BulkRun.py

Go to the documentation of this file.

# Copyright (C) 2002-2020 CERN for the benefit of the ATLAS collaboration
 
from __future__ import print_function
 
from glob import glob
import subprocess
import os
import pickle
import sys
 
class BulkRun :
  """Run on multiple pedestal files
 
  Used to run on multiple pedestal bytestream files located somewhere on disk.
  Puts results into seperate directories by run-number. It extracts pertinent
  information from the file names, which are expected to be of the form:
 
  data10_calib.00157081.calibration_pedCSC.daq.RAW._lb0000._CSC-EB._0001.data
 
  run(numToRun) - process NumToRun (at most) runs out of those not yet run out of. This is the primary interface. 
  ReadProcessedFilesList() - find all files already run on from a local text file
  AddProcessedFiles() - save newly run on files to disk
  FindFiles() - Find pattern and run number of a set of files that have not yet been run on
  RunAthena() - run athena job for a set of runs
 
  """
  def __init__(
      self,
      inputPattern = "/raid02/schernau/ped/csc/csc/*.data" ,
      processedFilesList = "ProcessedFiles.list",
      outputDirBase = "/raid02/lampen/datasets/csc/PedProcessing2",
      debug = False,
      allowDirOverwrite = False,
      ):
    """initialize internal variables"""
    self.InputPattern = inputPattern
    self.ProcessedFilesList = processedFilesList
    self.OutputDirBase = outputDirBase
    self.debug = debug
    self.AllowDirOverwrite = allowDirOverwrite
    print (self.InputPattern)
 
  def run(self, numToRun = 10) :
    """
    Run over all run numbers, find files for each, and submit each set to
    CscCalcPedMon.py
    """
 
    print ("Running on " + str(numToRun) + " runsSet.")
 
 
    runNumbers = []
    for runCnt in range(numToRun) :
      print (">>>>Running on runSet " + str(runCnt+1) + " of " + str(numToRun))
      pattern,runNumber = self.FindFiles()
      if(pattern != ""):
        #have files to run on
        runNumbers += [runNumber]
        self.RunAthena(pattern,runNumber)
      else:
        print ("No more unprocessed files. Congrats!")
        print ("N runs done: " + str(runCnt +1))
        print (runNumbers)
        return
    print ("finished all " + str(numToRun) )
    print ("Run numbers include:")
    print (runNumbers)
    print()
    print ("All Processed files:" )
    print (self.ReadProcessedFilesList())
 
  #Read list of previously processed files
  def ReadProcessedFilesList(self):
 
    ProcessedFiles = []
 
    #Get processed files
    f = open(self.ProcessedFilesList,"rb") 
    ProcessedFiles = pickle.load(f)
    f.close()
     
    #Remove newline character from each filename
 
    #for index in range(len(ProcessedFiles)) : 
    #  file = ProcessedFiles[index]
    #  ProcessedFiles[index] = file[0:len(file)-1] 
 
    print ('Processed String: ')
    print (ProcessedFiles)
 
    return ProcessedFiles
 
  def AddProcessedFiles(self,newFiles):
    """Save new processed files to disk"""
    ProcessedFiles = self.ReadProcessedFilesList()
    ProcessedFiles += newFiles
    f = open(self.ProcessedFilesList,"wb")
    pickle.dump(ProcessedFiles,f)
    f.close()
 
    return True
 
 
  #Read list of previously processed files
  def FindFiles(self):
 
    #Initial setup
    FoundUnprocessedFile = False
 
    #Get processed file list
    ProcessedFiles = self.ReadProcessedFilesList()
 
    #Get list of files in input dir
    print ("Input pattern: " + self.InputPattern)
    inputFiles = glob(self.InputPattern)
 
    if(self.debug):
      print()
      print ("Searching for file")
      print ("InputPattern: " + self.InputPattern)
      #print ("inputFiles: ")
      #print (inputFiles)
 
    pattern = ""
    runNumber = ""
 
    #Loop through list until find file that is 
    for file in inputFiles:
      if not ProcessedFiles.count(file):
 
        index = file.find("data10")
        if(index == -1):
          index = file.find("data11")
        if(index == -1):
          print ("ERROR! Index of -1!")
          raise Exception("Index error") 
        FoundUnprocessedFile = True
        pattern = file[0:index + 22] + "*" #includes run number
        runNumber = file[index + 13: index + 21]
 
    if(not FoundUnprocessedFile):
      return "",""
 
    if(self.debug) : 
      print ("Found unprocessed file with pattern: " + pattern)
      print ("This includes files:")
      print (glob(pattern))
 
    return pattern, runNumber
 
  def RunAthena(self, pattern, runNumber):
    """Run athena on a particular set of files matching pattern"""
    outputDirPath = self.OutputDirBase + "/" + runNumber
      
    if(self.AllowDirOverwrite):
      subprocess.call("rm -rf " + outputDirPath)
 
    print ("Making directory" + outputDirPath)
    #Create output directory
    os.mkdir(outputDirPath,0o755)
 
    #Change directory to output directory
    #os.chdir(outputDirPath)
 
    #Athena options
    athOpt = "outputPre='" + outputDirPath +"/" + runNumber 
    athOpt += "';inputPat='" + pattern 
    athOpt += "';reportPrefix='runNumber = " 
    athOpt += runNumber + "'"
 
    #athena arguments
    athArgs = ["athena.py", "-c", athOpt, "CscCalcPedMon.py"]
 
    #Output log file
    logFile = open(outputDirPath + "/run.log","w")
 
    print()
    print ("**************************************")
    print ("Starting running on run " + str(runNumber))
    sys.stdout.flush()
    subprocess.Popen(athArgs,stdout=logFile,stderr=subprocess.STDOUT).wait()
    print ("Finished run " + str(runNumber))
    print ("**************************************")
    print()
 
    logFile.close()
 
 
    #Add files we just ran on to file list
    newProcessedFiles = glob(pattern)
    self.AddProcessedFiles(newProcessedFiles)