BrechtDeMan@928: import xml.etree.ElementTree as ET BrechtDeMan@928: import os BrechtDeMan@928: import csv BrechtDeMan@928: BrechtDeMan@928: #TODO Remove DEBUG statements BrechtDeMan@928: BrechtDeMan@928: # get every XML file in folder BrechtDeMan@928: for file in os.listdir("."): # You have to put this in folder where output XML files are. BrechtDeMan@928: if file.endswith(".xml"): BrechtDeMan@928: tree = ET.parse(file) BrechtDeMan@928: root = tree.getroot() BrechtDeMan@928: #print ["DEBUG Reading " + file + "..."] BrechtDeMan@928: BrechtDeMan@928: # get subject ID from XML file BrechtDeMan@928: subject_id = file # file name as subject ID BrechtDeMan@928: BrechtDeMan@928: # get list of all pages this subject evaluated BrechtDeMan@928: for audioholder in root.findall("./audioholder"): # iterate over pages BrechtDeMan@928: page_name = audioholder.get('id') # get page name BrechtDeMan@928: #print ["DEBUG page " + page_name] BrechtDeMan@1030: BrechtDeMan@1030: if page_name is None: # ignore 'empty' audio_holders BrechtDeMan@1030: break BrechtDeMan@928: BrechtDeMan@928: file_name = 'ratings/'+page_name+'-ratings.csv' # score file name BrechtDeMan@928: BrechtDeMan@928: # create folder 'ratings if not yet created BrechtDeMan@928: if not os.path.exists('ratings'): BrechtDeMan@928: os.makedirs('ratings') BrechtDeMan@928: BrechtDeMan@928: # header: fragment IDs in 'alphabetical' order BrechtDeMan@928: # go to fragment column, or create new column if it doesn't exist yet BrechtDeMan@928: BrechtDeMan@928: # get array of audio elements and number of audio elements BrechtDeMan@928: audiolist = root.findall("*/[@id='"+page_name+"']/audioelement") BrechtDeMan@928: n_fragments = len(audiolist) BrechtDeMan@928: BrechtDeMan@928: # get alphabetical array of fragment IDs from this subject's XML BrechtDeMan@928: fragmentnamelist = [] # make empty list BrechtDeMan@928: for audioelement in audiolist: # iterate over all audioelements BrechtDeMan@928: fragmentnamelist.append(audioelement.get('id')) # add to list BrechtDeMan@928: BrechtDeMan@928: BrechtDeMan@928: # if file exists, get header and add 'new' fragments BrechtDeMan@928: if os.path.isfile(file_name): BrechtDeMan@928: #print ["DEBUG file " + file_name + " already exists - reading header"] BrechtDeMan@928: with open(file_name, 'r') as readfile: BrechtDeMan@928: filereader = csv.reader(readfile, delimiter=',') BrechtDeMan@928: headerrow = filereader.next() BrechtDeMan@928: #headerrow = headerrow[1:] # remove first column (empty) BrechtDeMan@928: BrechtDeMan@928: # Which of the fragmentes are in fragmentnamelist but not in headerrow? BrechtDeMan@928: newfragments = list(set(fragmentnamelist)-set(headerrow)) BrechtDeMan@928: newfragments = sorted(newfragments) # new fragments in alphabetical order BrechtDeMan@928: # If not empty, read file and rewrite adding extra columns BrechtDeMan@928: if newfragments: # if not empty BrechtDeMan@928: print ["DEBUG New fragments found: " + str(newfragments)] BrechtDeMan@928: with open('temp.csv', 'w') as writefile: BrechtDeMan@928: filewriter = csv.writer(writefile, delimiter=',') BrechtDeMan@928: filewriter.writerow(headerrow + newfragments) # write new header BrechtDeMan@928: for row in filereader: # rewrite row plus empty cells for every new fragment name BrechtDeMan@928: #print ["DEBUG Old row: " + str(row)] BrechtDeMan@928: filewriter.writerow(row + ['']*len(newfragments)) BrechtDeMan@928: #print ["DEBUG New row: " + str(row + ['']*len(newfragments))] BrechtDeMan@928: os.rename('temp.csv', file_name) # replace old file with temp file BrechtDeMan@928: headerrow = headerrow + newfragments BrechtDeMan@928: print ["DEBUG New header row: " + str(headerrow)] BrechtDeMan@928: BrechtDeMan@928: # if not, create file and make header BrechtDeMan@928: else: BrechtDeMan@928: #print ["DEBUG file " + file_name + " doesn't exist yet - making new one"] BrechtDeMan@928: headerrow = sorted(fragmentnamelist) # sort alphabetically BrechtDeMan@928: headerrow.insert(0,'') BrechtDeMan@928: fragmentnamelist = fragmentnamelist[1:] #HACKY FIX inserting in firstrow also affects fragmentnamelist BrechtDeMan@928: with open(file_name, 'w') as writefile: BrechtDeMan@928: filewriter = csv.writer(writefile, delimiter=',') BrechtDeMan@928: filewriter.writerow(headerrow) BrechtDeMan@928: BrechtDeMan@928: # open file to write for this page BrechtDeMan@928: writefile = open(file_name, 'a') BrechtDeMan@928: filewriter = csv.writer(writefile, delimiter=',') BrechtDeMan@928: BrechtDeMan@928: # prepare row to be written for this subject for this page BrechtDeMan@928: ratingrow = [subject_id] BrechtDeMan@928: BrechtDeMan@928: # get scores related to fragment [id] BrechtDeMan@928: for fragmentname in headerrow[1:]: # iterate over fragments in header (skip first empty column) BrechtDeMan@928: elementvalue = root.find("*/[@id='" BrechtDeMan@928: + page_name BrechtDeMan@928: + "']/audioelement/[@id='" BrechtDeMan@928: + fragmentname BrechtDeMan@928: + "']/value") BrechtDeMan@928: if hasattr(elementvalue, 'text'): # if rating for this fragment exists BrechtDeMan@928: ratingrow.append(elementvalue.text) # add to rating row BrechtDeMan@928: else: # if this subject has not rated this fragment BrechtDeMan@928: ratingrow.append('') # append empty cell BrechtDeMan@928: BrechtDeMan@928: # write row: [subject ID, rating fragment ID 1, ..., rating fragment ID M] nicholas@931: filewriter.writerow(ratingrow) nicholas@931: