nickjillings@1542: import xml.etree.ElementTree as ET
nickjillings@1542: import os
nickjillings@1542: import csv
nickjillings@1542: 
nickjillings@1542: #TODO Remove DEBUG statements
nickjillings@1542: 
nickjillings@1542: # get every XML file in folder
nickjillings@1542: for file in os.listdir("."): # You have to put this in folder where output XML files are.
nickjillings@1542:     if file.endswith(".xml"):
nickjillings@1542:         tree = ET.parse(file)
nickjillings@1542:         root = tree.getroot()
nickjillings@1542:         #print ["DEBUG Reading " + file + "..."]
nickjillings@1542: 
nickjillings@1542:         # get subject ID from XML file
nickjillings@1542:         subject_id = file # file name as subject ID
nickjillings@1542: 
nickjillings@1542:         # get list of all pages this subject evaluated
nickjillings@1542:         for audioholder in root.findall("./audioholder"):    # iterate over pages
nickjillings@1542:             page_name = audioholder.get('id') # get page name
nickjillings@1542:             #print ["DEBUG    page " + page_name]
nickjillings@1542:             
nickjillings@1542:             if page_name is None: # ignore 'empty' audio_holders
nickjillings@1542:                 break
nickjillings@1542: 
nickjillings@1542:             file_name = 'ratings/'+page_name+'-ratings.csv' # score file name
nickjillings@1542: 
nickjillings@1542:             # create folder 'ratings if not yet created
nickjillings@1542:             if not os.path.exists('ratings'):
nickjillings@1542:                 os.makedirs('ratings')
nickjillings@1542: 
nickjillings@1542:             # header: fragment IDs in 'alphabetical' order
nickjillings@1542:             # go to fragment column, or create new column if it doesn't exist yet
nickjillings@1542: 
nickjillings@1542:             # get array of audio elements and number of audio elements
nickjillings@1542:             audiolist = root.findall("*/[@id='"+page_name+"']/audioelement")
nickjillings@1542:             n_fragments = len(audiolist)
nickjillings@1542: 
nickjillings@1542:             # get alphabetical array of fragment IDs from this subject's XML
nickjillings@1542:             fragmentnamelist = []    # make empty list
nickjillings@1542:             for audioelement in audiolist: # iterate over all audioelements
nickjillings@1542:                 fragmentnamelist.append(audioelement.get('id')) # add to list
nickjillings@1542: 
nickjillings@1542: 
nickjillings@1542:             # if file exists, get header and add 'new' fragments
nickjillings@1542:             if os.path.isfile(file_name):
nickjillings@1542:                 #print ["DEBUG file " + file_name + " already exists - reading header"]
nickjillings@1542:                 with open(file_name, 'r') as readfile:
nickjillings@1542:                     filereader = csv.reader(readfile, delimiter=',')
nickjillings@1542:                     headerrow = filereader.next()
nickjillings@1542:                     #headerrow = headerrow[1:] # remove first column (empty)
nickjillings@1542: 
nickjillings@1542:                     # Which of the fragmentes are in fragmentnamelist but not in headerrow?
nickjillings@1542:                     newfragments = list(set(fragmentnamelist)-set(headerrow))
nickjillings@1542:                     newfragments = sorted(newfragments) # new fragments in alphabetical order
nickjillings@1542:                     # If not empty, read file and rewrite adding extra columns
nickjillings@1542:                     if newfragments: # if not empty
nickjillings@1542:                         print ["DEBUG New fragments found: " + str(newfragments)]
nickjillings@1542:                         with open('temp.csv', 'w') as writefile:
nickjillings@1542:                             filewriter = csv.writer(writefile, delimiter=',')
nickjillings@1542:                             filewriter.writerow(headerrow + newfragments) # write new header
nickjillings@1542:                             for row in filereader: # rewrite row plus empty cells for every new fragment name
nickjillings@1542:                                 #print ["DEBUG Old row: " + str(row)]
nickjillings@1542:                                 filewriter.writerow(row + ['']*len(newfragments))
nickjillings@1542:                                 #print ["DEBUG New row: " + str(row + ['']*len(newfragments))]
nickjillings@1542:                         os.rename('temp.csv', file_name) # replace old file with temp file
nickjillings@1542:                         headerrow = headerrow + newfragments
nickjillings@1542:                         print ["DEBUG New header row: " + str(headerrow)]
nickjillings@1542: 
nickjillings@1542:             # if not, create file and make header
nickjillings@1542:             else:
nickjillings@1542:                 #print ["DEBUG file " + file_name + " doesn't exist yet - making new one"]
nickjillings@1542:                 headerrow = sorted(fragmentnamelist) # sort alphabetically
nickjillings@1542:                 headerrow.insert(0,'')
nickjillings@1542:                 fragmentnamelist = fragmentnamelist[1:] #HACKY FIX inserting in firstrow also affects fragmentnamelist
nickjillings@1542:                 with open(file_name, 'w') as writefile:
nickjillings@1542:                     filewriter = csv.writer(writefile, delimiter=',')
nickjillings@1542:                     filewriter.writerow(headerrow)
nickjillings@1542: 
nickjillings@1542:             # open file to write for this page
nickjillings@1542:             writefile = open(file_name, 'a')
nickjillings@1542:             filewriter = csv.writer(writefile, delimiter=',')
nickjillings@1542: 
nickjillings@1542:             # prepare row to be written for this subject for this page
nickjillings@1542:             ratingrow = [subject_id]
nickjillings@1542: 
nickjillings@1542:             # get scores related to fragment [id]
nickjillings@1542:             for fragmentname in headerrow[1:]: # iterate over fragments in header (skip first empty column)
nickjillings@1542:                 elementvalue = root.find("*/[@id='"
nickjillings@1542:                                        + page_name
nickjillings@1542:                                        + "']/audioelement/[@id='"
nickjillings@1542:                                        + fragmentname
nickjillings@1542:                                        + "']/value")
nickjillings@1542:                 if hasattr(elementvalue, 'text'): # if rating for this fragment exists
nickjillings@1542:                     ratingrow.append(elementvalue.text) # add to rating row
nickjillings@1542:                 else: # if this subject has not rated this fragment
nickjillings@1542:                     ratingrow.append('') # append empty cell
nickjillings@1542: 
nickjillings@1542:             # write row: [subject ID, rating fragment ID 1, ..., rating fragment ID M]
nickjillings@1542:             filewriter.writerow(ratingrow)
nickjillings@1542: