Python LumiList.getRuns方法代码示例

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]
    def adjustLumisForCompletion(self, task, unprocessed):
        """Sets the run, lumi information in the task information for the
        completion jobs.  Returns True if completion jobs are needed,
        otherwise False.
        """
        missingDir = "automatic_splitting/missing_lumis/" #TODO in ServerUtilities to be shared with PJ

        try:
            available = set(os.listdir(missingDir)) & unprocessed
        except OSError:
            available = set()

        failed = set(self.failedJobs) & unprocessed

        if len(available) == 0 and len(failed) == 0:
            return False

        missing = LumiList()
        for missingFile in available:
            with open(os.path.join(missingDir, missingFile)) as fd:
                self.logger.info("Adding missing lumis from job %s", missingFile)
                missing = missing + LumiList(compactList=literal_eval(fd.read()))
        for failedId in failed:
            f = None
            try:
                tmpdir = tempfile.mkdtemp()
                f = tarfile.open("run_and_lumis.tar.gz")
                fn = "job_lumis_{0}.json".format(failedId)
                f.extract(fn, path=tmpdir)
                with open(os.path.join(tmpdir, fn)) as fd:
                    injson = json.load(fd)
                    missing = missing + LumiList(compactList=injson)
                    self.logger.info("Adding lumis from failed job %s", failedId)
            finally:
                if f:
                    f.close()
                shutil.rmtree(tmpdir)
        missing_compact = missing.getCompactList()
        runs = missing.getRuns()
        # Compact list is like
        # {
        # '1': [[1, 33], [35, 35], [37, 47], [49, 75], [77, 130], [133, 136]],
        # '2':[[1,45],[50,80]]
        # }
        # Now we turn lumis it into something like:
        # lumis=['1, 33, 35, 35, 37, 47, 49, 75, 77, 130, 133, 136','1,45,50,80']
        # which is the format expected by buildLumiMask in the splitting algorithm
        lumis = [",".join(str(l) for l in functools.reduce(lambda x, y:x + y, missing_compact[run])) for run in runs]

        task['tm_split_args']['runs'] = runs
        task['tm_split_args']['lumis'] = lumis

        return True

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]

#.........这里部分代码省略.........
        ## handled, we are sure that if we reached this point it will not raise EnvironmentException.
        ## But otherwise we should take this into account.
        with UserTarball(name=tarFilename, logger=self.logger, config=self.config) as tb:
            inputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'inputFiles', [])]
            tb.addFiles(userFiles=inputFiles, cfgOutputName=cfgOutputName)
            configArguments['adduserfiles'] = [os.path.basename(f) for f in inputFiles]
            try:
                uploadResult = tb.upload(filecacheurl = filecacheurl)
            except HTTPException as hte:
                if 'X-Error-Info' in hte.headers:
                    reason = hte.headers['X-Error-Info']
                    reason_re = re.compile(r'\AFile size is ([0-9]*)B\. This is bigger than the maximum allowed size of ([0-9]*)B\.$')
                    re_match = reason_re.match(reason)
                    if re_match:
                        ISBSize = int(re_match.group(1))
                        ISBSizeLimit = int(re_match.group(2))
                        reason  = "%sError%s:" % (colors.RED, colors.NORMAL)
                        reason += " Input sanbox size is ~%sMB. This is bigger than the maximum allowed size of %sMB." % (ISBSize/1024/1024, ISBSizeLimit/1024/1024)
                        ISBContent = sorted(tb.content, reverse=True)
                        biggestFileSize = ISBContent[0][0]
                        ndigits = int(math.ceil(math.log(biggestFileSize+1, 10)))
                        reason += "\nInput sanbox content sorted by size[Bytes]:"
                        for (size, name) in ISBContent:
                            reason += ("\n%" + str(ndigits) + "s\t%s") % (size, name)
                        raise ClientException(reason)
                raise hte
            except Exception as e:
                msg = ("Impossible to calculate the checksum of the sandbox tarball.\nError message: %s.\n"
                       "More details can be found in %s" % (e, self.logger.logfile))
                LOGGERS['CRAB3'].exception(msg) #the traceback is only printed into the logfile
                raise ClientException(msg)

        debugFilesUploadResult = None
        with UserTarball(name=debugTarFilename, logger=self.logger, config=self.config) as dtb:
            dtb.addMonFiles()
            try:
                debugFilesUploadResult = dtb.upload(filecacheurl = filecacheurl)
            except Exception as e:
                msg = ("Problem uploading debug_files.tar.gz.\nError message: %s.\n"
                       "More details can be found in %s" % (e, self.logger.logfile))
                LOGGERS['CRAB3'].exception(msg) #the traceback is only printed into the logfile

        configArguments['cacheurl'] = filecacheurl
        configArguments['cachefilename'] = "%s.tar.gz" % uploadResult
        if debugFilesUploadResult is not None:
            configArguments['debugfilename'] = "%s.tar.gz" % debugFilesUploadResult
        self.logger.debug("Result uploading input files: %(cachefilename)s " % configArguments)

        # Upload list of user-defined input files to process as the primary input
        userFilesList = getattr(self.config.Data, 'userInputFiles', None)
        if userFilesList:
            self.logger.debug("Attaching list of user-specified primary input files.")
            userFilesList = map(string.strip, userFilesList)
            userFilesList = [file for file in userFilesList if file]
            if len(userFilesList) != len(set(userFilesList)):
                msg  = "%sWarning%s:" % (colors.RED, colors.NORMAL)
                msg += " CRAB configuration parameter Data.userInputFiles contains duplicated entries."
                msg += " Duplicated entries will be removed."
                self.logger.warning(msg)
            configArguments['userfiles'] = set(userFilesList)
            configArguments['primarydataset'] = getattr(self.config.Data, 'outputPrimaryDataset', 'CRAB_UserFiles')

        lumi_mask_name = getattr(self.config.Data, 'lumiMask', None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % (lumi_mask_name))
            try:
                lumi_list = getLumiList(lumi_mask_name, logger = self.logger)
            except ValueError as ex:
                msg  = "%sError%s:" % (colors.RED, colors.NORMAL)
                msg += " Failed to load lumi mask %s : %s" % (lumi_mask_name, ex)
                raise ConfigurationException(msg)
        run_ranges = getattr(self.config.Data, 'runRange', None)
        if run_ranges:
            run_ranges_is_valid = re.match('^\d+((?!(-\d+-))(\,|\-)\d+)*$', run_ranges)
            if run_ranges_is_valid:
                run_list = getRunList(run_ranges)
                if lumi_list:
                    lumi_list.selectRuns(run_list)
                    if not lumi_list:
                        msg = "Invalid CRAB configuration: The intersection between the lumi mask and the run range is null."
                        raise ConfigurationException(msg)
                else:
                    if len(run_list) > 50000:
                        msg  = "CRAB configuration parameter Data.runRange includes %s runs." % str(len(run_list))
                        msg += " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                        raise ConfigurationException(msg)
                    lumi_list = LumiList(runs = run_list)
            else:
                msg = "Invalid CRAB configuration: Parameter Data.runRange should be a comma separated list of integers or (inclusive) ranges. Example: '12345,99900-99910'"
                raise ConfigurationException(msg)
        if lumi_list:
            configArguments['runs'] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments['lumis'] = [str(reduce(lambda x,y: x+y, lumi_mask[run]))[1:-1].replace(' ','') for run in configArguments['runs']]

        configArguments['jobtype'] = 'Analysis'

        return tarFilename, configArguments

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]
    def run(self, filecacheurl = None):
        """
        Override run() for JobType
        """

        taskDict, webdir = self.getTaskDict()
        addoutputfiles = literal_eval(getColumn(taskDict, 'tm_outfiles'))
        tfileoutfiles = literal_eval(getColumn(taskDict, 'tm_tfile_outfiles'))
        edmoutfiles = literal_eval(getColumn(taskDict, 'tm_edm_outfiles'))
        jobarch = getColumn(taskDict, 'tm_job_arch')
        jobsw = getColumn(taskDict, 'tm_job_sw')

        sandboxFilename = os.path.join(self.workdir, 'sandbox.tar.gz')
        getFileFromURL(webdir + '/sandbox.tar.gz', sandboxFilename, self.proxyfilename)

        configArguments = {'addoutputfiles' : addoutputfiles,
                           'tfileoutfiles' : tfileoutfiles,
                           'edmoutfiles' : edmoutfiles,
                           'jobarch' : jobarch,
                           'jobsw' : jobsw,
                          }

        # Maybe the user wnat to change the dataset
        if getattr(self.config.Data, 'inputDataset', None):
            configArguments['inputdata'] = self.config.Data.inputDataset

        ufc = CRABClient.Emulator.getEmulator('ufc')({'endpoint' : filecacheurl, "pycurl": True})
        result = ufc.upload(sandboxFilename, excludeList = NEW_USER_SANDBOX_EXCLUSIONS)
        if 'hashkey' not in result:
            self.logger.error("Failed to upload source files: %s" % str(result))
            raise CachefileNotFoundException

        configArguments['cacheurl'] = filecacheurl
        configArguments['cachefilename'] = "%s.tar.gz" % str(result['hashkey'])

        # Upload list of user-defined input files to process as the primary input
        userFilesList = getattr(self.config.Data, 'userInputFiles', None)
        if userFilesList:
            self.logger.debug("Attaching list of user-specified primary input files.")
            userFilesList = map(string.strip, userFilesList)
            userFilesList = [file for file in userFilesList if file]
            if len(userFilesList) != len(set(userFilesList)):
                msg  = "%sWarning%s:" % (colors.RED, colors.NORMAL)
                msg += " CRAB configuration parameter Data.userInputFiles contains duplicated entries."
                msg += " Duplicated entries will be removed."
                self.logger.warning(msg)
            configArguments['userfiles'] = set(userFilesList)
            configArguments['primarydataset'] = getattr(self.config.Data, 'outputPrimaryDataset', 'CRAB_UserFiles')

        lumi_mask_name = getattr(self.config.Data, 'lumiMask', None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % (lumi_mask_name))
            try:
                lumi_list = getLumiList(lumi_mask_name, logger = self.logger)
            except ValueError as ex:
                msg  = "%sError%s:" % (colors.RED, colors.NORMAL)
                msg += " Failed to load lumi mask %s : %s" % (lumi_mask_name, ex)
                raise ConfigurationException(msg)
        run_ranges = getattr(self.config.Data, 'runRange', None)
        if run_ranges:
            run_ranges_is_valid = re.match('^\d+((?!(-\d+-))(\,|\-)\d+)*$', run_ranges)
            if run_ranges_is_valid:
                run_list = getRunList(run_ranges)
                if lumi_list:
                    lumi_list.selectRuns(run_list)
                    if not lumi_list:
                        msg = "Invalid CRAB configuration: The intersection between the lumi mask and the run range is null."
                        raise ConfigurationException(msg)
                else:
                    if len(run_list) > 50000:
                        msg  = "CRAB configuration parameter Data.runRange includes %s runs." % str(len(run_list))
                        msg += " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                        raise ConfigurationException(msg)
                    lumi_list = LumiList(runs = run_list)
            else:
                msg = "Invalid CRAB configuration: Parameter Data.runRange should be a comma separated list of integers or (inclusive) ranges. Example: '12345,99900-99910'"
                raise ConfigurationException(msg)
        if lumi_list:
            configArguments['runs'] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments['lumis'] = [str(reduce(lambda x,y: x+y, lumi_mask[run]))[1:-1].replace(' ','') for run in configArguments['runs']]

        configArguments['jobtype'] = 'Analysis'

        return sandboxFilename, configArguments

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]

#.........这里部分代码省略.........
        # Get SCRAM environment
        scram = ScramEnvironment(logger=self.logger)

        configArguments.update({'jobarch'    : scram.scramArch,
                                'jobsw' : scram.cmsswVersion, })

        # Build tarball
        if self.workdir:
            tarUUID =  PandaInterface.wrappedUuidGen()
            self.logger.debug('UNIQUE NAME: tarUUID %s ' % tarUUID)
            if len(tarUUID):
                tarFilename   = os.path.join(self.workdir, tarUUID +'default.tgz')
                cfgOutputName = os.path.join(self.workdir, 'CMSSW_cfg.py')
            else:
                raise EnvironmentException('Problem with uuidgen while preparing for Sandbox upload.')
        else:
            _dummy, tarFilename   = tempfile.mkstemp(suffix='.tgz')
            _dummy, cfgOutputName = tempfile.mkstemp(suffix='_cfg.py')

        #configArguments['userisburl'] = 'https://'+ self.config.General.ufccacheUrl + '/crabcache/file?hashkey=' + uploadResults['hashkey']#XXX hardcoded
        #configArguments['userisburl'] = 'INSERTuserisburl'#XXX hardcoded
        if getattr(self.config.Data, 'inputDataset', None):
            configArguments['inputdata'] = self.config.Data.inputDataset
#        configArguments['ProcessingVersion'] = getattr(self.config.Data, 'processingVersion', None)

        # Create CMSSW config
        self.logger.debug("self.config: %s" % self.config)
        self.logger.debug("self.config.JobType.psetName: %s" % self.config.JobType.psetName)
        cmsswCfg = CMSSWConfig(config=self.config, logger=self.logger,
                               userConfig=self.config.JobType.psetName)

        ## Interogate CMSSW config and user config for output file names. For now no use for EDM files or TFiles here.
        edmfiles, tfiles = cmsswCfg.outputFiles()
        addoutputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'outputFiles', []) if re.sub(r'^file:', '', file) not in edmfiles+tfiles]
        self.logger.debug("The following EDM output files will be collected: %s" % edmfiles)
        self.logger.debug("The following TFile output files will be collected: %s" % tfiles)
        self.logger.debug("The following user output files will be collected: %s" % addoutputFiles)
        configArguments['edmoutfiles'] = edmfiles
        configArguments['tfileoutfiles'] = tfiles
        configArguments['addoutputfiles'].extend(addoutputFiles)

        # Write out CMSSW config
        cmsswCfg.writeFile(cfgOutputName)

        with UserTarball(name=tarFilename, logger=self.logger, config=self.config) as tb:
            inputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'inputFiles', [])]
            tb.addFiles(userFiles=inputFiles, cfgOutputName=cfgOutputName)
            configArguments['adduserfiles'] = [os.path.basename(f) for f in inputFiles]
            uploadResults = tb.upload()

        self.logger.debug("Result uploading input files: %s " % str(uploadResults))
        configArguments['cachefilename'] = uploadResults[1]
        configArguments['cacheurl'] = uploadResults[0]
        isbchecksum = uploadResults[2]

        # Upload list of user-defined input files to process as the primary input
        userFileName = getattr(self.config.Data, 'userInputFile', None)
        if userFileName:
            self.logger.debug("Attaching a list of user-specified primary input files from %s." % userFileName)
            fnames = []
            for fname in open(userFileName).readlines():
                fnames.append(fname.strip())
            configArguments['userfiles'] = filter(lambda x: x, fnames) #removing whitelines and empty objects

            primDS = getattr(self.config.Data, 'primaryDataset', None)
            if primDS:
                # Normalizes "foo/bar" and "/foo/bar" to "/foo/bar"
                primDS = "/" + os.path.join(*primDS.split("/"))
                if not re.match("/%(primDS)s.*" % lfnParts, primDS):
                    self.logger.warning("Invalid primary dataset name %s for private MC; publishing may fail" % primDS)
                configArguments['inputdata'] = primDS
            else:
                configArguments['inputdata'] = getattr(self.config.Data, 'inputDataset', '/CRAB_UserFiles')

        lumi_mask_name = getattr(self.config.Data, 'lumiMask', None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % lumi_mask_name)
            lumi_list = getLumiList(lumi_mask_name, logger = self.logger)
        run_ranges = getattr(self.config.Data, 'runRange', None)
        run_ranges_is_valid = run_ranges is not None and isinstance(run_ranges, str) and re.match('^\d+((?!(-\d+-))(\,|\-)\d+)*$', run_ranges)
        if run_ranges_is_valid:
            run_list = getRunList(run_ranges)
            if lumi_list:
                lumi_list.selectRuns(run_list)
            else:
                if len(run_list) > 50000:
                    msg  = "Data.runRange includes %s runs." % str(len(run_list))
                    msg += " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                    raise ConfigurationException(msg)
                lumi_list = LumiList(runs = run_list)
        if lumi_list:
            configArguments['runs'] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments['lumis'] = [str(reduce(lambda x,y: x+y, lumi_mask[run]))[1:-1].replace(' ','') for run in configArguments['runs']]

        configArguments['jobtype'] = 'Analysis'

        return tarFilename, configArguments, isbchecksum

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]

#.........这里部分代码省略.........
        configArguments["addoutputfiles"].extend(addoutputFiles)
        ## Give warning message in case no output file was detected in the CMSSW pset
        ## nor was any specified in the CRAB configuration.
        if (
            not configArguments["edmoutfiles"]
            and not configArguments["tfileoutfiles"]
            and not configArguments["addoutputfiles"]
        ):
            msg = "%sWarning%s:" % (colors.RED, colors.NORMAL)
            if getattr(
                self.config.JobType,
                "disableAutomaticOutputCollection",
                getParamDefaultValue("JobType.disableAutomaticOutputCollection"),
            ):
                msg += " Automatic detection of output files in the CMSSW configuration is disabled from the CRAB configuration"
                msg += " and no output file was explicitly specified in the CRAB configuration."
            else:
                msg += " CRAB could not detect any output file in the CMSSW configuration"
                msg += " nor was any explicitly specified in the CRAB configuration."
            msg += " Hence CRAB will not collect any output file from this task."
            self.logger.warning(msg)

        ## UserTarball calls ScramEnvironment which can raise EnvironmentException.
        ## Since ScramEnvironment is already called above and the exception is not
        ## handled, we are sure that if we reached this point it will not raise EnvironmentException.
        ## But otherwise we should take this into account.
        with UserTarball(name=tarFilename, logger=self.logger, config=self.config) as tb:
            inputFiles = [re.sub(r"^file:", "", file) for file in getattr(self.config.JobType, "inputFiles", [])]
            tb.addFiles(userFiles=inputFiles, cfgOutputName=cfgOutputName)
            configArguments["adduserfiles"] = [os.path.basename(f) for f in inputFiles]
            uploadResults = tb.upload(filecacheurl=filecacheurl)

        self.logger.debug("Result uploading input files: %s " % str(uploadResults))
        configArguments["cacheurl"] = filecacheurl
        configArguments["cachefilename"] = uploadResults[0]
        isbchecksum = uploadResults[1]

        # Upload list of user-defined input files to process as the primary input
        userFilesList = getattr(self.config.Data, "userInputFiles", None)
        if userFilesList:
            self.logger.debug("Attaching list of user-specified primary input files.")
            userFilesList = map(string.strip, userFilesList)
            userFilesList = [file for file in userFilesList if file]
            if len(userFilesList) != len(set(userFilesList)):
                msg = "%sWarning%s:" % (colors.RED, colors.NORMAL)
                msg += " CRAB configuration parameter Data.userInputFiles contains duplicated entries."
                msg += " Duplicated entries will be removed."
                self.logger.warning(msg)
            configArguments["userfiles"] = set(userFilesList)
            ## Get the user-specified primary dataset name.
            primaryDataset = getattr(self.config.Data, "primaryDataset", "CRAB_UserFiles")
            # Normalizes "foo/bar" and "/foo/bar" to "/foo/bar"
            primaryDataset = "/" + os.path.join(*primaryDataset.split("/"))
            if not re.match("/%(primDS)s.*" % (lfnParts), primaryDataset):
                self.logger.warning("Invalid primary dataset name %s; publication may fail." % (primaryDataset))
            configArguments["inputdata"] = primaryDataset

        lumi_mask_name = getattr(self.config.Data, "lumiMask", None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % (lumi_mask_name))
            try:
                lumi_list = getLumiList(lumi_mask_name, logger=self.logger)
            except ValueError as ex:
                msg = "%sError%s:" % (colors.RED, colors.NORMAL)
                msg += " Failed to load lumi mask %s : %s" % (lumi_mask_name, ex)
                raise ConfigurationException(msg)
        run_ranges = getattr(self.config.Data, "runRange", None)
        if run_ranges:
            run_ranges_is_valid = re.match("^\d+((?!(-\d+-))(\,|\-)\d+)*$", run_ranges)
            if run_ranges_is_valid:
                run_list = getRunList(run_ranges)
                if lumi_list:
                    lumi_list.selectRuns(run_list)
                    if not lumi_list:
                        msg = "Invalid CRAB configuration: The intersection between the lumi mask and the run range is null."
                        raise ConfigurationException(msg)
                else:
                    if len(run_list) > 50000:
                        msg = "CRAB configuration parameter Data.runRange includes %s runs." % str(len(run_list))
                        msg += (
                            " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                        )
                        raise ConfigurationException(msg)
                    lumi_list = LumiList(runs=run_list)
            else:
                msg = "Invalid CRAB configuration: Parameter Data.runRange should be a comma separated list of integers or (inclusive) ranges. Example: '12345,99900-99910'"
                raise ConfigurationException(msg)
        if lumi_list:
            configArguments["runs"] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments["lumis"] = [
                str(reduce(lambda x, y: x + y, lumi_mask[run]))[1:-1].replace(" ", "")
                for run in configArguments["runs"]
            ]

        configArguments["jobtype"] = "Analysis"

        return tarFilename, configArguments, isbchecksum

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]

#.........这里部分代码省略.........

        configArguments.update({'jobarch'    : scram.scramArch,
                                'jobsw' : scram.cmsswVersion, })

        # Build tarball
        if self.workdir:
            tarUUID =  PandaInterface.wrappedUuidGen()
            self.logger.debug('UNIQUE NAME: tarUUID %s ' % tarUUID)
            if len(tarUUID):
                tarFilename   = os.path.join(self.workdir, tarUUID +'default.tgz')
                cfgOutputName = os.path.join(self.workdir, 'CMSSW_cfg.py')
            else:
                raise EnvironmentException('Problem with uuidgen while preparing for Sandbox upload.')
        else:
            _dummy, tarFilename   = tempfile.mkstemp(suffix='.tgz')
            _dummy, cfgOutputName = tempfile.mkstemp(suffix='_cfg.py')

        if getattr(self.config.Data, 'inputDataset', None):
            configArguments['inputdata'] = self.config.Data.inputDataset
#        configArguments['ProcessingVersion'] = getattr(self.config.Data, 'processingVersion', None)

        # Create CMSSW config
        self.logger.debug("self.config: %s" % self.config)
        self.logger.debug("self.config.JobType.psetName: %s" % self.config.JobType.psetName)
        cmsswCfg = CMSSWConfig(config=self.config, logger=self.logger,
                               userConfig=self.config.JobType.psetName)

        ## Interogate CMSSW config and user config for output file names. For now no use for EDM files or TFiles here.
        edmfiles, tfiles = cmsswCfg.outputFiles()
        addoutputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'outputFiles', []) if re.sub(r'^file:', '', file) not in edmfiles+tfiles]
        self.logger.debug("The following EDM output files will be collected: %s" % edmfiles)
        self.logger.debug("The following TFile output files will be collected: %s" % tfiles)
        self.logger.debug("The following user output files will be collected: %s" % addoutputFiles)
        configArguments['edmoutfiles'] = edmfiles
        configArguments['tfileoutfiles'] = tfiles
        configArguments['addoutputfiles'].extend(addoutputFiles)

        # Write out CMSSW config
        cmsswCfg.writeFile(cfgOutputName)

        ## UserTarball calls ScramEnvironment which can raise EnvironmentException.
        ## Since ScramEnvironment is already called above and the exception is not
        ## handled, we are sure that if we reached this point it will not raise EnvironmentException.
        ## But otherwise we should take this into account.
        with UserTarball(name=tarFilename, logger=self.logger, config=self.config) as tb:
            inputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'inputFiles', [])]
            tb.addFiles(userFiles=inputFiles, cfgOutputName=cfgOutputName)
            configArguments['adduserfiles'] = [os.path.basename(f) for f in inputFiles]
            uploadResults = tb.upload(filecacheurl = filecacheurl)

        self.logger.debug("Result uploading input files: %s " % str(uploadResults))
        configArguments['cacheurl'] = filecacheurl
        configArguments['cachefilename'] = uploadResults[0]
        isbchecksum = uploadResults[1]

        # Upload list of user-defined input files to process as the primary input
        userFilesList = getattr(self.config.Data, 'userInputFiles', None)
        if userFilesList:
            self.logger.debug("Attaching list of user-specified primary input files.")
            userFilesList = map(string.strip, userFilesList)
            userFilesList = [file for file in userFilesList if file]
            if len(userFilesList) != len(set(userFilesList)):
                msg  = "%sWarning%s: CRAB configuration parameter Data.userInputFiles contains duplicated entries." % (colors.RED, colors.NORMAL)
                msg += " Duplicated entries will be removed."    
                self.logger.warning(msg)
            configArguments['userfiles'] = set(userFilesList)
            ## Get the user-specified primary dataset name.
            primaryDataset = getattr(self.config.Data, 'primaryDataset', 'CRAB_UserFiles')
            # Normalizes "foo/bar" and "/foo/bar" to "/foo/bar"
            primaryDataset = "/" + os.path.join(*primaryDataset.split("/"))
            if not re.match("/%(primDS)s.*" % (lfnParts), primaryDataset):
                self.logger.warning("Invalid primary dataset name %s; publication may fail." % (primaryDataset))
            configArguments['inputdata'] = primaryDataset

        lumi_mask_name = getattr(self.config.Data, 'lumiMask', None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % lumi_mask_name)
            lumi_list = getLumiList(lumi_mask_name, logger = self.logger)
        run_ranges = getattr(self.config.Data, 'runRange', None)
        run_ranges_is_valid = run_ranges is not None and isinstance(run_ranges, str) and re.match('^\d+((?!(-\d+-))(\,|\-)\d+)*$', run_ranges)
        if run_ranges_is_valid:
            run_list = getRunList(run_ranges)
            if lumi_list:
                lumi_list.selectRuns(run_list)
            else:
                if len(run_list) > 50000:
                    msg  = "Data.runRange includes %s runs." % str(len(run_list))
                    msg += " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                    raise ConfigurationException(msg)
                lumi_list = LumiList(runs = run_list)
        if lumi_list:
            configArguments['runs'] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments['lumis'] = [str(reduce(lambda x,y: x+y, lumi_mask[run]))[1:-1].replace(' ','') for run in configArguments['runs']]

        configArguments['jobtype'] = 'Analysis'

        return tarFilename, configArguments, isbchecksum

# 需要导入模块: from WMCore.DataStructs.LumiList import LumiList [as 别名]
# 或者: from WMCore.DataStructs.LumiList.LumiList import getRuns [as 别名]

#.........这里部分代码省略.........
        ## in the sense that a second loading of the same pset may not produce the same
        ## result. Therefore there is a cache in CMSSWConfig to avoid loading any CMSSW
        ## pset twice. However, some "complicated" psets seem to evade the caching.
        ## Thus, to be safe, keep the CMSSWConfig instance in a class variable, so that
        ## it can be reused later if wanted (for example, in PrivateMC when checking if
        ## the pset has an LHE source) instead of having to load the pset again.
        ## As for what does "complicated" psets mean, Daniel Riley said that there are
        ## some psets where one module modifies the configuration from another module.
        self.cmsswCfg = CMSSWConfig(config=self.config, logger=self.logger,
                                    userConfig=self.config.JobType.psetName)

        ## Interrogate the CMSSW pset for output files (only output files produced by
        ## PoolOutputModule or TFileService are identified automatically). Do this
        ## automatic detection even if JobType.disableAutomaticOutputCollection = True,
        ## so that we can still classify the output files in EDM, TFile and additional
        ## output files in the Task DB (and the job ad).
        ## TODO: Do we really need this classification at all? cmscp and PostJob read
        ## the FJR to know if an output file is EDM, TFile or other.
        edmfiles, tfiles = self.cmsswCfg.outputFiles()
        ## If JobType.disableAutomaticOutputCollection = True, ignore the EDM and TFile
        ## output files that are not listed in JobType.outputFiles.
        if getattr(self.config.JobType, 'disableAutomaticOutputCollection', getParamDefaultValue('JobType.disableAutomaticOutputCollection')):
            outputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'outputFiles', [])]
            edmfiles = [file for file in edmfiles if file in outputFiles]
            tfiles = [file for file in tfiles if file in outputFiles]
        ## Get the list of additional output files that have to be collected as given
        ## in JobType.outputFiles, but remove duplicates listed already as EDM files or
        ## TFiles.
        addoutputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'outputFiles', []) if re.sub(r'^file:', '', file) not in edmfiles+tfiles]
        self.logger.debug("The following EDM output files will be collected: %s" % edmfiles)
        self.logger.debug("The following TFile output files will be collected: %s" % tfiles)
        self.logger.debug("The following user output files will be collected: %s" % addoutputFiles)
        configArguments['edmoutfiles'] = edmfiles
        configArguments['tfileoutfiles'] = tfiles
        configArguments['addoutputfiles'].extend(addoutputFiles)

