BrechtDeMan@884: #!/usr/bin/python BrechtDeMan@884: nicholas@872: import xml.etree.ElementTree as ET nicholas@872: import os nicholas@872: import csv nicholas@872: nicholas@872: #TODO Remove DEBUG statements nicholas@872: BrechtDeMan@883: # XML results files location (modify as needed): BrechtDeMan@883: folder_name = "../saves" # Looks in 'saves/' folder from 'scripts/' folder BrechtDeMan@883: nicholas@872: # get every XML file in folder BrechtDeMan@885: for file in os.listdir(folder_name): nicholas@872: if file.endswith(".xml"): BrechtDeMan@883: tree = ET.parse(folder_name + '/' + file) nicholas@872: root = tree.getroot() BrechtDeMan@884: #print "DEBUG Reading " + file + "..." nicholas@872: nicholas@872: # get subject ID from XML file BrechtDeMan@884: subject_id = file[:-4] # file name (without extension) as subject ID nicholas@872: nicholas@872: # get list of all pages this subject evaluated nicholas@872: for audioholder in root.findall("./audioholder"): # iterate over pages nicholas@872: page_name = audioholder.get('id') # get page name BrechtDeMan@884: nicholas@872: if page_name is None: # ignore 'empty' audio_holders nicholas@872: break nicholas@872: BrechtDeMan@883: file_name = folder_name+'/ratings/'+page_name+'-ratings.csv' # score file name nicholas@872: nicholas@872: # create folder 'ratings if not yet created BrechtDeMan@883: if not os.path.exists(folder_name + '/ratings'): BrechtDeMan@883: os.makedirs(folder_name + '/ratings') nicholas@872: nicholas@872: # header: fragment IDs in 'alphabetical' order nicholas@872: # go to fragment column, or create new column if it doesn't exist yet nicholas@872: nicholas@872: # get array of audio elements and number of audio elements BrechtDeMan@883: audiolist = audioholder.findall("./audioelement") nicholas@872: n_fragments = len(audiolist) nicholas@872: nicholas@872: # get alphabetical array of fragment IDs from this subject's XML nicholas@872: fragmentnamelist = [] # make empty list nicholas@872: for audioelement in audiolist: # iterate over all audioelements nicholas@872: fragmentnamelist.append(audioelement.get('id')) # add to list nicholas@872: nicholas@872: nicholas@872: # if file exists, get header and add 'new' fragments nicholas@872: if os.path.isfile(file_name): BrechtDeMan@884: #print "DEBUG file " + file_name + " already exists - reading header" nicholas@872: with open(file_name, 'r') as readfile: nicholas@872: filereader = csv.reader(readfile, delimiter=',') nicholas@872: headerrow = filereader.next() nicholas@872: BrechtDeMan@884: # Which of the fragmentes are in fragmentnamelist but not in headerrow? BrechtDeMan@884: newfragments = list(set(fragmentnamelist)-set(headerrow)) BrechtDeMan@884: newfragments = sorted(newfragments) # new fragments in alphabetical order BrechtDeMan@884: # If not empty, read file and rewrite adding extra columns BrechtDeMan@884: if newfragments: # if not empty BrechtDeMan@884: #print "DEBUG New fragments found: " + str(newfragments) BrechtDeMan@884: with open('temp.csv', 'w') as writefile: BrechtDeMan@884: filewriter = csv.writer(writefile, delimiter=',') BrechtDeMan@884: filewriter.writerow(headerrow + newfragments) # write new header BrechtDeMan@884: #print " "+str(headerrow + newfragments) # DEBUG BrechtDeMan@884: with open(file_name, 'r') as readfile: BrechtDeMan@884: filereader = csv.reader(readfile, delimiter=',') BrechtDeMan@884: filereader.next() # skip header nicholas@872: for row in filereader: # rewrite row plus empty cells for every new fragment name BrechtDeMan@884: #print " Old row: " + str(row) # DEBUG nicholas@872: filewriter.writerow(row + ['']*len(newfragments)) BrechtDeMan@884: #print " New row: " + str(row + ['']*len(newfragments)) # DEBUG BrechtDeMan@884: os.rename('temp.csv', file_name) # replace old file with temp file BrechtDeMan@884: headerrow = headerrow + newfragments BrechtDeMan@884: #print "DEBUG New header row: " + str(headerrow) nicholas@872: nicholas@872: # if not, create file and make header nicholas@872: else: nicholas@872: #print ["DEBUG file " + file_name + " doesn't exist yet - making new one"] nicholas@872: headerrow = sorted(fragmentnamelist) # sort alphabetically nicholas@872: headerrow.insert(0,'') nicholas@872: fragmentnamelist = fragmentnamelist[1:] #HACKY FIX inserting in firstrow also affects fragmentnamelist nicholas@872: with open(file_name, 'w') as writefile: nicholas@872: filewriter = csv.writer(writefile, delimiter=',') nicholas@872: filewriter.writerow(headerrow) nicholas@872: nicholas@872: # open file to write for this page nicholas@872: writefile = open(file_name, 'a') nicholas@872: filewriter = csv.writer(writefile, delimiter=',') nicholas@872: nicholas@872: # prepare row to be written for this subject for this page nicholas@872: ratingrow = [subject_id] nicholas@872: nicholas@872: # get scores related to fragment [id] nicholas@872: for fragmentname in headerrow[1:]: # iterate over fragments in header (skip first empty column) BrechtDeMan@883: elementvalue = audioholder.find("./audioelement/[@id='" nicholas@872: + fragmentname nicholas@872: + "']/value") nicholas@872: if hasattr(elementvalue, 'text'): # if rating for this fragment exists nicholas@872: ratingrow.append(elementvalue.text) # add to rating row nicholas@872: else: # if this subject has not rated this fragment nicholas@872: ratingrow.append('') # append empty cell nicholas@872: nicholas@872: # write row: [subject ID, rating fragment ID 1, ..., rating fragment ID M] BrechtDeMan@885: if any(ratingrow[1:]): # append to file if row non-empty (except subject name) BrechtDeMan@885: filewriter.writerow(ratingrow) nicholas@872: