:
300  profilerFormat = "%s___%s___%s___%s___%s___%s___%%I.gz" % (self._options.evt_type.replace("_cfi", ""),
301  self._options.step,
302  self._options.pileup,
303  self._options.conditions,
304  self._options.datatier,
305  self._options.profileTypeLabel)
306  if not profilerJobFormat and profilerFormat.endswith(".gz"):
307  profilerJobFormat = profilerFormat.replace(".gz", "_EndOfJob.gz")
308  elif not profilerJobFormat:
309  profilerJobFormat = profilerFormat + "_EndOfJob.gz"
310 
311  return (profilerStart,profilerInterval,profilerFormat,profilerJobFormat)
312 
313  def load(self,includeFile):
314  includeFile = includeFile.replace('/','.')
315  self.process.load(includeFile)
316  return sys.modules[includeFile]
317 
318  def loadAndRemember(self, includeFile):
319  """helper routine to load am memorize imports"""
320  # we could make the imports a on-the-fly data method of the process instance itself
321  # not sure if the latter is a good idea
322  includeFile = includeFile.replace('/','.')
323  self.imports.append(includeFile)
324  self.process.load(includeFile)
325  return sys.modules[includeFile]
326 
327  def executeAndRemember(self, command):
328  """helper routine to remember replace statements"""
329  self.additionalCommands.append(command)
330  if not command.strip().startswith("#"):
331  # substitute: process.foo = process.bar -> self.process.foo = self.process.bar
332  import re
333  exec(re.sub(r"([^a-zA-Z_0-9]|^)(process)([^a-zA-Z_0-9])",r"\1self.process\3",command))
334  #exec(command.replace("process.","self.process."))
335 
336  def addCommon(self):
337  if 'HARVESTING' in self.stepMap.keys() or 'ALCAHARVEST' in self.stepMap.keys():
338  self.process.options = cms.untracked.PSet( Rethrow = cms.untracked.vstring('ProductNotFound'),fileMode = cms.untracked.string('FULLMERGE'))
339  else:
340  self.process.options = cms.untracked.PSet( )
341 
342  self.addedObjects.append(("","options"))
343 
344  if self._options.lazy_download:
345  self.process.AdaptorConfig = cms.Service("AdaptorConfig",
346  stats = cms.untracked.bool(True),
347  enable = cms.untracked.bool(True),
348  cacheHint = cms.untracked.string("lazy-download"),
349  readHint = cms.untracked.string("read-ahead-buffered")
350  )
351  self.addedObjects.append(("Setup lazy download","AdaptorConfig"))
352 
353  #self.process.cmsDriverCommand = cms.untracked.PSet( command=cms.untracked.string('cmsDriver.py '+self._options.arguments) )
354  #self.addedObjects.append(("what cmsDriver command was used","cmsDriverCommand"))
355 
356  if self._options.profile:
357  (start, interval, eventFormat, jobFormat)=self.profileOptions()
358  self.process.IgProfService = cms.Service("IgProfService",
359  reportFirstEvent = cms.untracked.int32(start),
360  reportEventInterval = cms.untracked.int32(interval),
361  reportToFileAtPostEvent = cms.untracked.string("| gzip -c > %s"%(eventFormat)),
362  reportToFileAtPostEndJob = cms.untracked.string("| gzip -c > %s"%(jobFormat)))
363  self.addedObjects.append(("Setup IGProf Service for profiling","IgProfService"))
364 
365  def addMaxEvents(self):
366  """Here we decide how many evts will be processed"""
367  self.process.maxEvents=cms.untracked.PSet(input=cms.untracked.int32(int(self._options.number)))
368  if self._options.number_out:
369  self.process.maxEvents.output = cms.untracked.int32(int(self._options.number_out))
370  self.addedObjects.append(("","maxEvents"))
371 
372  def addSource(self):
373  """Here the source is built. Priority: file, generator"""
374  self.addedObjects.append(("Input source","source"))
375 
376  def filesFromOption(self):
377  for entry in self._options.filein.split(','):
378  print("entry",entry)
379  if entry.startswith("filelist:"):
380  filesFromList(entry[9:],self.process.source)
381  elif entry.startswith("dbs:") or entry.startswith("das:"):
382  filesFromDASQuery('file dataset = %s'%(entry[4:]),self._options.dasoption,self.process.source)
383  else:
384  self.process.source.fileNames.append(self._options.dirin+entry)
385  if self._options.secondfilein:
386  if not hasattr(self.process.source,"secondaryFileNames"):
387  raise Exception("--secondfilein not compatible with "+self._options.filetype+"input type")
388  for entry in self._options.secondfilein.split(','):
389  print("entry",entry)
390  if entry.startswith("filelist:"):
391  self.process.source.secondaryFileNames.extend((filesFromList(entry[9:]))[0])
392  elif entry.startswith("dbs:") or entry.startswith("das:"):
393  self.process.source.secondaryFileNames.extend((filesFromDASQuery('file dataset = %s'%(entry[4:]),self._options.dasoption))[0])
394  else:
395  self.process.source.secondaryFileNames.append(self._options.dirin+entry)
396 
397  if self._options.filein or self._options.dasquery:
398  if self._options.filetype == "EDM":
399  self.process.source=cms.Source("PoolSource",
400  fileNames = cms.untracked.vstring(),
401  secondaryFileNames= cms.untracked.vstring())
402  filesFromOption(self)
403  elif self._options.filetype == "DAT":
404  self.process.source=cms.Source("NewEventStreamFileReader",fileNames = cms.untracked.vstring())
405  filesFromOption(self)
406  elif self._options.filetype == "LHE":
407  self.process.source=cms.Source("LHESource", fileNames = cms.untracked.vstring())
408  if self._options.filein.startswith("lhe:"):
409  #list the article directory automatically
410  args=self._options.filein.split(':')
411  article=args[1]
412  print('LHE input from article ',article)
413  location='/store/lhe/'
414  import os
415  textOfFiles=os.popen('cmsLHEtoEOSManager.py -l '+article)
416  for line in textOfFiles:
417  for fileName in [x for x in line.split() if '.lhe' in x]:
418  self.process.source.fileNames.append(location+article+'/'+fileName)
419  #check first if list of LHE files is loaded (not empty)
420  if len(line)<2:
421  print('Issue to load LHE files, please check and try again.')
422  sys.exit(-1)
423  #Additional check to protect empty fileNames in process.source
424  if len(self.process.source.fileNames)==0:
425  print('Issue with empty filename, but can pass line check')
426  sys.exit(-1)
427  if len(args)>2:
428  self.process.source.skipEvents = cms.untracked.uint32(int(args[2]))
429  else:
430  filesFromOption(self)
431 
432  elif self._options.filetype == "DQM":
433  self.process.source=cms.Source("DQMRootSource",
434  fileNames = cms.untracked.vstring())
435  filesFromOption(self)
436 
437  elif self._options.filetype == "DQMDAQ":
438  # FIXME: how to configure it if there are no input files specified?
439  self.process.source=cms.Source("DQMStreamerReader")
440 
441 
442  if ('HARVESTING' in self.stepMap.keys() or 'ALCAHARVEST' in self.stepMap.keys()) and (not self._options.filetype == "DQM"):
443  self.process.source.processingMode = cms.untracked.string("RunsAndLumis")
444 
445  if self._options.dasquery!='':
446  self.process.source=cms.Source("PoolSource", fileNames = cms.untracked.vstring(),secondaryFileNames = cms.untracked.vstring())
447  filesFromDASQuery(self._options.dasquery,self._options.dasoption,self.process.source)
448 
449  if ('HARVESTING' in self.stepMap.keys() or 'ALCAHARVEST' in self.stepMap.keys()) and (not self._options.filetype == "DQM"):
450  self.process.source.processingMode = cms.untracked.string("RunsAndLumis")
451 
452  ##drop LHEXMLStringProduct on input to save memory if appropriate
453  if 'GEN' in self.stepMap.keys():
454  if self._options.inputCommands:
455  self._options.inputCommands+=',drop LHEXMLStringProduct_*_*_*,'
456  else:
457  self._options.inputCommands='keep *, drop LHEXMLStringProduct_*_*_*,'
458 
459  if self.process.source and self._options.inputCommands:
460  if not hasattr(self.process.source,'inputCommands'): self.process.source.inputCommands=cms.untracked.vstring()
461  for command in self._options.inputCommands.split(','):
462  # remove whitespace around the keep/drop statements
463  command = command.strip()
464  if command=='': continue
465  self.process.source.inputCommands.append(command)
466  if not self._options.dropDescendant:
467  self.process.source.dropDescendantsOfDroppedBranches = cms.untracked.bool(False)
468 
469  if self._options.lumiToProcess:
470  import FWCore.PythonUtilities.LumiList as LumiList
471  self.process.source.lumisToProcess = cms.untracked.VLuminosityBlockRange( LumiList.LumiList(self._options.lumiToProcess).getCMSSWString().split(',') )
472 
473  if 'GEN' in self.stepMap.keys() or 'LHE' in self.stepMap or (not self._options.filein and hasattr(self._options, "evt_type")):
474  if self.process.source is None:
475  self.process.source=cms.Source("EmptySource")
476 
477  # modify source in case of run-dependent MC
478  self.runsAndWeights=None
479  if self._options.runsAndWeightsForMC or self._options.runsScenarioForMC :
480  if not self._options.isMC :
481  raise Exception("options --runsAndWeightsForMC and --runsScenarioForMC are only valid for MC")
482  if self._options.runsAndWeightsForMC:
483  self.runsAndWeights = eval(self._options.runsAndWeightsForMC)
484  else:
485  from Configuration.StandardSequences.RunsAndWeights import RunsAndWeights
486  if isinstance(RunsAndWeights[self._options.runsScenarioForMC], str):
487  __import__(RunsAndWeights[self._options.runsScenarioForMC])
488  self.runsAndWeights = sys.modules[RunsAndWeights[self._options.runsScenarioForMC]].runProbabilityDistribution
489  else:
490  self.runsAndWeights = RunsAndWeights[self._options.runsScenarioForMC]
491 
492  if self.runsAndWeights:
493  import SimGeneral.Configuration.ThrowAndSetRandomRun as ThrowAndSetRandomRun
495  self.additionalCommands.append('import SimGeneral.Configuration.ThrowAndSetRandomRun as ThrowAndSetRandomRun')
496  self.additionalCommands.append('ThrowAndSetRandomRun.throwAndSetRandomRun(process.source,%s)'%(self.runsAndWeights))
497 
498  return
499 
500  def addOutput(self):
501  """ Add output module to the process """
502  result=""
503  if self._options.outputDefinition:
504  if self._options.datatier:
505  print("--datatier & --eventcontent options ignored")
506 
507  #new output convention with a list of dict
508  outList = eval(self._options.outputDefinition)
509  for (id,outDefDict) in enumerate(outList):
510  outDefDictStr=outDefDict.__str__()
511  if not isinstance(outDefDict,dict):
512  raise Exception("--output needs to be passed a list of dict"+self._options.outputDefinition+" is invalid")
513  #requires option: tier
514  theTier=anyOf(['t','tier','dataTier'],outDefDict)
515  #optional option: eventcontent, filtername, selectEvents, moduleLabel, filename
516  ## event content
517  theStreamType=anyOf(['e','ec','eventContent','streamType'],outDefDict,theTier)
518  theFilterName=anyOf(['f','ftN','filterName'],outDefDict,'')
519  theSelectEvent=anyOf(['s','sE','selectEvents'],outDefDict,'')
520  theModuleLabel=anyOf(['l','mL','moduleLabel'],outDefDict,'')
521  theExtraOutputCommands=anyOf(['o','oC','outputCommands'],outDefDict,'')
522  # module label has a particular role
523  if not theModuleLabel:
524  tryNames=[theStreamType.replace(theTier.replace('-',''),'')+theTier.replace('-','')+'output',
525  theStreamType.replace(theTier.replace('-',''),'')+theTier.replace('-','')+theFilterName+'output',
526  theStreamType.replace(theTier.replace('-',''),'')+theTier.replace('-','')+theFilterName+theSelectEvent.split(',')[0].replace(':','for').replace(' ','')+'output'
527  ]
528  for name in tryNames:
529  if not hasattr(self.process,name):
530  theModuleLabel=name
531  break
532  if not theModuleLabel:
533  raise Exception("cannot find a module label for specification: "+outDefDictStr)
534  if id==0:
535  defaultFileName=self._options.outfile_name
536  else:
537  defaultFileName=self._options.outfile_name.replace('.root','_in'+theTier+'.root')
538 
539  theFileName=self._options.dirout+anyOf(['fn','fileName'],outDefDict,defaultFileName)
540  if not theFileName.endswith('.root'):
541  theFileName+='.root'
542 
543  if len(outDefDict):
544  raise Exception("unused keys from --output options: "+','.join(outDefDict.keys()))
545  if theStreamType=='DQMIO': theStreamType='DQM'
546  if theStreamType=='ALL':
547  theEventContent = cms.PSet(outputCommands = cms.untracked.vstring('keep *'))
548  else:
549  theEventContent = getattr(self.process, theStreamType+"EventContent")
550 
551 
552  addAlCaSelects=False
553  if theStreamType=='ALCARECO' and not theFilterName:
554  theFilterName='StreamALCACombined'
555  addAlCaSelects=True
556 
557  CppType='PoolOutputModule'
558  if self._options.timeoutOutput:
559  CppType='TimeoutPoolOutputModule'
560  if theStreamType=='DQM' and theTier=='DQMIO': CppType='DQMRootOutputModule'
561  output = cms.OutputModule(CppType,
562  theEventContent.clone(),
563  fileName = cms.untracked.string(theFileName),
564  dataset = cms.untracked.PSet(
565  dataTier = cms.untracked.string(theTier),
566  filterName = cms.untracked.string(theFilterName))
567  )
568  if not theSelectEvent and hasattr(self.process,'generation_step') and theStreamType!='LHE':
569  output.SelectEvents = cms.untracked.PSet(SelectEvents = cms.vstring('generation_step'))
570  if not theSelectEvent and hasattr(self.process,'filtering_step'):
571  output.SelectEvents = cms.untracked.PSet(SelectEvents = cms.vstring('filtering_step'))
572  if theSelectEvent:
573  output.SelectEvents =cms.untracked.PSet(SelectEvents = cms.vstring(theSelectEvent))
574 
575  if addAlCaSelects:
576  if not hasattr(output,'SelectEvents'):
577  output.SelectEvents=cms.untracked.PSet(SelectEvents=cms.vstring())
578  for alca in self.AlCaPaths:
579  output.SelectEvents.SelectEvents.extend(getattr(self.process,'OutALCARECO'+alca).SelectEvents.SelectEvents)
580 
581 
582  if hasattr(self.process,theModuleLabel):
583  raise Exception("the current process already has a module "+theModuleLabel+" defined")
584  #print "creating output module ",theModuleLabel
585  setattr(self.process,theModuleLabel,output)
586  outputModule=getattr(self.process,theModuleLabel)
587  setattr(self.process,theModuleLabel+'_step',cms.EndPath(outputModule))
588  path=getattr(self.process,theModuleLabel+'_step')
589  self.schedule.append(path)
590 
591  if not self._options.inlineEventContent and hasattr(self.process,theStreamType+"EventContent"):
592  def doNotInlineEventContent(instance,label = "cms.untracked.vstring(process."+theStreamType+"EventContent.outputCommands)"):
593  return label
594  outputModule.outputCommands.__dict__["dumpPython"] = doNotInlineEventContent
595  if theExtraOutputCommands:
596  if not isinstance(theExtraOutputCommands,list):
597  raise Exception("extra ouput command in --option must be a list of strings")
598  if hasattr(self.process,theStreamType+"EventContent"):
599  self.executeAndRemember('process.%s.outputCommands.extend(%s)'%(theModuleLabel,theExtraOutputCommands))
600  else:
601  outputModule.outputCommands.extend(theExtraOutputCommands)
602 
603  result+="\nprocess."+theModuleLabel+" = "+outputModule.dumpPython()
604 
605  ##ends the --output options model
606  return result
607 
608  streamTypes=self._options.eventcontent.split(',')
609  tiers=self._options.datatier.split(',')
610  if not self._options.outputDefinition and len(streamTypes)!=len(tiers):
611  raise Exception("number of event content arguments does not match number of datatier arguments")
612 
613  # if the only step is alca we don't need to put in an output
614  if self._options.step.split(',')[0].split(':')[0] == 'ALCA':
615  return "\n"
616 
617  for i,(streamType,tier) in enumerate(zip(streamTypes,tiers)):
618  if streamType=='': continue
619  if streamType == 'ALCARECO' and not 'ALCAPRODUCER' in self._options.step: continue
620  if streamType=='DQMIO': streamType='DQM'
621  eventContent=streamType
622  ## override streamType to eventContent in case NANOEDM
623  if streamType == "NANOEDMAOD" :
624  eventContent = "NANOAOD"
625  elif streamType == "NANOEDMAODSIM" :
626  eventContent = "NANOAODSIM"
627  theEventContent = getattr(self.process, eventContent+"EventContent")
628  if i==0:
629  theFileName=self._options.outfile_name
630  theFilterName=self._options.filtername
631  else:
632  theFileName=self._options.outfile_name.replace('.root','_in'+streamType+'.root')
633  theFilterName=self._options.filtername
634  CppType='PoolOutputModule'
635  if self._options.timeoutOutput:
636  CppType='TimeoutPoolOutputModule'
637  if streamType=='DQM' and tier=='DQMIO': CppType='DQMRootOutputModule'
638  if "NANOAOD" in streamType : CppType='NanoAODOutputModule'
639  output = cms.OutputModule(CppType,
640  theEventContent,
641  fileName = cms.untracked.string(theFileName),
642  dataset = cms.untracked.PSet(dataTier = cms.untracked.string(tier),
643  filterName = cms.untracked.string(theFilterName)
644  )
645  )
646  if hasattr(self.process,"generation_step") and streamType!='LHE':
647  output.SelectEvents = cms.untracked.PSet(SelectEvents = cms.vstring('generation_step'))
648  if hasattr(self.process,"filtering_step"):
649  output.SelectEvents = cms.untracked.PSet(SelectEvents = cms.vstring('filtering_step'))
650 
651  if streamType=='ALCARECO':
652  output.dataset.filterName = cms.untracked.string('StreamALCACombined')
653 
654  if "MINIAOD" in streamType:
655  from PhysicsTools.PatAlgos.slimming.miniAOD_tools import miniAOD_customizeOutput
657 
658  outputModuleName=streamType+'output'
659  setattr(self.process,outputModuleName,output)
660  outputModule=getattr(self.process,outputModuleName)
661  setattr(self.process,outputModuleName+'_step',cms.EndPath(outputModule))
662  path=getattr(self.process,outputModuleName+'_step')
663  self.schedule.append(path)
664 
665  if self._options.outputCommands and streamType!='DQM':
666  for evct in self._options.outputCommands.split(','):
667  if not evct: continue
668  self.executeAndRemember("process.%s.outputCommands.append('%s')"%(outputModuleName,evct.strip()))
669 
670  if not self._options.inlineEventContent:
671  tmpstreamType=streamType
672  if "NANOEDM" in tmpstreamType :
673  tmpstreamType=tmpstreamType.replace("NANOEDM","NANO")
674  def doNotInlineEventContent(instance,label = "process."+tmpstreamType+"EventContent.outputCommands"):
675  return label
676  outputModule.outputCommands.__dict__["dumpPython"] = doNotInlineEventContent
677 
678  result+="\nprocess."+outputModuleName+" = "+outputModule.dumpPython()
679 
680  return result
681 
683  """
684  Add selected standard sequences to the process
685  """
686  # load the pile up file
687  if self._options.pileup:
688  pileupSpec=self._options.pileup.split(',')[0]
689 
690  # Does the requested pile-up scenario exist?
691  from Configuration.StandardSequences.Mixing import Mixing,defineMixing
692  if not pileupSpec in Mixing and '.' not in pileupSpec and 'file:' not in pileupSpec:
693  message = pileupSpec+' is not a know mixing scenario:\n available are: '+'\n'.join(Mixing.keys())
694  raise Exception(message)
695 
696  # Put mixing parameters in a dictionary
697  if '.' in pileupSpec:
698  mixingDict={'file':pileupSpec}
699  elif pileupSpec.startswith('file:'):
700  mixingDict={'file':pileupSpec[5:]}
701  else:
702  import copy
703  mixingDict=copy.copy(Mixing[pileupSpec])
704  if len(self._options.pileup.split(','))>1:
705  mixingDict.update(eval(self._options.pileup[self._options.pileup.find(',')+1:]))
706 
707  # Load the pu cfg file corresponding to the requested pu scenario
708  if 'file:' in pileupSpec:
709  #the file is local
710  self.process.load(mixingDict['file'])
711  print("inlining mixing module configuration")
712  self._options.inlineObjets+=',mix'
713  else:
714  self.loadAndRemember(mixingDict['file'])
715 
716  mixingDict.pop('file')
717  if not "DATAMIX" in self.stepMap.keys(): # when DATAMIX is present, pileup_input refers to pre-mixed GEN-RAW
718  if self._options.pileup_input:
719  if self._options.pileup_input.startswith('dbs:') or self._options.pileup_input.startswith('das:'):
720  mixingDict['F']=filesFromDASQuery('file dataset = %s'%(self._options.pileup_input[4:],),self._options.pileup_dasoption)[0]
721  elif self._options.pileup_input.startswith("filelist:"):
722  mixingDict['F']=(filesFromList(self._options.pileup_input[9:]))[0]
723  else:
724  mixingDict['F']=self._options.pileup_input.split(',')
725  specialization=defineMixing(mixingDict)
726  for command in specialization:
727  self.executeAndRemember(command)
728  if len(mixingDict)!=0:
729  raise Exception('unused mixing specification: '+mixingDict.keys().__str__())
730 
731 
732  # load the geometry file
733  try:
734  if len(self.stepMap):
735  self.loadAndRemember(self.GeometryCFF)
736  if ('SIM' in self.stepMap or 'reSIM' in self.stepMap) and not self._options.fast:
738  if self.geometryDBLabel:
739  self.executeAndRemember('process.XMLFromDBSource.label = cms.string("%s")'%(self.geometryDBLabel))
740  except ImportError:
741  print("Geometry option",self._options.geometry,"unknown.")
742  raise
743 
744  if len(self.stepMap):
745  self.loadAndRemember(self.magFieldCFF)
746 
747  for stepName in self.stepKeys:
748  stepSpec = self.stepMap[stepName]
749  print("Step:", stepName,"Spec:",stepSpec)
750  if stepName.startswith('re'):
751  ##add the corresponding input content
752  if stepName[2:] not in self._options.donotDropOnInput:
753  self._options.inputEventContent='%s,%s'%(stepName.upper(),self._options.inputEventContent)
754  stepName=stepName[2:]
755  if stepSpec=="":
756  getattr(self,"prepare_"+stepName)(sequence = getattr(self,stepName+"DefaultSeq"))
757  elif isinstance(stepSpec, list):
758  getattr(self,"prepare_"+stepName)(sequence = '+'.join(stepSpec))
759  elif isinstance(stepSpec, tuple):
760  getattr(self,"prepare_"+stepName)(sequence = ','.join([stepSpec[1],'+'.join(stepSpec[0])]))
761  else:
762  raise ValueError("Invalid step definition")
763 
764  if self._options.restoreRNDSeeds!=False:
765  #it is either True, or a process name
766  if self._options.restoreRNDSeeds==True:
767  self.executeAndRemember('process.RandomNumberGeneratorService.restoreStateLabel=cms.untracked.string("randomEngineStateProducer")')
768  else:
769  self.executeAndRemember('process.RandomNumberGeneratorService.restoreStateTag=cms.untracked.InputTag("randomEngineStateProducer","","%s")'%(self._options.restoreRNDSeeds))
770  if self._options.inputEventContent or self._options.inputCommands:
771  if self._options.inputCommands:
772  self._options.inputCommands+='keep *_randomEngineStateProducer_*_*,'
773  else:
774  self._options.inputCommands='keep *_randomEngineStateProducer_*_*,'
775 
776 
778  if self._options.inputEventContent:
779  import copy
780  def dropSecondDropStar(iec):
781  #drop occurence of 'drop *' in the list
782  count=0
783  for item in iec:
784  if item=='drop *':
785  if count!=0:
786  iec.remove(item)
787  count+=1
788 
789 
790  ## allow comma separated input eventcontent
791  if not hasattr(self.process.source,'inputCommands'): self.process.source.inputCommands=cms.untracked.vstring()
792  for evct in self._options.inputEventContent.split(','):
793  if evct=='': continue
794  theEventContent = getattr(self.process, evct+"EventContent")
795  if hasattr(theEventContent,'outputCommands'):
796  self.process.source.inputCommands.extend(copy.copy(theEventContent.outputCommands))
797  if hasattr(theEventContent,'inputCommands'):
798  self.process.source.inputCommands.extend(copy.copy(theEventContent.inputCommands))
799 
800  dropSecondDropStar(self.process.source.inputCommands)
801 
802  if not self._options.dropDescendant:
803  self.process.source.dropDescendantsOfDroppedBranches = cms.untracked.bool(False)
804 
805 
806  return
807 
808  def addConditions(self):
809  """Add conditions to the process"""
810  if not self._options.conditions: return
811 
812  if 'FrontierConditions_GlobalTag' in self._options.conditions:
813  print('using FrontierConditions_GlobalTag in --conditions is not necessary anymore and will be deprecated soon. please update your command line')
814  self._options.conditions = self._options.conditions.replace("FrontierConditions_GlobalTag,",'')
815 
817  from Configuration.AlCa.GlobalTag import GlobalTag
818  self.process.GlobalTag = GlobalTag(self.process.GlobalTag, self._options.conditions, self._options.custom_conditions)
819  self.additionalCommands.append('from Configuration.AlCa.GlobalTag import GlobalTag')
820  self.additionalCommands.append('process.GlobalTag = GlobalTag(process.GlobalTag, %s, %s)' % (repr(self._options.conditions), repr(self._options.custom_conditions)))
821 
822 
823  def addCustomise(self,unsch=0):
824  """Include the customise code """
825 
826  custOpt=[]
827  if unsch==0:
828  for c in self._options.customisation_file:
829  custOpt.extend(c.split(","))
830  else:
831  for c in self._options.customisation_file_unsch:
832  custOpt.extend(c.split(","))
833 
834  custMap=DictTypes.SortedKeysDict()
835  for opt in custOpt:
836  if opt=='': continue
837  if opt.count('.')>1:
838  raise Exception("more than . in the specification:"+opt)
839  fileName=opt.split('.')[0]
840  if opt.count('.')==0: rest='customise'
841  else:
842  rest=opt.split('.')[1]
843  if rest=='py': rest='customise' #catch the case of --customise file.py
844 
845  if fileName in custMap:
846  custMap[fileName].extend(rest.split('+'))
847  else:
848  custMap[fileName]=rest.split('+')
849 
850  if len(custMap)==0:
851  final_snippet='\n'
852  else:
853  final_snippet='\n# customisation of the process.\n'
854 
855  allFcn=[]
856  for opt in custMap:
857  allFcn.extend(custMap[opt])
858  for fcn in allFcn:
859  if allFcn.count(fcn)!=1:
860  raise Exception("cannot specify twice "+fcn+" as a customisation method")
861 
862  for f in custMap:
863  # let python search for that package and do syntax checking at the same time
864  packageName = f.replace(".py","").replace("/",".")
865  __import__(packageName)
866  package = sys.modules[packageName]
867 
868  # now ask the package for its definition and pick .py instead of .pyc
869  customiseFile = re.sub(r'\.pyc$', '.py', package.__file__)
870 
871  final_snippet+='\n# Automatic addition of the customisation function from '+packageName+'\n'
872  if self._options.inline_custom:
873  for line in file(customiseFile,'r'):
874  if "import FWCore.ParameterSet.Config" in line:
875  continue
876  final_snippet += line
877  else:
878  final_snippet += 'from %s import %s \n'%(packageName,','.join(custMap[f]))
879  for fcn in custMap[f]:
880  print("customising the process with",fcn,"from",f)
881  if not hasattr(package,fcn):
882  #bound to fail at run time
883  raise Exception("config "+f+" has no function "+fcn)
884  #execute the command
885  self.process=getattr(package,fcn)(self.process)
886  #and print it in the configuration
887  final_snippet += "\n#call to customisation function "+fcn+" imported from "+packageName
888  final_snippet += "\nprocess = %s(process)\n"%(fcn,)
889 
890  if len(custMap)!=0:
891  final_snippet += '\n# End of customisation functions\n'
892 
893  ### now for a useful command
894  return final_snippet
895 
897  final_snippet='\n# Customisation from command line\n'
898  if self._options.customise_commands:
899  import string
900  for com in self._options.customise_commands.split('\\n'):
901  com=string.lstrip(com)
902  self.executeAndRemember(com)
903  final_snippet +='\n'+com
904 
905  return final_snippet
906 
907  #----------------------------------------------------------------------------
908  # here the methods to define the python includes for each step or
909  # conditions
910  #----------------------------------------------------------------------------
911  def define_Configs(self):
912  if len(self.stepMap):
913  self.loadAndRemember('Configuration/StandardSequences/Services_cff')
914  if self._options.particleTable not in defaultOptions.particleTableList:
915  print('Invalid particle table provided. Options are:')
916  print(defaultOptions.particleTable)
917  sys.exit(-1)
918  else:
919  if len(self.stepMap):
920  self.loadAndRemember('SimGeneral.HepPDTESSource.'+self._options.particleTable+'_cfi')
921 
922  self.loadAndRemember('FWCore/MessageService/MessageLogger_cfi')
923 
924  self.ALCADefaultCFF="Configuration/StandardSequences/AlCaRecoStreams_cff"
925  self.GENDefaultCFF="Configuration/StandardSequences/Generator_cff"
926  self.SIMDefaultCFF="Configuration/StandardSequences/Sim_cff"
927  self.DIGIDefaultCFF="Configuration/StandardSequences/Digi_cff"
928  self.DIGI2RAWDefaultCFF="Configuration/StandardSequences/DigiToRaw_cff"
929  self.L1EMDefaultCFF='Configuration/StandardSequences/SimL1Emulator_cff'
930  self.L1MENUDefaultCFF="Configuration/StandardSequences/L1TriggerDefaultMenu_cff"
931  self.HLTDefaultCFF="Configuration/StandardSequences/HLTtable_cff"
932  self.RAW2DIGIDefaultCFF="Configuration/StandardSequences/RawToDigi_Data_cff"
933  if self._options.isRepacked: self.RAW2DIGIDefaultCFF="Configuration/StandardSequences/RawToDigi_DataMapper_cff"
934  self.L1RecoDefaultCFF="Configuration/StandardSequences/L1Reco_cff"
935  self.L1TrackTriggerDefaultCFF="Configuration/StandardSequences/L1TrackTrigger_cff"
936  self.RECODefaultCFF="Configuration/StandardSequences/Reconstruction_Data_cff"
937  self.RECOSIMDefaultCFF="Configuration/StandardSequences/RecoSim_cff"
938  self.PATDefaultCFF="Configuration/StandardSequences/PAT_cff"
939  self.NANODefaultCFF="PhysicsTools/NanoAOD/nano_cff"
940  self.EIDefaultCFF=None
941  self.SKIMDefaultCFF="Configuration/StandardSequences/Skims_cff"
942  self.POSTRECODefaultCFF="Configuration/StandardSequences/PostRecoGenerator_cff"
943  self.VALIDATIONDefaultCFF="Configuration/StandardSequences/Validation_cff"
944  self.L1HwValDefaultCFF = "Configuration/StandardSequences/L1HwVal_cff"
945  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOffline_cff"
946  self.HARVESTINGDefaultCFF="Configuration/StandardSequences/Harvesting_cff"
947  self.ALCAHARVESTDefaultCFF="Configuration/StandardSequences/AlCaHarvesting_cff"
948  self.ENDJOBDefaultCFF="Configuration/StandardSequences/EndOfProcess_cff"
949  self.ConditionsDefaultCFF = "Configuration/StandardSequences/FrontierConditions_GlobalTag_cff"
950  self.CFWRITERDefaultCFF = "Configuration/StandardSequences/CrossingFrameWriter_cff"
951  self.REPACKDefaultCFF="Configuration/StandardSequences/DigiToRaw_Repack_cff"
952 
953  if "DATAMIX" in self.stepMap.keys():
954  self.DATAMIXDefaultCFF="Configuration/StandardSequences/DataMixer"+self._options.datamix+"_cff"
955  self.DIGIDefaultCFF="Configuration/StandardSequences/DigiDM_cff"
956  self.DIGI2RAWDefaultCFF="Configuration/StandardSequences/DigiToRawDM_cff"
957  self.L1EMDefaultCFF='Configuration/StandardSequences/SimL1EmulatorDM_cff'
958 
959  self.ALCADefaultSeq=None
960  self.LHEDefaultSeq='externalLHEProducer'
961  self.GENDefaultSeq='pgen'
962  self.SIMDefaultSeq='psim'
963  self.DIGIDefaultSeq='pdigi'
965  self.DIGI2RAWDefaultSeq='DigiToRaw'
966  self.HLTDefaultSeq='GRun'
967  self.L1DefaultSeq=None
972  self.RAW2DIGIDefaultSeq='RawToDigi'
973  self.L1RecoDefaultSeq='L1Reco'
974  self.L1TrackTriggerDefaultSeq='L1TrackTrigger'
975  if self._options.fast or ('RAW2DIGI' in self.stepMap and 'RECO' in self.stepMap):
976  self.RECODefaultSeq='reconstruction'
977  else:
978  self.RECODefaultSeq='reconstruction_fromRECO'
979  self.RECOSIMDefaultSeq='recosim'
980  self.EIDefaultSeq='top'
982  self.L1HwValDefaultSeq='L1HwVal'
983  self.DQMDefaultSeq='DQMOffline'
985  self.ENDJOBDefaultSeq='endOfProcess'
986  self.REPACKDefaultSeq='DigiToRawRepack'
987  self.PATDefaultSeq='miniAOD'
988  self.PATGENDefaultSeq='miniGEN'
989  self.NANODefaultSeq='nanoSequence'
990 
991  self.EVTCONTDefaultCFF="Configuration/EventContent/EventContent_cff"
992 
993  if not self._options.beamspot:
994  self._options.beamspot=VtxSmearedDefaultKey
995 
996  # if its MC then change the raw2digi
997  if self._options.isMC==True:
998  self.RAW2DIGIDefaultCFF="Configuration/StandardSequences/RawToDigi_cff"
999  self.RECODefaultCFF="Configuration/StandardSequences/Reconstruction_cff"
1000  self.PATDefaultCFF="Configuration/StandardSequences/PATMC_cff"
1001  self.PATGENDefaultCFF="Configuration/StandardSequences/PATGEN_cff"
1002  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOfflineMC_cff"
1003  self.ALCADefaultCFF="Configuration/StandardSequences/AlCaRecoStreamsMC_cff"
1004  self.NANODefaultSeq='nanoSequenceMC'
1005  else:
1006  self._options.beamspot = None
1007 
1008  #patch for gen, due to backward incompatibility
1009  if 'reGEN' in self.stepMap:
1010  self.GENDefaultSeq='fixGenInfo'
1011 
1012  if self._options.scenario=='cosmics':
1013  self._options.pileup='Cosmics'
1014  self.DIGIDefaultCFF="Configuration/StandardSequences/DigiCosmics_cff"
1015  self.RECODefaultCFF="Configuration/StandardSequences/ReconstructionCosmics_cff"
1016  self.SKIMDefaultCFF="Configuration/StandardSequences/SkimsCosmics_cff"
1017  self.EVTCONTDefaultCFF="Configuration/EventContent/EventContentCosmics_cff"
1018  self.VALIDATIONDefaultCFF="Configuration/StandardSequences/ValidationCosmics_cff"
1019  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOfflineCosmics_cff"
1020  if self._options.isMC==True:
1021  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOfflineCosmicsMC_cff"
1022  self.HARVESTINGDefaultCFF="Configuration/StandardSequences/HarvestingCosmics_cff"
1023  self.RECODefaultSeq='reconstructionCosmics'
1024  self.DQMDefaultSeq='DQMOfflineCosmics'
1025 
1026  if self._options.scenario=='HeavyIons':
1027  if not self._options.beamspot:
1028  self._options.beamspot=VtxSmearedHIDefaultKey
1029  self.HLTDefaultSeq = 'HIon'
1030  self.VALIDATIONDefaultCFF="Configuration/StandardSequences/ValidationHeavyIons_cff"
1031  self.VALIDATIONDefaultSeq=''
1032  self.EVTCONTDefaultCFF="Configuration/EventContent/EventContentHeavyIons_cff"
1033  self.RECODefaultCFF="Configuration/StandardSequences/ReconstructionHeavyIons_cff"
1034  self.RECODefaultSeq='reconstructionHeavyIons'
1035  self.ALCADefaultCFF = "Configuration/StandardSequences/AlCaRecoStreamsHeavyIons_cff"
1036  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOfflineHeavyIons_cff"
1037  self.DQMDefaultSeq='DQMOfflineHeavyIons'
1038  self.SKIMDefaultCFF="Configuration/StandardSequences/SkimsHeavyIons_cff"
1039  self.HARVESTINGDefaultCFF="Configuration/StandardSequences/HarvestingHeavyIons_cff"
1040  if self._options.isMC==True:
1041  self.DQMOFFLINEDefaultCFF="DQMOffline/Configuration/DQMOfflineHeavyIonsMC_cff"
1042 
1043 
1045 
1046  self.USERDefaultSeq='user'
1047  self.USERDefaultCFF=None
1048 
1049  # the magnetic field
1050  if self._options.isData:
1051  if self._options.magField==defaultOptions.magField:
1052  print("magnetic field option forced to: AutoFromDBCurrent")
1053  self._options.magField='AutoFromDBCurrent'
1054  self.magFieldCFF = 'Configuration/StandardSequences/MagneticField_'+self._options.magField.replace('.','')+'_cff'
1055  self.magFieldCFF = self.magFieldCFF.replace("__",'_')
1056 
1057  # the geometry
1058  self.GeometryCFF='Configuration/StandardSequences/GeometryRecoDB_cff'
1060  simGeometry=''
1061  if self._options.fast:
1062  if 'start' in self._options.conditions.lower():
1063  self.GeometryCFF='FastSimulation/Configuration/Geometries_START_cff'
1064  else:
1065  self.GeometryCFF='FastSimulation/Configuration/Geometries_MC_cff'
1066  else:
1067  def inGeometryKeys(opt):
1068  from Configuration.StandardSequences.GeometryConf import GeometryConf
1069  if opt in GeometryConf:
1070  return GeometryConf[opt]
1071  else:
1072  return opt
1073 
1074  geoms=self._options.geometry.split(',')
1075  if len(geoms)==1: geoms=inGeometryKeys(geoms[0]).split(',')
1076  if len(geoms)==2:
1077  #may specify the reco geometry
1078  if '/' in geoms[1] or '_cff' in geoms[1]:
1079  self.GeometryCFF=geoms[1]
1080  else:
1081  self.GeometryCFF='Configuration/Geometry/Geometry'+geoms[1]+'_cff'
1082 
1083  if (geoms[0].startswith('DB:')):
1084  self.SimGeometryCFF='Configuration/StandardSequences/GeometrySimDB_cff'
1085  self.geometryDBLabel=geoms[0][3:]
1086  print("with DB:")
1087  else:
1088  if '/' in geoms[0] or '_cff' in geoms[0]:
1089  self.SimGeometryCFF=geoms[0]
1090  else:
1091  simGeometry=geoms[0]
1092  if self._options.gflash==True:
1093  self.SimGeometryCFF='Configuration/Geometry/Geometry'+geoms[0]+'GFlash_cff'
1094  else:
1095  self.SimGeometryCFF='Configuration/Geometry/Geometry'+geoms[0]+'_cff'
1096 
1097  # synchronize the geometry configuration and the FullSimulation sequence to be used
1098  if simGeometry not in defaultOptions.geometryExtendedOptions:
1099  self.SIMDefaultCFF="Configuration/StandardSequences/SimIdeal_cff"
1100 
1101  if self._options.scenario=='nocoll' or self._options.scenario=='cosmics':
1102  self.SIMDefaultCFF="Configuration/StandardSequences/SimNOBEAM_cff"
1103  self._options.beamspot='NoSmear'
1104 
1105  # fastsim requires some changes to the default cff files and sequences
1106  if self._options.fast:
1107  self.SIMDefaultCFF = 'FastSimulation.Configuration.SimIdeal_cff'
1108  self.RECODefaultCFF= 'FastSimulation.Configuration.Reconstruction_AftMix_cff'
1109  self.RECOBEFMIXDefaultCFF = 'FastSimulation.Configuration.Reconstruction_BefMix_cff'
1110  self.RECOBEFMIXDefaultSeq = 'reconstruction_befmix'
1111  self.NANODefaultSeq = 'nanoSequenceFS'
1112  self.DQMOFFLINEDefaultCFF="FastSimulation.Configuration.DQMOfflineMC_cff"
1113 
1114  # Mixing
1115  if self._options.pileup=='default':
1116  from Configuration.StandardSequences.Mixing import MixingDefaultKey
1117  self._options.pileup=MixingDefaultKey
1118 
1119 
1120  #not driven by a default cff anymore
1121  if self._options.isData:
1122  self._options.pileup=None
1123 
1124 
1126 
1127  # for alca, skims, etc
1128  def addExtraStream(self,name,stream,workflow='full'):
1129  # define output module and go from there
1130  output = cms.OutputModule("PoolOutputModule")
1131  if stream.selectEvents.parameters_().__len__()!=0:
1132  output.SelectEvents = stream.selectEvents
1133  else:
1134  output.SelectEvents = cms.untracked.PSet()
1135  output.SelectEvents.SelectEvents=cms.vstring()
1136  if isinstance(stream.paths,tuple):
1137  for path in stream.paths:
1138  output.SelectEvents.SelectEvents.append(path.label())
1139  else:
1140  output.SelectEvents.SelectEvents.append(stream.paths.label())
1141 
1142 
1143 
1144  if isinstance(stream.content,str):
1145  evtPset=getattr(self.process,stream.content)
1146  for p in evtPset.parameters_():
1147  setattr(output,p,getattr(evtPset,p))
1148  if not self._options.inlineEventContent:
1149  def doNotInlineEventContent(instance,label = "process."+stream.content+".outputCommands"):
1150  return label
1151  output.outputCommands.__dict__["dumpPython"] = doNotInlineEventContent
1152  else:
1153  output.outputCommands = stream.content
1154 
1155 
1156  output.fileName = cms.untracked.string(self._options.dirout+stream.name+'.root')
1157 
1158  output.dataset = cms.untracked.PSet( dataTier = stream.dataTier,
1159  filterName = cms.untracked.string(stream.name))
1160 
1161  if self._options.filtername:
1162  output.dataset.filterName= cms.untracked.string(self._options.filtername+"_"+stream.name)
1163 
1164  #add an automatic flushing to limit memory consumption
1165  output.eventAutoFlushCompressedSize=cms.untracked.int32(5*1024*1024)
1166 
1167  if workflow in ("producers,full"):
1168  if isinstance(stream.paths,tuple):
1169  for path in stream.paths:
1170  self.schedule.append(path)
1171  else:
1172  self.schedule.append(stream.paths)
1173 
1174 
1175  # in case of relvals we don't want to have additional outputs
1176  if (not self._options.relval) and workflow in ("full","output"):
1177  self.additionalOutputs[name] = output
1178  setattr(self.process,name,output)
1179 
1180  if workflow == 'output':
1181  # adjust the select events to the proper trigger results from previous process
1182  filterList = output.SelectEvents.SelectEvents
1183  for i, filter in enumerate(filterList):
1184  filterList[i] = filter+":"+self._options.triggerResultsProcess
1185 
1186  return output
1187 
1188  #----------------------------------------------------------------------------
1189  # here the methods to create the steps. Of course we are doing magic here ;)
1190  # prepare_STEPNAME modifies self.process and what else's needed.
1191  #----------------------------------------------------------------------------
1192 
1193  def loadDefaultOrSpecifiedCFF(self, sequence,defaultCFF):
1194  if ( len(sequence.split('.'))==1 ):
1195  l=self.loadAndRemember(defaultCFF)
1196  elif ( len(sequence.split('.'))==2 ):
1197  l=self.loadAndRemember(sequence.split('.')[0])
1198  sequence=sequence.split('.')[1]
1199  else:
1200  print("sub sequence configuration must be of the form dir/subdir/cff.a+b+c or cff.a")
1201  print(sequence,"not recognized")
1202  raise
1203  return l
1204 
1205  def scheduleSequence(self,seq,prefix,what='Path'):
1206  if '*' in seq:
1207  #create only one path with all sequences in it
1208  for i,s in enumerate(seq.split('*')):
1209  if i==0:
1210  setattr(self.process,prefix,getattr(cms,what)( getattr(self.process, s) ))
1211  else:
1212  p=getattr(self.process,prefix)
1213  tmp = getattr(self.process, s)
1214  if isinstance(tmp, cms.Task):
1215  p.associate(tmp)
1216  else:
1217  p+=tmp
1218  self.schedule.append(getattr(self.process,prefix))
1219  return
1220  else:
1221  #create as many path as many sequences
1222  if not '+' in seq:
1223  if self.nextScheduleIsConditional:
1224  self.conditionalPaths.append(prefix)
1225  setattr(self.process,prefix,getattr(cms,what)( getattr(self.process, seq) ))
1226  self.schedule.append(getattr(self.process,prefix))
1227  else:
1228  for i,s in enumerate(seq.split('+')):
1229  sn=prefix+'%d'%(i)
1230  setattr(self.process,sn,getattr(cms,what)( getattr(self.process, s) ))
1231  self.schedule.append(getattr(self.process,sn))
1232  return
1233 
1234  def scheduleSequenceAtEnd(self,seq,prefix):
1235  self.scheduleSequence(seq,prefix,what='EndPath')
1236  return
1237 
1238  def prepare_ALCAPRODUCER(self, sequence = None):
1239  self.prepare_ALCA(sequence, workflow = "producers")
1240 
1241  def prepare_ALCAOUTPUT(self, sequence = None):
1242  self.prepare_ALCA(sequence, workflow = "output")
1243 
1244  def prepare_ALCA(self, sequence = None, workflow = 'full'):
1245  """ Enrich the process with alca streams """
1246  alcaConfig=self.loadDefaultOrSpecifiedCFF(sequence,self.ALCADefaultCFF)
1247  sequence = sequence.split('.')[-1]
1248 
1249  # decide which ALCA paths to use
1250  alcaList = sequence.split("+")
1251  maxLevel=0
1252  from Configuration.AlCa.autoAlca import autoAlca
1253  # support @X from autoAlca.py, and recursion support: i.e T0:@Mu+@EG+...
1254  self.expandMapping(alcaList,autoAlca)
1255  self.AlCaPaths=[]
1256  for name in alcaConfig.__dict__:
1257  alcastream = getattr(alcaConfig,name)
1258  shortName = name.replace('ALCARECOStream','')
1259  if shortName in alcaList and isinstance(alcastream,cms.FilteredStream):
1260  output = self.addExtraStream(name,alcastream, workflow = workflow)
1261  self.executeAndRemember('process.ALCARECOEventContent.outputCommands.extend(process.OutALCARECO'+shortName+'_noDrop.outputCommands)')
1262  self.AlCaPaths.append(shortName)
1263  if 'DQM' in alcaList:
1264  if not self._options.inlineEventContent and hasattr(self.process,name):
1265  self.executeAndRemember('process.' + name + '.outputCommands.append("keep *_MEtoEDMConverter_*_*")')
1266  else:
1267  output.outputCommands.append("keep *_MEtoEDMConverter_*_*")
1268 
1269  #rename the HLT process name in the alca modules
1270  if self._options.hltProcess or 'HLT' in self.stepMap:
1271  if isinstance(alcastream.paths,tuple):
1272  for path in alcastream.paths:
1273  self.renameHLTprocessInSequence(path.label())
1274  else:
1275  self.renameHLTprocessInSequence(alcastream.paths.label())
1276 
1277  for i in range(alcaList.count(shortName)):
1278  alcaList.remove(shortName)
1279 
1280  # DQM needs a special handling
1281  elif name == 'pathALCARECODQM' and 'DQM' in alcaList:
1282  path = getattr(alcaConfig,name)
1283  self.schedule.append(path)
1284  alcaList.remove('DQM')
1285 
1286  if isinstance(alcastream,cms.Path):
1287  #black list the alca path so that they do not appear in the cfg
1288  self.blacklist_paths.append(alcastream)
1289 
1290 
1291  if len(alcaList) != 0:
1292  available=[]
1293  for name in alcaConfig.__dict__:
1294  alcastream = getattr(alcaConfig,name)
1295  if isinstance(alcastream,cms.FilteredStream):
1296  available.append(name.replace('ALCARECOStream',''))
1297  print("The following alcas could not be found "+str(alcaList))
1298  print("available ",available)
1299  #print "verify your configuration, ignoring for now"
1300  raise Exception("The following alcas could not be found "+str(alcaList))
1301 
1302  def prepare_LHE(self, sequence = None):
1303  #load the fragment
1304  ##make it loadable
1305  loadFragment = self._options.evt_type.replace('.py','',).replace('.','_').replace('python/','').replace('/','.')
1306  print("Loading lhe fragment from",loadFragment)
1307  __import__(loadFragment)
1308  self.process.load(loadFragment)
1309  ##inline the modules
1310  self._options.inlineObjets+=','+sequence
1311 
1312  getattr(self.process,sequence).nEvents = int(self._options.number)
1313 
1314  #schedule it
1315  self.process.lhe_step = cms.Path( getattr( self.process,sequence) )
1316  self.excludedPaths.append("lhe_step")
1317  self.schedule.append( self.process.lhe_step )
1318 
1319  def prepare_GEN(self, sequence = None):
1320  """ load the fragment of generator configuration """
1321  loadFailure=False
1322  #remove trailing .py
1323  #support old style .cfi by changing into something.cfi into something_cfi
1324  #remove python/ from the name
1325  loadFragment = self._options.evt_type.replace('.py','',).replace('.','_').replace('python/','')
1326  #standard location of fragments
1327  if not '/' in loadFragment:
1328  loadFragment='Configuration.Generator.'+loadFragment
1329  else:
1330  loadFragment=loadFragment.replace('/','.')
1331  try:
1332  print("Loading generator fragment from",loadFragment)
1333  __import__(loadFragment)
1334  except:
1335  loadFailure=True
1336  #if self.process.source and self.process.source.type_()=='EmptySource':
1337  if not (self._options.filein or self._options.dasquery):
1338  raise Exception("Neither gen fragment of input files provided: this is an inconsistent GEN step configuration")
1339 
1340  if not loadFailure:
1341  generatorModule=sys.modules[loadFragment]
1342  genModules=generatorModule.__dict__
1343  #remove lhe producer module since this should have been
1344  #imported instead in the LHE step
1345  if self.LHEDefaultSeq in genModules:
1346  del genModules[self.LHEDefaultSeq]
1347 
1348  if self._options.hideGen:
1349  self.loadAndRemember(loadFragment)
1350  else:
1351  self.process.load(loadFragment)
1352  # expose the objects from that fragment to the configuration
1353  import FWCore.ParameterSet.Modules as cmstypes
1354  for name in genModules:
1355  theObject = getattr(generatorModule,name)
1356  if isinstance(theObject, cmstypes._Module):
1357  self._options.inlineObjets=name+','+self._options.inlineObjets
1358  elif isinstance(theObject, cms.Sequence) or isinstance(theObject, cmstypes.ESProducer):
1359  self._options.inlineObjets+=','+name
1360 
1361  if sequence == self.GENDefaultSeq or sequence == 'pgen_genonly':
1362  if 'ProductionFilterSequence' in genModules and ('generator' in genModules):
1363  self.productionFilterSequence = 'ProductionFilterSequence'
1364  elif 'generator' in genModules:
1365  self.productionFilterSequence = 'generator'
1366 
1367  """ Enrich the schedule with the rest of the generation step """
1368  self.loadDefaultOrSpecifiedCFF(sequence,self.GENDefaultCFF)
1369  genSeqName=sequence.split('.')[-1]
1370 
1371  if True:
1372  try:
1373  from Configuration.StandardSequences.VtxSmeared import VtxSmeared
1374  cffToBeLoaded=VtxSmeared[self._options.beamspot]
1375  self.loadAndRemember(cffToBeLoaded)
1376  except ImportError:
1377  raise Exception("VertexSmearing type or beamspot "+self._options.beamspot+" unknown.")
1378 
1379  if self._options.scenario == 'HeavyIons':
1380  if self._options.pileup=='HiMixGEN':
1381  self.loadAndRemember("Configuration/StandardSequences/GeneratorMix_cff")
1382  else:
1383  self.loadAndRemember("Configuration/StandardSequences/GeneratorHI_cff")
1384 
1385  self.process.generation_step = cms.Path( getattr(self.process,genSeqName) )
1386  self.schedule.append(self.process.generation_step)
1387 
1388  #register to the genstepfilter the name of the path (static right now, but might evolve)
1389  self.executeAndRemember('process.genstepfilter.triggerConditions=cms.vstring("generation_step")')
1390 
1391  if 'reGEN' in self.stepMap:
1392  #stop here
1393  return
1394 
1395  """ Enrich the schedule with the summary of the filter step """
1396  #the gen filter in the endpath
1397  self.loadAndRemember("GeneratorInterface/Core/genFilterSummary_cff")
1398  self.scheduleSequenceAtEnd('genFilterSummary','genfiltersummary_step')
1399  return
1400 
1401  def prepare_SIM(self, sequence = None):
1402  """ Enrich the schedule with the simulation step"""
1403  self.loadDefaultOrSpecifiedCFF(sequence,self.SIMDefaultCFF)
1404  if not self._options.fast:
1405  if self._options.gflash==True:
1406  self.loadAndRemember("Configuration/StandardSequences/GFlashSIM_cff")
1407 
1408  if self._options.magField=='0T':
1409  self.executeAndRemember("process.g4SimHits.UseMagneticField = cms.bool(False)")
1410  else:
1411  if self._options.magField=='0T':
1412  self.executeAndRemember("process.fastSimProducer.detectorDefinition.magneticFieldZ = cms.untracked.double(0.)")
1413 
1414  self.scheduleSequence(sequence.split('.')[-1],'simulation_step')
1415  return
1416 
1417  def prepare_DIGI(self, sequence = None):
1418  """ Enrich the schedule with the digitisation step"""
1419  self.loadDefaultOrSpecifiedCFF(sequence,self.DIGIDefaultCFF)
1420 
1421  if self._options.gflash==True:
1422  self.loadAndRemember("Configuration/StandardSequences/GFlashDIGI_cff")
1423 
1424  if sequence == 'pdigi_valid' or sequence == 'pdigi_hi':
1425  self.executeAndRemember("process.mix.digitizers = cms.PSet(process.theDigitizersValid)")
1426 
1427  if sequence != 'pdigi_nogen' and sequence != 'pdigi_valid_nogen' and sequence != 'pdigi_hi_nogen' and not self.process.source.type_()=='EmptySource':
1428  if self._options.inputEventContent=='':
1429  self._options.inputEventContent='REGEN'
1430  else:
1431  self._options.inputEventContent=self._options.inputEventContent+',REGEN'
1432 
1433 
1434  self.scheduleSequence(sequence.split('.')[-1],'digitisation_step')
1435  return
1436 
1437  def prepare_CFWRITER(self, sequence = None):
1438  """ Enrich the schedule with the crossing frame writer step"""
1440  self.scheduleSequence('pcfw','cfwriter_step')
1441  return
1442 
1443  def prepare_DATAMIX(self, sequence = None):
1444  """ Enrich the schedule with the digitisation step"""
1446  self.scheduleSequence('pdatamix','datamixing_step')
1447 
1448  if self._options.pileup_input:
1449  theFiles=''
1450  if self._options.pileup_input.startswith('dbs:') or self._options.pileup_input.startswith('das:'):
1451  theFiles=filesFromDASQuery('file dataset = %s'%(self._options.pileup_input[4:],),self._options.pileup_dasoption)[0]
1452  elif self._options.pileup_input.startswith("filelist:"):
1453  theFiles= (filesFromList(self._options.pileup_input[9:]))[0]
1454  else:
1455  theFiles=self._options.pileup_input.split(',')
1456  #print theFiles
1457  self.executeAndRemember( "process.mixData.input.fileNames = cms.untracked.vstring(%s)"%( theFiles ) )
1458 
1459  return
1460 
1461  def prepare_DIGI2RAW(self, sequence = None):
1462  self.loadDefaultOrSpecifiedCFF(sequence,self.DIGI2RAWDefaultCFF)
1463  self.scheduleSequence(sequence.split('.')[-1],'digi2raw_step')
1464  return
1465 
1466  def prepare_REPACK(self, sequence = None):
1467  self.loadDefaultOrSpecifiedCFF(sequence,self.REPACKDefaultCFF)
1468  self.scheduleSequence(sequence.split('.')[-1],'digi2repack_step')
1469  return
1470 
1471  def prepare_L1(self, sequence = None):
1472  """ Enrich the schedule with the L1 simulation step"""
1473  assert(sequence == None)
1474  self.loadAndRemember(self.L1EMDefaultCFF)
1475  self.scheduleSequence('SimL1Emulator','L1simulation_step')
1476  return
1477 
1478  def prepare_L1REPACK(self, sequence = None):
1479  """ Enrich the schedule with the L1 simulation step, running the L1 emulator on data unpacked from the RAW collection, and repacking the result in a new RAW collection"""
1480  supported = ['GT','GT1','GT2','GCTGT','Full','FullSimTP','FullMC','Full2015Data','uGT','CalouGT']
1481  if sequence in supported:
1482  self.loadAndRemember('Configuration/StandardSequences/SimL1EmulatorRepack_%s_cff'%sequence)
1483  if self._options.scenario == 'HeavyIons':
1484  self.renameInputTagsInSequence("SimL1Emulator","rawDataCollector","rawDataRepacker")
1485  self.scheduleSequence('SimL1Emulator','L1RePack_step')
1486  else:
1487  print("L1REPACK with '",sequence,"' is not supported! Supported choices are: ",supported)
1488  raise Exception('unsupported feature')
1489 
1490 
1491  def prepare_HLT(self, sequence = None):
1492  """ Enrich the schedule with the HLT simulation step"""
1493  if not sequence:
1494  print("no specification of the hlt menu has been given, should never happen")
1495  raise Exception('no HLT sequence provided')
1496 
1497  if '@' in sequence:
1498  # case where HLT:@something was provided
1499  from Configuration.HLT.autoHLT import autoHLT
1500  key = sequence[1:]
1501  if key in autoHLT:
1502  sequence = autoHLT[key]
1503  else:
1504  raise ValueError('no HLT mapping key "%s" found in autoHLT' % key)
1505 
1506  if ',' in sequence:
1507  #case where HLT:something:something was provided
1508  self.executeAndRemember('import HLTrigger.Configuration.Utilities')
1509  optionsForHLT = {}
1510  if self._options.scenario == 'HeavyIons':
1511  optionsForHLT['type'] = 'HIon'
1512  else:
1513  optionsForHLT['type'] = 'GRun'
1514  optionsForHLTConfig = ', '.join('%s=%s' % (key, repr(val)) for (key, val) in six.iteritems(optionsForHLT))
1515  if sequence == 'run,fromSource':
1516  if hasattr(self.process.source,'firstRun'):
1517  self.executeAndRemember('process.loadHltConfiguration("run:%%d"%%(process.source.firstRun.value()),%s)'%(optionsForHLTConfig))
1518  elif hasattr(self.process.source,'setRunNumber'):
1519  self.executeAndRemember('process.loadHltConfiguration("run:%%d"%%(process.source.setRunNumber.value()),%s)'%(optionsForHLTConfig))
1520  else:
1521  raise Exception('Cannot replace menu to load %s'%(sequence))
1522  else:
1523  self.executeAndRemember('process.loadHltConfiguration("%s",%s)'%(sequence.replace(',',':'),optionsForHLTConfig))
1524  else:
1525  self.loadAndRemember('HLTrigger/Configuration/HLT_%s_cff' % sequence)
1526 
1527  if self._options.isMC:
1528  self._options.customisation_file.append("HLTrigger/Configuration/customizeHLTforMC.customizeHLTforMC")
1529 
1530  if self._options.name != 'HLT':
1531  self.additionalCommands.append('from HLTrigger.Configuration.CustomConfigs import ProcessName')
1532  self.additionalCommands.append('process = ProcessName(process)')
1533  self.additionalCommands.append('')
1534  from HLTrigger.Configuration.CustomConfigs import ProcessName
1535  self.process = ProcessName(self.process)
1536 
1537  self.schedule.append(self.process.HLTSchedule)
1538  [self.blacklist_paths.append(path) for path in self.process.HLTSchedule if isinstance(path,(cms.Path,cms.EndPath))]
1539 
1540  #this is a fake, to be removed with fastim migration and HLT menu dump
1541  if self._options.fast:
1542  if not hasattr(self.process,'HLTEndSequence'):
1543  self.executeAndRemember("process.HLTEndSequence = cms.Sequence( process.dummyModule )")
1544 
1545 
1546  def prepare_RAW2RECO(self, sequence = None):
1547  if ','in sequence:
1548  seqReco=sequence.split(',')[1]
1549  seqDigi=sequence.split(',')[0]
1550  else:
1551  print("RAW2RECO requires two specifications",sequence,"insufficient")
1552 
1553  self.prepare_RAW2DIGI(seqDigi)
1554  self.prepare_RECO(seqReco)
1555  return
1556 
1557  def prepare_RAW2DIGI(self, sequence = "RawToDigi"):
1558  self.loadDefaultOrSpecifiedCFF(sequence,self.RAW2DIGIDefaultCFF)
1559  self.scheduleSequence(sequence,'raw2digi_step')
1560  # if self._options.isRepacked:
1561  #self.renameInputTagsInSequence(sequence)
1562  return
1563 
1564  def prepare_PATFILTER(self, sequence=None):
1565  self.loadAndRemember("PhysicsTools/PatAlgos/slimming/metFilterPaths_cff")
1566  from PhysicsTools.PatAlgos.slimming.metFilterPaths_cff import allMetFilterPaths
1567  for filt in allMetFilterPaths:
1568  self.schedule.append(getattr(self.process,'Flag_'+filt))
1569 
1570  def prepare_L1HwVal(self, sequence = 'L1HwVal'):
1571  ''' Enrich the schedule with L1 HW validation '''
1572  self.loadDefaultOrSpecifiedCFF(sequence,self.L1HwValDefaultCFF)
1573  #self.scheduleSequence(sequence.split('.')[-1],'l1hwval_step')
1574  print('\n\n\n DEPRECATED this has no action \n\n\n')
1575  return
1576 
1577  def prepare_L1Reco(self, sequence = "L1Reco"):
1578  ''' Enrich the schedule with L1 reconstruction '''
1579  self.loadDefaultOrSpecifiedCFF(sequence,self.L1RecoDefaultCFF)
1580  self.scheduleSequence(sequence.split('.')[-1],'L1Reco_step')
1581  return
1582 
1583  def prepare_L1TrackTrigger(self, sequence = "L1TrackTrigger"):
1584  ''' Enrich the schedule with L1 reconstruction '''
1586  self.scheduleSequence(sequence.split('.')[-1],'L1TrackTrigger_step')
1587  return
1588 
1589  def prepare_FILTER(self, sequence = None):
1590  ''' Enrich the schedule with a user defined filter sequence '''
1591  ## load the relevant part
1592  filterConfig=self.load(sequence.split('.')[0])
1593  filterSeq=sequence.split('.')[-1]
1594  ## print it in the configuration
1595  class PrintAllModules(object):
1596  def __init__(self):
1597  self.inliner=''
1598  pass
1599  def enter(self,visitee):
1600  try:
1601  label=visitee.label()
1602  ##needs to be in reverse order
1603  self.inliner=label+','+self.inliner
1604  except:
1605  pass
1606  def leave(self,v): pass
1607 
1608  expander=PrintAllModules()
1609  getattr(self.process,filterSeq).visit( expander )
1610  self._options.inlineObjets+=','+expander.inliner
1611  self._options.inlineObjets+=','+filterSeq
1612 
1613  ## put the filtering path in the schedule
1614  self.scheduleSequence(filterSeq,'filtering_step')
1615  self.nextScheduleIsConditional=True
1616  ## put it before all the other paths
1617  self.productionFilterSequence = filterSeq
1618 
1619  return
1620 
1621  def prepare_RECO(self, sequence = "reconstruction"):
1622  ''' Enrich the schedule with reconstruction '''
1623  self.loadDefaultOrSpecifiedCFF(sequence,self.RECODefaultCFF)
1624  self.scheduleSequence(sequence.split('.')[-1],'reconstruction_step')
1625  return
1626 
1627  def prepare_RECOSIM(self, sequence = "recosim"):
1628  ''' Enrich the schedule with reconstruction '''
1629  self.loadDefaultOrSpecifiedCFF(sequence,self.RECOSIMDefaultCFF)
1630  self.scheduleSequence(sequence.split('.')[-1],'recosim_step')
1631  return
1632 
1633  def prepare_RECOBEFMIX(self, sequence = "reconstruction"):
1634  ''' Enrich the schedule with the part of reconstruction that is done before mixing in FastSim'''
1635  if not self._options.fast:
1636  print("ERROR: this step is only implemented for FastSim")
1637  sys.exit()
1639  self.scheduleSequence(sequence.split('.')[-1],'reconstruction_befmix_step')
1640  return
1641 
1642  def prepare_PAT(self, sequence = "miniAOD"):
1643  ''' Enrich the schedule with PAT '''
1644  self.prepare_PATFILTER(self)
1645  self.loadDefaultOrSpecifiedCFF(sequence,self.PATDefaultCFF)
1646  self.labelsToAssociate.append('patTask')
1647  if not self._options.runUnscheduled:
1648  raise Exception("MiniAOD production can only run in unscheduled mode, please run cmsDriver with --runUnscheduled")
1649  if self._options.isData:
1650  self._options.customisation_file_unsch.insert(0,"PhysicsTools/PatAlgos/slimming/miniAOD_tools.miniAOD_customizeAllData")
1651  else:
1652  if self._options.fast:
1653  self._options.customisation_file_unsch.insert(0,"PhysicsTools/PatAlgos/slimming/miniAOD_tools.miniAOD_customizeAllMCFastSim")
1654  else:
1655  self._options.customisation_file_unsch.insert(0,"PhysicsTools/PatAlgos/slimming/miniAOD_tools.miniAOD_customizeAllMC")
1656 
1657  if self._options.hltProcess:
1658  if len(self._options.customise_commands) > 1:
1659  self._options.customise_commands = self._options.customise_commands + " \n"
1660  self._options.customise_commands = self._options.customise_commands + "process.patTrigger.processName = \""+self._options.hltProcess+"\"\n"
1661  self._options.customise_commands = self._options.customise_commands + "process.slimmedPatTrigger.triggerResults= cms.InputTag( 'TriggerResults::"+self._options.hltProcess+"' )\n"
1662  self._options.customise_commands = self._options.customise_commands + "process.patMuons.triggerResults= cms.InputTag( 'TriggerResults::"+self._options.hltProcess+"' )\n"
1663 
1664 # self.renameHLTprocessInSequence(sequence)
1665 
1666  return
1667 
1668  def prepare_PATGEN(self, sequence = "miniGEN"):
1669  ''' Enrich the schedule with PATGEN '''
1670  self.loadDefaultOrSpecifiedCFF(sequence,self.PATGENDefaultCFF) #this is unscheduled
1671  self.labelsToAssociate.append('patGENTask')
1672  if not self._options.runUnscheduled:
1673  raise Exception("MiniGEN production can only run in unscheduled mode, please run cmsDriver with --runUnscheduled")
1674  if self._options.isData:
1675  raise Exception("PATGEN step can only run on MC")
1676  return
1677 
1678  def prepare_NANO(self, sequence = "nanoAOD"):
1679  ''' Enrich the schedule with NANO '''
1680  self.loadDefaultOrSpecifiedCFF(sequence,self.NANODefaultCFF)
1681  self.scheduleSequence(sequence.split('.')[-1],'nanoAOD_step')
1682  custom = "nanoAOD_customizeData" if self._options.isData else "nanoAOD_customizeMC"
1683  if self._options.runUnscheduled:
1684  self._options.customisation_file_unsch.insert(0,"PhysicsTools/NanoAOD/nano_cff."+custom)
1685  else:
1686  self._options.customisation_file.insert(0,"PhysicsTools/NanoAOD/nano_cff."+custom)
1687  if self._options.hltProcess:
1688  if len(self._options.customise_commands) > 1:
1689  self._options.customise_commands = self._options.customise_commands + " \n"
1690  self._options.customise_commands = self._options.customise_commands + "process.unpackedPatTrigger.triggerResults= cms.InputTag( 'TriggerResults::"+self._options.hltProcess+"' )\n"
1691 
1692 
1693  def prepare_EI(self, sequence = None):
1694  ''' Enrich the schedule with event interpretation '''
1695  from Configuration.StandardSequences.EventInterpretation import EventInterpretation
1696  if sequence in EventInterpretation:
1697  self.EIDefaultCFF = EventInterpretation[sequence]
1698  sequence = 'EIsequence'
1699  else:
1700  raise Exception('Cannot set %s event interpretation'%( sequence) )
1701  self.loadDefaultOrSpecifiedCFF(sequence,self.EIDefaultCFF)
1702  self.scheduleSequence(sequence.split('.')[-1],'eventinterpretaion_step')
1703  return
1704 
1705  def prepare_SKIM(self, sequence = "all"):
1706  ''' Enrich the schedule with skimming fragments'''
1707  skimConfig = self.loadDefaultOrSpecifiedCFF(sequence,self.SKIMDefaultCFF)
1708  sequence = sequence.split('.')[-1]
1709 
1710  skimlist=sequence.split('+')
1711  ## support @Mu+DiJet+@Electron configuration via autoSkim.py
1712  from Configuration.Skimming.autoSkim import autoSkim
1713  self.expandMapping(skimlist,autoSkim)
1714 
1715  #print "dictionnary for skims:",skimConfig.__dict__
1716  for skim in skimConfig.__dict__:
1717  skimstream = getattr(skimConfig,skim)
1718  if isinstance(skimstream,cms.Path):
1719  #black list the alca path so that they do not appear in the cfg
1720  self.blacklist_paths.append(skimstream)
1721  if (not isinstance(skimstream,cms.FilteredStream)):
1722  continue
1723  shortname = skim.replace('SKIMStream','')
1724  if (sequence=="all"):
1725  self.addExtraStream(skim,skimstream)
1726  elif (shortname in skimlist):
1727  self.addExtraStream(skim,skimstream)
1728  #add a DQM eventcontent for this guy
1729  if self._options.datatier=='DQM':
1730  self.process.load(self.EVTCONTDefaultCFF)
1731  skimstreamDQM = cms.FilteredStream(
1732  responsible = skimstream.responsible,
1733  name = skimstream.name+'DQM',
1734  paths = skimstream.paths,
1735  selectEvents = skimstream.selectEvents,
1736  content = self._options.datatier+'EventContent',
1737  dataTier = cms.untracked.string(self._options.datatier)
1738  )
1739  self.addExtraStream(skim+'DQM',skimstreamDQM)
1740  for i in range(skimlist.count(shortname)):
1741  skimlist.remove(shortname)
1742 
1743 
1744 
1745  if (skimlist.__len__()!=0 and sequence!="all"):
1746  print('WARNING, possible typo with SKIM:'+'+'.join(skimlist))
1747  raise Exception('WARNING, possible typo with SKIM:'+'+'.join(skimlist))
1748 
1749  def prepare_USER(self, sequence = None):
1750  ''' Enrich the schedule with a user defined sequence '''
1751  self.loadDefaultOrSpecifiedCFF(sequence,self.USERDefaultCFF)
1752  self.scheduleSequence(sequence.split('.')[-1],'user_step')
1753  return
1754 
1755  def prepare_POSTRECO(self, sequence = None):
1756  """ Enrich the schedule with the postreco step """
1758  self.scheduleSequence('postreco_generator','postreco_step')
1759  return
1760 
1761 
1762  def prepare_VALIDATION(self, sequence = 'validation'):
1763  print(sequence,"in preparing validation")
1765  from Validation.Configuration.autoValidation import autoValidation
1766  #in case VALIDATION:something:somethingelse -> something,somethingelse
1767  sequence=sequence.split('.')[-1]
1768  if sequence.find(',')!=-1:
1769  prevalSeqName=sequence.split(',')[0].split('+')
1770  valSeqName=sequence.split(',')[1].split('+')
1771  self.expandMapping(prevalSeqName,autoValidation,index=0)
1772  self.expandMapping(valSeqName,autoValidation,index=1)
1773  else:
1774  if '@' in sequence:
1775  prevalSeqName=sequence.split('+')
1776  valSeqName=sequence.split('+')
1777  self.expandMapping(prevalSeqName,autoValidation,index=0)
1778  self.expandMapping(valSeqName,autoValidation,index=1)
1779  else:
1780  postfix=''
1781  if sequence:
1782  postfix='_'+sequence
1783  prevalSeqName=['prevalidation'+postfix]
1784  valSeqName=['validation'+postfix]
1785  if not hasattr(self.process,valSeqName[0]):
1786  prevalSeqName=['']
1787  valSeqName=[sequence]
1788 
1789  def NFI(index):
1790  ##name from index, required to keep backward compatibility
1791  if index==0:
1792  return ''
1793  else:
1794  return '%s'%index
1795 
1796 
1797  #rename the HLT process in validation steps
1798  if ('HLT' in self.stepMap and not self._options.fast) or self._options.hltProcess:
1799  for s in valSeqName+prevalSeqName:
1800  if s:
1802  for (i,s) in enumerate(prevalSeqName):
1803  if s:
1804  setattr(self.process,'prevalidation_step%s'%NFI(i), cms.Path( getattr(self.process, s)) )
1805  self.schedule.append(getattr(self.process,'prevalidation_step%s'%NFI(i)))
1806 
1807  for (i,s) in enumerate(valSeqName):
1808  setattr(self.process,'validation_step%s'%NFI(i), cms.EndPath( getattr(self.process, s)))
1809  self.schedule.append(getattr(self.process,'validation_step%s'%NFI(i)))
1810 
1811  #needed in case the miniAODValidation sequence is run starting from AODSIM
1812  if 'PAT' in self.stepMap and not 'RECO' in self.stepMap:
1813  return
1814 
1815  if not 'DIGI' in self.stepMap and not self._options.fast and not any(map( lambda s : s.startswith('genvalid'), valSeqName)):
1816  if self._options.restoreRNDSeeds==False and not self._options.restoreRNDSeeds==True:
1817  self._options.restoreRNDSeeds=True
1818 
1819  if not 'DIGI' in self.stepMap and not self._options.fast:
1820  self.executeAndRemember("process.mix.playback = True")
1821  self.executeAndRemember("process.mix.digitizers = cms.PSet()")
1822  self.executeAndRemember("for a in process.aliases: delattr(process, a)")
1823  self._options.customisation_file.append("SimGeneral/MixingModule/fullMixCustomize_cff.setCrossingFrameOn")
1824 
1825  if hasattr(self.process,"genstepfilter") and len(self.process.genstepfilter.triggerConditions):
1826  #will get in the schedule, smoothly
1827  for (i,s) in enumerate(valSeqName):
1828  getattr(self.process,'validation_step%s'%NFI(i)).insert(0, self.process.genstepfilter)
1829 
1830  return
1831 
1832 
1834  """Visitor that travels within a cms.Sequence, looks for a parameter and replace its value
1835  It will climb down within PSets, VPSets and VInputTags to find its target"""
1836  def __init__(self, paramSearch, paramReplace, verbose=False, whitelist=()):
1837  self._paramReplace = paramReplace
1838  self._paramSearch = paramSearch
1839  self._verbose = verbose
1840  self._whitelist = whitelist
1841 
1842  def doIt(self,pset,base):
1843  if isinstance(pset, cms._Parameterizable):
1844  for name in pset.parameters_().keys():
1845  # skip whitelisted parameters
1846  if name in self._whitelist:
1847  continue
1848  # if I use pset.parameters_().items() I get copies of the parameter values
1849  # so I can't modify the nested pset
1850  value = getattr(pset,name)
1851  type = value.pythonTypeName()
1852  if type in ('cms.PSet', 'cms.untracked.PSet'):
1853  self.doIt(value,base+"."+name)
1854  elif type in ('cms.VPSet', 'cms.untracked.VPSet'):
1855  for (i,ps) in enumerate(value): self.doIt(ps, "%s.%s[%d]"%(base,name,i) )
1856  elif type in ('cms.string', 'cms.untracked.string'):
1857  if value.value() == self._paramSearch:
1858  if self._verbose: print("set string process name %s.%s %s ==> %s"% (base, name, value, self._paramReplace))
1859  setattr(pset, name,self._paramReplace)
1860  elif type in ('cms.VInputTag', 'cms.untracked.VInputTag'):
1861  for (i,n) in enumerate(value):
1862  if not isinstance(n, cms.InputTag):
1863  n=cms.InputTag(n)
1864  if n.processName == self._paramSearch:
1865  # VInputTag can be declared as a list of strings, so ensure that n is formatted correctly
1866  if self._verbose:print("set process name %s.%s[%d] %s ==> %s " % (base, name, i, n, self._paramReplace))
1867  setattr(n,"processName",self._paramReplace)
1868  value[i]=n
1869  elif type in ('cms.vstring', 'cms.untracked.vstring'):
1870  for (i,n) in enumerate(value):
1871  if n==self._paramSearch:
1872  getattr(pset,name)[i]=self._paramReplace
1873  elif type in ('cms.InputTag', 'cms.untracked.InputTag'):
1874  if value.processName == self._paramSearch:
1875  if self._verbose: print("set process name %s.%s %s ==> %s " % (base, name, value, self._paramReplace))
1876  setattr(getattr(pset, name),"processName",self._paramReplace)
1877 
1878  def enter(self,visitee):
1879  label = ''
1880  try:
1881  label = visitee.label()
1882  except AttributeError:
1883  label = '<Module not in a Process>'
1884  except:
1885  label = 'other execption'
1886  self.doIt(visitee, label)
1887 
1888  def leave(self,visitee):
1889  pass
1890 
1891  #visit a sequence to repalce all input tags
1892  def renameInputTagsInSequence(self,sequence,oldT="rawDataCollector",newT="rawDataRepacker"):
1893  print("Replacing all InputTag %s => %s"%(oldT,newT))
1894  from PhysicsTools.PatAlgos.tools.helpers import massSearchReplaceAnyInputTag
1895  massSearchReplaceAnyInputTag(getattr(self.process,sequence),oldT,newT)
1896  loadMe='from PhysicsTools.PatAlgos.tools.helpers import massSearchReplaceAnyInputTag'
1897  if not loadMe in self.additionalCommands:
1898  self.additionalCommands.append(loadMe)
1899  self.additionalCommands.append('massSearchReplaceAnyInputTag(process.%s,"%s","%s",False,True)'%(sequence,oldT,newT))
1900 
1901  #change the process name used to address HLT results in any sequence
1902  def renameHLTprocessInSequence(self,sequence,proc=None,HLTprocess='HLT'):
1903  if self._options.hltProcess:
1904  proc=self._options.hltProcess
1905  else:
1906  proc=self.process.name_()
1907  if proc==HLTprocess: return
1908  # look up all module in dqm sequence
1909  print("replacing %s process name - sequence %s will use '%s'" % (HLTprocess,sequence, proc))
1910  getattr(self.process,sequence).visit(ConfigBuilder.MassSearchReplaceProcessNameVisitor(HLTprocess,proc,whitelist = ("subSystemFolder",)))
1911  if 'from Configuration.Applications.ConfigBuilder import ConfigBuilder' not in self.additionalCommands:
1912  self.additionalCommands.append('from Configuration.Applications.ConfigBuilder import ConfigBuilder')
1913  self.additionalCommands.append('process.%s.visit(ConfigBuilder.MassSearchReplaceProcessNameVisitor("%s", "%s", whitelist = ("subSystemFolder",)))'% (sequence,HLTprocess, proc))
1914 
1915 
1916  def expandMapping(self,seqList,mapping,index=None):
1917  maxLevel=20
1918  level=0
1919  while '@' in repr(seqList) and level<maxLevel:
1920  level+=1
1921  for specifiedCommand in seqList:
1922  if specifiedCommand.startswith('@'):
1923  location=specifiedCommand[1:]
1924  if not location in mapping:
1925  raise Exception("Impossible to map "+location+" from "+repr(mapping))
1926  mappedTo=mapping[location]
1927  if index!=None:
1928  mappedTo=mappedTo[index]
1929  seqList.remove(specifiedCommand)
1930  seqList.extend(mappedTo.split('+'))
1931  break;
1932  if level==maxLevel:
1933  raise Exception("Could not fully expand "+repr(seqList)+" from "+repr(mapping))
1934 
1935  def prepare_DQM(self, sequence = 'DQMOffline'):
1936  # this one needs replacement
1937 
1938  # any 'DQM' job should use DQMStore in non-legacy mode (but not HARVESTING)
1939  self.loadAndRemember("DQMServices/Core/DQMStoreNonLegacy_cff")
1941  sequenceList=sequence.split('.')[-1].split('+')
1942  postSequenceList=sequence.split('.')[-1].split('+')
1943  from DQMOffline.Configuration.autoDQM import autoDQM
1944  self.expandMapping(sequenceList,autoDQM,index=0)
1945  self.expandMapping(postSequenceList,autoDQM,index=1)
1946 
1947  if len(set(sequenceList))!=len(sequenceList):
1948  sequenceList=list(set(sequenceList))
1949  print("Duplicate entries for DQM:, using",sequenceList)
1950 
1951  pathName='dqmoffline_step'
1952  for (i,sequence) in enumerate(sequenceList):
1953  if (i!=0):
1954  pathName='dqmoffline_%d_step'%(i)
1955 
1956  if 'HLT' in self.stepMap.keys() or self._options.hltProcess:
1957  self.renameHLTprocessInSequence(sequence)
1958 
1959  setattr(self.process,pathName, cms.EndPath( getattr(self.process,sequence ) ) )
1960  self.schedule.append(getattr(self.process,pathName))
1961 
1962  if hasattr(self.process,"genstepfilter") and len(self.process.genstepfilter.triggerConditions):
1963  #will get in the schedule, smoothly
1964  getattr(self.process,pathName).insert(0,self.process.genstepfilter)
1965 
1966  pathName='dqmofflineOnPAT_step'
1967  for (i,sequence) in enumerate(postSequenceList):
1968  if (i!=0):
1969  pathName='dqmofflineOnPAT_%d_step'%(i)
1970 
1971  setattr(self.process,pathName, cms.EndPath( getattr(self.process, sequence ) ) )
1972  self.schedule.append(getattr(self.process,pathName))
1973 
1974  def prepare_HARVESTING(self, sequence = None):
1975  """ Enrich the process with harvesting step """
1976  self.DQMSaverCFF='Configuration/StandardSequences/DQMSaver'+self._options.harvesting+'_cff'
1977  self.loadAndRemember(self.DQMSaverCFF)
1978 
1979  harvestingConfig = self.loadDefaultOrSpecifiedCFF(sequence,self.HARVESTINGDefaultCFF)
1980  sequence = sequence.split('.')[-1]
1981 
1982  # decide which HARVESTING paths to use
1983  harvestingList = sequence.split("+")
1984  from DQMOffline.Configuration.autoDQM import autoDQM
1985  from Validation.Configuration.autoValidation import autoValidation
1986  import copy
1987  combined_mapping = copy.deepcopy( autoDQM )
1988  combined_mapping.update( autoValidation )
1989  self.expandMapping(harvestingList,combined_mapping,index=-1)
1990 
1991  if len(set(harvestingList))!=len(harvestingList):
1992  harvestingList=list(set(harvestingList))
1993  print("Duplicate entries for HARVESTING, using",harvestingList)
1994 
1995  for name in harvestingList:
1996  if not name in harvestingConfig.__dict__:
1997  print(name,"is not a possible harvesting type. Available are",harvestingConfig.__dict__.keys())
1998  continue
1999  harvestingstream = getattr(harvestingConfig,name)
2000  if isinstance(harvestingstream,cms.Path):
2001  self.schedule.append(harvestingstream)
2002  self.blacklist_paths.append(harvestingstream)
2003  if isinstance(harvestingstream,cms.Sequence):
2004  setattr(self.process,name+"_step",cms.Path(harvestingstream))
2005  self.schedule.append(getattr(self.process,name+"_step"))
2006 
2007  self.scheduleSequence('DQMSaver','dqmsave_step')
2008  return
2009 
2010  def prepare_ALCAHARVEST(self, sequence = None):
2011  """ Enrich the process with AlCaHarvesting step """
2012  harvestingConfig = self.loadAndRemember(self.ALCAHARVESTDefaultCFF)
2013  sequence=sequence.split(".")[-1]
2014 
2015  # decide which AlcaHARVESTING paths to use
2016  harvestingList = sequence.split("+")
2017 
2018 
2019 
2020  from Configuration.AlCa.autoPCL import autoPCL
2021  self.expandMapping(harvestingList,autoPCL)
2022 
2023  for name in harvestingConfig.__dict__:
2024  harvestingstream = getattr(harvestingConfig,name)
2025  if name in harvestingList and isinstance(harvestingstream,cms.Path):
2026  self.schedule.append(harvestingstream)
2027  if isinstance(getattr(harvestingConfig,"ALCAHARVEST" + name + "_dbOutput"), cms.VPSet) and \
2028  isinstance(getattr(harvestingConfig,"ALCAHARVEST" + name + "_metadata"), cms.VPSet):
2029  self.executeAndRemember("process.PoolDBOutputService.toPut.extend(process.ALCAHARVEST" + name + "_dbOutput)")
2030  self.executeAndRemember("process.pclMetadataWriter.recordsToMap.extend(process.ALCAHARVEST" + name + "_metadata)")
2031  else:
2032  self.executeAndRemember("process.PoolDBOutputService.toPut.append(process.ALCAHARVEST" + name + "_dbOutput)")
2033  self.executeAndRemember("process.pclMetadataWriter.recordsToMap.append(process.ALCAHARVEST" + name + "_metadata)")
2034  harvestingList.remove(name)
2035  # append the common part at the end of the sequence
2036  lastStep = getattr(harvestingConfig,"ALCAHARVESTDQMSaveAndMetadataWriter")
2037  self.schedule.append(lastStep)
2038 
2039  if len(harvestingList) != 0 and 'dummyHarvesting' not in harvestingList :
2040  print("The following harvesting could not be found : ", harvestingList)
2041  raise Exception("The following harvesting could not be found : "+str(harvestingList))
2042 
2043 
2044 
2045  def prepare_ENDJOB(self, sequence = 'endOfProcess'):
2046  self.loadDefaultOrSpecifiedCFF(sequence,self.ENDJOBDefaultCFF)
2047  self.scheduleSequenceAtEnd(sequence.split('.')[-1],'endjob_step')
2048  return
2049 
2051  self.process.reconstruction = cms.Path(self.process.reconstructionWithFamos)
2052  self.schedule.append(self.process.reconstruction)
2053 
2054 
2055  def build_production_info(self, evt_type, evtnumber):
2056  """ Add useful info for the production. """
2057  self.process.configurationMetadata=cms.untracked.PSet\
2058  (version=cms.untracked.string("$Revision: 1.19 $"),
2059  name=cms.untracked.string("Applications"),
2060  annotation=cms.untracked.string(evt_type+ " nevts:"+str(evtnumber))
2061  )
2062 
2063  self.addedObjects.append(("Production Info","configurationMetadata"))
2064 
2065 
2066  def create_process(self):
2067  self.pythonCfgCode = "# Auto generated configuration file\n"
2068  self.pythonCfgCode += "# using: \n# "+__version__[1:-1]+"\n# "+__source__[1:-1]+'\n'
2069  self.pythonCfgCode += "# with command line options: "+self._options.arguments+'\n'
2070  self.pythonCfgCode += "import FWCore.ParameterSet.Config as cms\n\n"
2071 
2072  # now set up the modifies
2073  modifiers=[]
2074  modifierStrings=[]
2075  modifierImports=[]
2076 
2077  if hasattr(self._options,"era") and self._options.era :
2078  # Multiple eras can be specified in a comma seperated list
2079  from Configuration.StandardSequences.Eras import eras
2080  for requestedEra in self._options.era.split(",") :
2081  modifierStrings.append(requestedEra)
2082  modifierImports.append(eras.pythonCfgLines[requestedEra])
2083  modifiers.append(getattr(eras,requestedEra))
2084 
2085 
2086  if hasattr(self._options,"procModifiers") and self._options.procModifiers:
2087  import importlib
2088  thingsImported=[]
2089  for pm in self._options.procModifiers.split(','):
2090  modifierStrings.append(pm)
2091  modifierImports.append('from Configuration.ProcessModifiers.'+pm+'_cff import '+pm)
2092  modifiers.append(getattr(importlib.import_module('Configuration.ProcessModifiers.'+pm+'_cff'),pm))
2093 
2094  self.pythonCfgCode += '\n'.join(modifierImports)+'\n\n'
2095  self.pythonCfgCode += "process = cms.Process('"+self._options.name+"'" # Start of the line, finished after the loop
2096 
2097 
2098  if len(modifierStrings)>0:
2099  self.pythonCfgCode+= ','+','.join(modifierStrings)
2100  self.pythonCfgCode+=')\n\n'
2101 
2102  #yes, the cfg code gets out of sync here if a process is passed in. That could be fixed in the future
2103  #assuming there is some way for the fwk to get the list of modifiers (and their stringified name)
2104  if self.process == None:
2105  if len(modifiers)>0:
2106  self.process = cms.Process(self._options.name,*modifiers)
2107  else:
2108  self.process = cms.Process(self._options.name)
2109 
2110 
2111 
2112 
2113  def prepare(self, doChecking = False):
2114  """ Prepare the configuration string and add missing pieces."""
2115 
2116  self.loadAndRemember(self.EVTCONTDefaultCFF) #load the event contents regardless
2117  self.addMaxEvents()
2118  if self.with_input:
2119  self.addSource()
2120  self.addStandardSequences()
2121  ##adding standard sequences might change the inputEventContent option and therefore needs to be finalized after
2122  self.completeInputCommand()
2123  self.addConditions()
2124 
2125 
2126  outputModuleCfgCode=""
2127  if not 'HARVESTING' in self.stepMap.keys() and not 'ALCAHARVEST' in self.stepMap.keys() and not 'ALCAOUTPUT' in self.stepMap.keys() and self.with_output:
2128  outputModuleCfgCode=self.addOutput()
2129 
2130  self.addCommon()
2131 
2132  self.pythonCfgCode += "# import of standard configurations\n"
2133  for module in self.imports:
2134  self.pythonCfgCode += ("process.load('"+module+"')\n")
2135 
2136  # production info
2137  if not hasattr(self.process,"configurationMetadata"):
2138  self.build_production_info(self._options.evt_type, self._options.number)
2139  else:
2140  #the PSet was added via a load
2141  self.addedObjects.append(("Production Info","configurationMetadata"))
2142 
2143  self.pythonCfgCode +="\n"
2144  for comment,object in self.addedObjects:
2145  if comment!="":
2146  self.pythonCfgCode += "\n# "+comment+"\n"
2147  self.pythonCfgCode += dumpPython(self.process,object)
2148 
2149  # dump the output definition
2150  self.pythonCfgCode += "\n# Output definition\n"
2151  self.pythonCfgCode += outputModuleCfgCode
2152 
2153  # dump all additional outputs (e.g. alca or skim streams)
2154  self.pythonCfgCode += "\n# Additional output definition\n"
2155  #I do not understand why the keys are not normally ordered.
2156  nl=sorted(self.additionalOutputs.keys())
2157  for name in nl:
2158  output = self.additionalOutputs[name]
2159  self.pythonCfgCode += "process.%s = %s" %(name, output.dumpPython())
2160  tmpOut = cms.EndPath(output)
2161  setattr(self.process,name+'OutPath',tmpOut)
2162  self.schedule.append(tmpOut)
2163 
2164  # dump all additional commands
2165  self.pythonCfgCode += "\n# Other statements\n"
2166  for command in self.additionalCommands:
2167  self.pythonCfgCode += command + "\n"
2168 
2169  #comma separated list of objects that deserve to be inlined in the configuration (typically from a modified config deep down)
2170  for object in self._options.inlineObjets.split(','):
2171  if not object:
2172  continue
2173  if not hasattr(self.process,object):
2174  print('cannot inline -'+object+'- : not known')
2175  else:
2176  self.pythonCfgCode +='\n'
2177  self.pythonCfgCode +=dumpPython(self.process,object)
2178 
2179  # dump all paths
2180  self.pythonCfgCode += "\n# Path and EndPath definitions\n"
2181  for path in self.process.paths:
2182  if getattr(self.process,path) not in self.blacklist_paths:
2183  self.pythonCfgCode += dumpPython(self.process,path)
2184 
2185  for endpath in self.process.endpaths:
2186  if getattr(self.process,endpath) not in self.blacklist_paths:
2187  self.pythonCfgCode += dumpPython(self.process,endpath)
2188 
2189  # dump the schedule
2190  self.pythonCfgCode += "\n# Schedule definition\n"
2191  result = "process.schedule = cms.Schedule("
2192 
2193  # handling of the schedule
2194  self.process.schedule = cms.Schedule()
2195  for item in self.schedule:
2196  if not isinstance(item, cms.Schedule):
2197  self.process.schedule.append(item)
2198  else:
2199  self.process.schedule.extend(item)
2200 
2201  if hasattr(self.process,"HLTSchedule"):
2202  beforeHLT = self.schedule[:self.schedule.index(self.process.HLTSchedule)]
2203  afterHLT = self.schedule[self.schedule.index(self.process.HLTSchedule)+1:]
2204  pathNames = ['process.'+p.label_() for p in beforeHLT]
2205  result += ','.join(pathNames)+')\n'
2206  result += 'process.schedule.extend(process.HLTSchedule)\n'
2207  pathNames = ['process.'+p.label_() for p in afterHLT]
2208  result += 'process.schedule.extend(['+','.join(pathNames)+'])\n'
2209  else:
2210  pathNames = ['process.'+p.label_() for p in self.schedule]
2211  result ='process.schedule = cms.Schedule('+','.join(pathNames)+')\n'
2212 
2213  self.pythonCfgCode += result
2214 
2215  for labelToAssociate in self.labelsToAssociate:
2216  self.process.schedule.associate(getattr(self.process, labelToAssociate))
2217  self.pythonCfgCode += 'process.schedule.associate(process.' + labelToAssociate + ')\n'
2218 
2219  from PhysicsTools.PatAlgos.tools.helpers import associatePatAlgosToolsTask
2221  self.pythonCfgCode+="from PhysicsTools.PatAlgos.tools.helpers import associatePatAlgosToolsTask\n"
2222  self.pythonCfgCode+="associatePatAlgosToolsTask(process)\n"
2223 
2224  if self._options.nThreads is not "1":
2225  self.pythonCfgCode +="\n"
2226  self.pythonCfgCode +="#Setup FWK for multithreaded\n"
2227  self.pythonCfgCode +="process.options.numberOfThreads=cms.untracked.uint32("+self._options.nThreads+")\n"
2228  self.pythonCfgCode +="process.options.numberOfStreams=cms.untracked.uint32("+self._options.nStreams+")\n"
2229  self.pythonCfgCode +="process.options.numberOfConcurrentLuminosityBlocks=cms.untracked.uint32("+self._options.nConcurrentLumis+")\n"
2230  #repacked version
2231  if self._options.isRepacked:
2232  self.pythonCfgCode +="\n"
2233  self.pythonCfgCode +="from Configuration.Applications.ConfigBuilder import MassReplaceInputTag\n"
2234  self.pythonCfgCode +="MassReplaceInputTag(process, new=\"rawDataMapperByLabel\", old=\"rawDataCollector\")\n"
2235  MassReplaceInputTag(self.process, new="rawDataMapperByLabel", old="rawDataCollector")
2236 
2237  # special treatment in case of production filter sequence 2/2
2238  if self.productionFilterSequence:
2239  self.pythonCfgCode +='# filter all path with the production filter sequence\n'
2240  self.pythonCfgCode +='for path in process.paths:\n'
2241  if len(self.conditionalPaths):
2242  self.pythonCfgCode +='\tif not path in %s: continue\n'%str(self.conditionalPaths)
2243  if len(self.excludedPaths):
2244  self.pythonCfgCode +='\tif path in %s: continue\n'%str(self.excludedPaths)
2245  self.pythonCfgCode +='\tgetattr(process,path).insert(0, process.%s)\n'%(self.productionFilterSequence,)
2246  pfs = getattr(self.process,self.productionFilterSequence)
2247  for path in self.process.paths:
2248  if not path in self.conditionalPaths: continue
2249  if path in self.excludedPaths: continue
2250  getattr(self.process,path).insert(0, pfs)
2251 
2252 
2253  # dump customise fragment
2254  self.pythonCfgCode += self.addCustomise()
2255 
2256  if self._options.runUnscheduled:
2257  # prune and delete paths
2258  #this is not supporting the blacklist at this point since I do not understand it
2259  self.pythonCfgCode+="#do not add changes to your config after this point (unless you know what you are doing)\n"
2260  self.pythonCfgCode+="from FWCore.ParameterSet.Utilities import convertToUnscheduled\n"
2261  self.pythonCfgCode+="process=convertToUnscheduled(process)\n"
2262 
2263  from FWCore.ParameterSet.Utilities import convertToUnscheduled
2265 
2266  self.pythonCfgCode += self.addCustomise(1)
2267 
2268  self.pythonCfgCode += self.addCustomiseCmdLine()
2269 
2270  if hasattr(self.process,"logErrorHarvester"):
2271  #configure logErrorHarvester to wait for same EDProducers to finish as the OutputModules
2272  self.pythonCfgCode +="\n#Have logErrorHarvester wait for the same EDProducers to finish as those providing data for the OutputModule\n"
2273  self.pythonCfgCode +="from FWCore.Modules.logErrorHarvester_cff import customiseLogErrorHarvesterUsingOutputCommands\n"
2274  self.pythonCfgCode +="process = customiseLogErrorHarvesterUsingOutputCommands(process)\n"
2275  from FWCore.Modules.logErrorHarvester_cff import customiseLogErrorHarvesterUsingOutputCommands
2277 
2278  # Temporary hack to put the early delete customization after
2279  # everything else
2280  #
2281  # FIXME: remove when no longer needed
2282  self.pythonCfgCode += "\n# Add early deletion of temporary data products to reduce peak memory need\n"
2283  self.pythonCfgCode += "from Configuration.StandardSequences.earlyDeleteSettings_cff import customiseEarlyDelete\n"
2284  self.pythonCfgCode += "process = customiseEarlyDelete(process)\n"
2285  self.pythonCfgCode += "# End adding early deletion\n"
2286  from Configuration.StandardSequences.earlyDeleteSettings_cff import customiseEarlyDelete
2287  self.process = customiseEarlyDelete(self.process)
2288 
2289 
2290  # make the .io file
2291 
2292  if self._options.io:
2293  #io=open(self._options.python_filename.replace('.py','.io'),'w')
2294  if not self._options.io.endswith('.io'): self._option.io+='.io'
2295  io=open(self._options.io,'w')
2296  ioJson={}
2297  if hasattr(self.process.source,"fileNames"):
2298  if len(self.process.source.fileNames.value()):
2299  ioJson['primary']=self.process.source.fileNames.value()
2300  if hasattr(self.process.source,"secondaryFileNames"):
2301  if len(self.process.source.secondaryFileNames.value()):
2302  ioJson['secondary']=self.process.source.secondaryFileNames.value()
2303  if self._options.pileup_input and (self._options.pileup_input.startswith('dbs:') or self._options.pileup_input.startswith('das:')):
2304  ioJson['pileup']=self._options.pileup_input[4:]
2305  for (o,om) in self.process.outputModules_().items():
2306  ioJson[o]=om.fileName.value()
2307  ioJson['GT']=self.process.GlobalTag.globaltag.value()
2308  if self.productionFilterSequence:
2309  ioJson['filter']=self.productionFilterSequence
2310  import json
2311  io.write(json.dumps(ioJson))
2312  return
2313 
def load(self, includeFile)
def prepare_ENDJOB(self, sequence='endOfProcess')
def prepare_LHE(self, sequence=None)
def prepare_DATAMIX(self, sequence=None)
def expandMapping(self, seqList, mapping, index=None)
def prepare_SIM(self, sequence=None)
def prepare_HARVESTING(self, sequence=None)
def prepare_USER(self, sequence=None)
def prepare_GEN(self, sequence=None)
def loadDefaultOrSpecifiedCFF(self, sequence, defaultCFF)
bool any(const std::vector< T > &v, const T &what)
Definition: ECalSD.cc:37
def massSearchReplaceAnyInputTag(sequence, oldInputTag, newInputTag, verbose=False, moduleLabelOnly=False, skipLabelTest=False)
Definition: MassReplace.py:73
def prepare_L1REPACK(self, sequence=None)
inliner
load the relevant part
def renameInputTagsInSequence(self, sequence, oldT="rawDataCollector", newT="rawDataRepacker")
def prepare_POSTRECO(self, sequence=None)
process
adding standard sequences might change the inputEventContent option and therefore needs to be finaliz...
def replace(string, replacements)
def prepare_REPACK(self, sequence=None)
def prepare_EI(self, sequence=None)
def prepare_RECOBEFMIX(self, sequence="reconstruction")
def prepare_FILTER(self, sequence=None)
S & print(S &os, JobReport::InputFile const &f)
Definition: JobReport.cc:66
def filesFromDASQuery(query, option="", s=None)
def prepare_PATFILTER(self, sequence=None)
def build_production_info(self, evt_type, evtnumber)
def ProcessName(process)
Definition: CustomConfigs.py:8
def __init__(self, paramSearch, paramReplace, verbose=False, whitelist=())
def visit(visitdir)
Retrieve data from a perf suite output (sub) directory, only examines TimeSize at the moment...
def prepare_L1HwVal(self, sequence='L1HwVal')
def prepare_CFWRITER(self, sequence=None)
def prepare_RAW2DIGI(self, sequence="RawToDigi")
def prepare_RECO(self, sequence="reconstruction")
OutputIterator zip(InputIterator1 first1, InputIterator1 last1, InputIterator2 first2, InputIterator2 last2, OutputIterator result, Compare comp)
def anyOf(listOfKeys, dict, opt=None)
def convertToUnscheduled(proc)
def prepare_PATGEN(self, sequence="miniGEN")
def addExtraStream(self, name, stream, workflow='full')
def prepare_DQM(self, sequence='DQMOffline')
def prepare_L1Reco(self, sequence="L1Reco")
def renameHLTprocessInSequence(self, sequence, proc=None, HLTprocess='HLT')
def prepare_RECOSIM(self, sequence="recosim")
productionFilterSequence
put it before all the other paths
runsAndWeights
drop LHEXMLStringProduct on input to save memory if appropriate
def addCustomise(self, unsch=0)
def prepare_ALCAPRODUCER(self, sequence=None)
def prepare_ALCAOUTPUT(self, sequence=None)
def scheduleSequence(self, seq, prefix, what='Path')
bool insert(Storage &iStorage, ItemType *iItem, const IdTag &iIdTag)
Definition: HCMethods.h:50
def prepare_DIGI(self, sequence=None)
def prepare_DIGI2RAW(self, sequence=None)
def loadAndRemember(self, includeFile)
def throwAndSetRandomRun(source, runsAndProbs)
def prepare_NANO(self, sequence="nanoAOD")
def prepare_SKIM(self, sequence="all")
static std::string join(char **cmd)
Definition: RemoteFile.cc:18
def prepare_ALCAHARVEST(self, sequence=None)
def defineMixing(dict)
Definition: Mixing.py:183
def dumpPython(process, name)
def miniAOD_customizeOutput(out)
def prepare_L1TrackTrigger(self, sequence="L1TrackTrigger")
def prepare(self, doChecking=False)
def associatePatAlgosToolsTask(process)
Definition: helpers.py:25
def prepare_ALCA(self, sequence=None, workflow='full')
def prepare_HLT(self, sequence=None)
def __init__(self, options, process=None, with_output=False, with_input=False)
def prepare_VALIDATION(self, sequence='validation')
def prepare_L1(self, sequence=None)
def scheduleSequenceAtEnd(self, seq, prefix)
#define str(s)
def filesFromList(fileName, s=None)
def prepare_RAW2RECO(self, sequence=None)
def customiseLogErrorHarvesterUsingOutputCommands(process)
double split
Definition: MVATrainer.cc:139
How EventSelector::AcceptEvent() decides whether to accept an event for output otherwise it is excluding the probing of A single or multiple positive and the trigger will pass if any such matching triggers are PASS or EXCEPTION[A criterion thatmatches no triggers at all is detected and causes a throw.] A single negative with an expectation of appropriate bit checking in the decision and the trigger will pass if any such matching triggers are FAIL or EXCEPTION A wildcarded negative criterion that matches more than one trigger in the trigger list("!*","!HLTx*"if it matches 2 triggers or more) will accept the event if all the matching triggers are FAIL.It will reject the event if any of the triggers are PASS or EXCEPTION(this matches the behavior of"!*"before the partial wildcard feature was incorporated).Triggers which are in the READY state are completely ignored.(READY should never be returned since the trigger paths have been run
def executeAndRemember(self, command)
def prepare_PAT(self, sequence="miniAOD")
nextScheduleIsConditional
put the filtering path in the schedule