        # Write out CMSSW config
        self.cmsswCfg.writeFile(cfgOutputName)

        ## UserTarball calls ScramEnvironment which can raise EnvironmentException.
        ## Since ScramEnvironment is already called above and the exception is not
        ## handled, we are sure that if we reached this point it will not raise EnvironmentException.
        ## But otherwise we should take this into account.
        with UserTarball(name=tarFilename, logger=self.logger, config=self.config) as tb:
            inputFiles = [re.sub(r'^file:', '', file) for file in getattr(self.config.JobType, 'inputFiles', [])]
            tb.addFiles(userFiles=inputFiles, cfgOutputName=cfgOutputName)
            configArguments['adduserfiles'] = [os.path.basename(f) for f in inputFiles]
            uploadResults = tb.upload(filecacheurl = filecacheurl)

        self.logger.debug("Result uploading input files: %s " % str(uploadResults))
        configArguments['cacheurl'] = filecacheurl
        configArguments['cachefilename'] = uploadResults[0]
        isbchecksum = uploadResults[1]

        # Upload list of user-defined input files to process as the primary input
        userFilesList = getattr(self.config.Data, 'userInputFiles', None)
        if userFilesList:
            self.logger.debug("Attaching list of user-specified primary input files.")
            userFilesList = map(string.strip, userFilesList)
            userFilesList = [file for file in userFilesList if file]
            if len(userFilesList) != len(set(userFilesList)):
                msg  = "%sWarning%s:" % (colors.RED, colors.NORMAL)
                msg += " CRAB configuration parameter Data.userInputFiles contains duplicated entries."
                msg += " Duplicated entries will be removed."    
                self.logger.warning(msg)
            configArguments['userfiles'] = set(userFilesList)
            ## Get the user-specified primary dataset name.
            primaryDataset = getattr(self.config.Data, 'primaryDataset', 'CRAB_UserFiles')
            # Normalizes "foo/bar" and "/foo/bar" to "/foo/bar"
            primaryDataset = "/" + os.path.join(*primaryDataset.split("/"))
            if not re.match("/%(primDS)s.*" % (lfnParts), primaryDataset):
                self.logger.warning("Invalid primary dataset name %s; publication may fail." % (primaryDataset))
            configArguments['inputdata'] = primaryDataset

        lumi_mask_name = getattr(self.config.Data, 'lumiMask', None)
        lumi_list = None
        if lumi_mask_name:
            self.logger.debug("Attaching lumi mask %s to the request" % (lumi_mask_name))
            lumi_list = getLumiList(lumi_mask_name, logger = self.logger)
        run_ranges = getattr(self.config.Data, 'runRange', None)
        run_ranges_is_valid = run_ranges is not None and isinstance(run_ranges, str) and re.match('^\d+((?!(-\d+-))(\,|\-)\d+)*$', run_ranges)
        if run_ranges_is_valid:
            run_list = getRunList(run_ranges)
            if lumi_list:
                lumi_list.selectRuns(run_list)
            else:
                if len(run_list) > 50000:
                    msg  = "CRAB configuration parameter Data.runRange includes %s runs." % str(len(run_list))
                    msg += " When Data.lumiMask is not specified, Data.runRange can not include more than 50000 runs."
                    raise ConfigurationException(msg)
                lumi_list = LumiList(runs = run_list)
        if lumi_list:
            configArguments['runs'] = lumi_list.getRuns()
            ## For each run we encode the lumis as a string representing a list of integers: [[1,2],[5,5]] ==> '1,2,5,5'
            lumi_mask = lumi_list.getCompactList()
            configArguments['lumis'] = [str(reduce(lambda x,y: x+y, lumi_mask[run]))[1:-1].replace(' ','') for run in configArguments['runs']]

        configArguments['jobtype'] = 'Analysis'

        return tarFilename, configArguments, isbchecksum