giuliomoro@0: #!/usr/bin/python
giuliomoro@0: 
giuliomoro@0: import xml.etree.ElementTree as ET
giuliomoro@0: import os
giuliomoro@0: import sys
giuliomoro@0: import csv
giuliomoro@0: 
giuliomoro@0: # COMMAND LINE ARGUMENTS
giuliomoro@0: 
giuliomoro@0: assert len(sys.argv)<3, "score_parser takes at most 1 command line argument\n"+\
giuliomoro@0:                         "Use: python score_parser.py [rating_folder_location]"
giuliomoro@0: 
giuliomoro@0: # XML results files location
giuliomoro@0: if len(sys.argv) == 1:
giuliomoro@0:     folder_name = "../saves"    # Looks in 'saves/' folder from 'scripts/' folder
giuliomoro@0:     print "Use: python score_parser.py [rating_folder_location]"
giuliomoro@0:     print "Using default path: " + folder_name
giuliomoro@0: elif len(sys.argv) == 2:
giuliomoro@0:     folder_name = sys.argv[1]   # First command line argument is folder
giuliomoro@0: 
giuliomoro@0: # check if folder_name exists
giuliomoro@0: if not os.path.exists(folder_name):
giuliomoro@0:     #the file is not there
giuliomoro@0:     print "Folder '"+folder_name+"' does not exist."
giuliomoro@0:     sys.exit() # terminate script execution
giuliomoro@0: elif not os.access(os.path.dirname(folder_name), os.W_OK):
giuliomoro@0:     #the file does exist but write privileges are not given
giuliomoro@0:     print "No write privileges in folder '"+folder_name+"'."
giuliomoro@0: 
giuliomoro@0:     
giuliomoro@0: # CODE
giuliomoro@0: 
giuliomoro@0: # remember which files have been opened this time
giuliomoro@0: file_history = []
giuliomoro@0: 
giuliomoro@0: # get every XML file in folder
giuliomoro@0: for file_name in os.listdir(folder_name):
giuliomoro@0:     if file_name.endswith(".xml"):
giuliomoro@0:         tree = ET.parse(folder_name + '/' + file_name)
giuliomoro@0:         root = tree.getroot()
giuliomoro@0: 
giuliomoro@0:         # get subject ID from XML file
giuliomoro@0:         subject_id = file_name[:-4] # file name (without extension) as subject ID
giuliomoro@0: 
giuliomoro@0:         # get list of all pages this subject evaluated
giuliomoro@0:         for audioholder in root.findall("./page"):    # iterate over pages
giuliomoro@0:             page_name = audioholder.get('id') # get page name
giuliomoro@0:                        
giuliomoro@0:             if page_name is None: # ignore 'empty' audio_holders
giuliomoro@0:                 print "WARNING: " + file_name + " contains empty audio holder. (score_parser.py)"
giuliomoro@0:                 break
giuliomoro@0: 
giuliomoro@0:             file_name = folder_name+'/ratings/'+page_name+'-ratings.csv' # score file name
giuliomoro@0: 
giuliomoro@0:             # create folder 'ratings' if not yet created
giuliomoro@0:             if not os.path.exists(folder_name + '/ratings'):
giuliomoro@0:                 os.makedirs(folder_name + '/ratings')
giuliomoro@0: 
giuliomoro@0:             # header: fragment IDs in 'alphabetical' order
giuliomoro@0:             # go to fragment column, or create new column if it doesn't exist yet
giuliomoro@0: 
giuliomoro@0:             # get array of audio elements and number of audio elements
giuliomoro@0:             audiolist = audioholder.findall("./audioelement")
giuliomoro@0:             n_fragments = len(audiolist)
giuliomoro@0: 
giuliomoro@0:             # get alphabetical array of fragment IDs from this subject's XML
giuliomoro@0:             fragmentnamelist = []    # make empty list
giuliomoro@0:             for audioelement in audiolist: # iterate over all audioelements
giuliomoro@0:                 fragmentnamelist.append(audioelement.get('id')) # add to list
giuliomoro@0: 
giuliomoro@0: 
giuliomoro@0:             # if file exists, get header and add any 'new' fragments not yet in the header
giuliomoro@0:             if os.path.isfile(file_name):
giuliomoro@0:                 with open(file_name, 'r') as readfile:
giuliomoro@0:                     filereader = csv.reader(readfile, delimiter=',')
giuliomoro@0:                     headerrow = filereader.next()
giuliomoro@0: 
giuliomoro@0:                 # If file hasn't been opened yet this time, remove all rows except header
giuliomoro@0:                 if file_name not in file_history:
giuliomoro@0:                     with open(file_name, 'w') as writefile:
giuliomoro@0:                         filewriter = csv.writer(writefile, delimiter=',')
giuliomoro@0:                         headerrow = sorted(headerrow)
giuliomoro@0:                         filewriter.writerow(headerrow)
giuliomoro@0:                     file_history.append(file_name)
giuliomoro@0: 
giuliomoro@0:                 # Which of the fragments are in fragmentnamelist but not in headerrow?
giuliomoro@0:                 newfragments = list(set(fragmentnamelist)-set(headerrow))
giuliomoro@0:                 newfragments = sorted(newfragments) # new fragments in alphabetical order
giuliomoro@0:                 # If not empty, read file and rewrite adding extra columns
giuliomoro@0:                 if newfragments: # if not empty
giuliomoro@0:                     with open('temp.csv', 'w') as writefile:
giuliomoro@0:                         filewriter = csv.writer(writefile, delimiter=',')
giuliomoro@0:                         filewriter.writerow(headerrow + newfragments) # write new header
giuliomoro@0:                         with open(file_name, 'r') as readfile:
giuliomoro@0:                             filereader = csv.reader(readfile, delimiter=',')
giuliomoro@0:                             filereader.next() # skip header
giuliomoro@0:                             for row in filereader: # rewrite row plus empty cells for every new fragment name
giuliomoro@0:                                 filewriter.writerow(row + ['']*len(newfragments))
giuliomoro@0:                     os.rename('temp.csv', file_name) # replace old file with temp file
giuliomoro@0:                     headerrow = headerrow + newfragments
giuliomoro@0:                     
giuliomoro@0: 
giuliomoro@0:             # if file does not exist yet, create file and make header
giuliomoro@0:             else:
giuliomoro@0:                 headerrow = sorted(fragmentnamelist) # sort alphabetically
giuliomoro@0:                 headerrow.insert(0,'')
giuliomoro@0:                 fragmentnamelist = fragmentnamelist[1:] #HACKY FIX inserting in firstrow also affects fragmentnamelist
giuliomoro@0:                 with open(file_name, 'w') as writefile:
giuliomoro@0:                     filewriter = csv.writer(writefile, delimiter=',')
giuliomoro@0:                     filewriter.writerow(headerrow)
giuliomoro@0:                 file_history.append(file_name)
giuliomoro@0: 
giuliomoro@0:             # open file to write for this page
giuliomoro@0:             writefile = open(file_name, 'a')
giuliomoro@0:             filewriter = csv.writer(writefile, delimiter=',')
giuliomoro@0: 
giuliomoro@0:             # prepare row to be written for this subject for this page
giuliomoro@0:             ratingrow = [subject_id]
giuliomoro@0: 
giuliomoro@0:             # get scores related to fragment [id]
giuliomoro@0:             for fragmentname in headerrow[1:]: # iterate over fragments in header (skip first empty column)
giuliomoro@0:                 elementvalue = audioholder.find("./audioelement/[@id='"
giuliomoro@0:                                        + fragmentname
giuliomoro@0:                                        + "']/value")
giuliomoro@0:                 if hasattr(elementvalue, 'text'): # if rating for this fragment exists
giuliomoro@0:                     ratingrow.append(elementvalue.text) # add to rating row
giuliomoro@0:                 else: # if this subject has not rated this fragment
giuliomoro@0:                     ratingrow.append('') # append empty cell
giuliomoro@0: 
giuliomoro@0:             # write row: [subject ID, rating fragment ID 1, ..., rating fragment ID M]
giuliomoro@0:             if any(ratingrow[1:]): # append to file if row non-empty (except subject name)
giuliomoro@0:                 filewriter.writerow(ratingrow)