b@2264: #!/usr/bin/python
b@2264: 
b@2264: import xml.etree.ElementTree as ET
b@2264: import os
b@2264: import sys
b@2264: import csv
brecht@2988: import re
b@2264: 
b@2264: # COMMAND LINE ARGUMENTS
b@2264: 
b@2264: assert len(sys.argv)<3, "score_parser takes at most 1 command line argument\n"+\
b@2264:                         "Use: python score_parser.py [rating_folder_location]"
b@2264: 
b@2264: # XML results files location
b@2264: if len(sys.argv) == 1:
b@2264:     folder_name = "../saves"    # Looks in 'saves/' folder from 'scripts/' folder
b@2266:     print("Use: python score_parser.py [rating_folder_location]")
b@2266:     print("Using default path: " + folder_name)
b@2264: elif len(sys.argv) == 2:
b@2264:     folder_name = sys.argv[1]   # First command line argument is folder
b@2264: 
b@2264: # check if folder_name exists
b@2264: if not os.path.exists(folder_name):
b@2264:     #the file is not there
b@2266:     print("Folder '"+folder_name+"' does not exist.")
b@2264:     sys.exit() # terminate script execution
b@2264: elif not os.access(os.path.dirname(folder_name), os.W_OK):
b@2264:     #the file does exist but write privileges are not given
b@2266:     print("No write privileges in folder '"+folder_name+"'.")
nicholas@2524:     
b@2264: 
b@2264: # CODE
b@2264: 
nicholas@2524: storage = {}
b@2264: 
nicholas@2524: # create folder 'ratings' if not yet created
nicholas@2524: if not os.path.exists(folder_name + '/ratings'):
nicholas@2524:     os.makedirs(folder_name + '/ratings')
nicholas@2524: 
nicholas@2524: # Get every XML file in the folder
b@2264: for file_name in os.listdir(folder_name):
nicholas@2524:     if (file_name.endswith(".xml")):
b@2264:         tree = ET.parse(folder_name + '/' + file_name)
b@2264:         root = tree.getroot()
nicholas@2524:         
nicholas@2524:         subject_id = root.get('key');
nicholas@2524:         
nicholas@2524:         # get the list of the pages this subject evaluated
nicholas@2524:         for page in root.findall("./page"):     # iterate over pages
nicholas@2524:             page_name = page.get('ref') # get page ID
nicholas@2524:             
b@2264:             if page_name is None: # ignore 'empty' audio_holders
b@2266:                 print("WARNING: " + file_name + " contains empty audio holder. (score_parser.py)")
b@2264:                 break
b@2264:                 
b@2264:             if page.get('state') != "complete":
b@2266:                 print("WARNING: " + file_name + " contains incomplete page " +page_name+ ". (score_parser.py)")
b@2264:                 break;
nicholas@2524:             
nicholas@2524:             # Check if page in the store
nicholas@2524:             if storage.get(page_name) == None:
nicholas@2524:                 storage[page_name] = {'header':[], 'axis':{}} # add to the store
nicholas@2524:             
brecht@2988:             # strip repetitions
brecht@2988:             page_name_root = re.sub('-repeat-.$', '', page_name)
brecht@2988: 
nicholas@2524:             # Get the axis names
brecht@2988:             pageConfig = root.find('./waet/page/[@id="'+page_name_root+'"]')
me@2990:             if pageConfig is None:
me@2990:                 interfaceName = "default"
nicholas@2524:                 if storage[page_name]['axis'].get(interfaceName) == None:
nicholas@2524:                     storage[page_name]['axis'][interfaceName] = {}  # If not in store for page, add empty dict
me@2990:                 storage[page_name]['axis'][interfaceName][subject_id] = []
me@2990:             else:
me@2990:                 for interface in pageConfig.findall('./interface'):    # Get the <interface> nodes
me@2990:                     interfaceName = interface.get("name"); # Get the axis name
me@2990:                     if interfaceName == None:
me@2990:                         interfaceName = "default"   # If name not set, make name 'default'
me@2990:                     if storage[page_name]['axis'].get(interfaceName) == None:
me@2990:                         storage[page_name]['axis'][interfaceName] = {}  # If not in store for page, add empty dict
me@2990:                     storage[page_name]['axis'][interfaceName][subject_id] = [] # Add the store for the session
nicholas@2524:                     
b@2264:             # header: fragment IDs in 'alphabetical' order
b@2264:             # go to fragment column, or create new column if it doesn't exist yet
nicholas@2524:             
b@2264:             # get alphabetical array of fragment IDs from this subject's XML
b@2264:             fragmentnamelist = []    # make empty list
nicholas@2524:             for audioelement in page.findall("./audioelement"): # iterate over all audioelements
epdamskagg@2947:                 if audioelement is not None and audioelement.get('type') != "outside-reference":
epdamskagg@2947:                     fragmentnamelist.append(audioelement.get('ref')) # add to list
nicholas@2524:             
nicholas@2524:             fragmentnamelist = sorted(fragmentnamelist);    # Sort the list
nicholas@2524:             storage[page_name]['header'] = fragmentnamelist;
nicholas@2524:             
nicholas@2524:             for fragmentname in fragmentnamelist:
nicholas@2524:                 audioElement = page.find("./audioelement/[@ref='"+ fragmentname+ "']") # Get the element
nicholas@2524:                 for value in audioElement.findall('./value'):
nicholas@2524:                     axisName = value.get('interface-name')
nicholas@2831:                     if axisName == None or axisName == "null":
nicholas@2524:                         axisName = 'default'
nicholas@2524:                     axisStore = storage[page_name]['axis'][axisName]
nicholas@2524:                     if hasattr(value, 'text'):
nicholas@2524:                         axisStore[subject_id].append(value.text)
nicholas@2524:                     else:
nicholas@2524:                         axisStore[subject_id].append('')
b@2264: 
nicholas@2524: # Now create the individual files
nicholas@2524: for page_name in storage:
nicholas@2524:     for axis_name in storage[page_name]['axis']:
nicholas@2524:         
nicholas@2524:         file_name = folder_name+'/ratings/'+page_name+'-'+axis_name+'-ratings.csv' # score file name
nicholas@2524:         
nicholas@2524:         # I'm not as elegant, I say burn the files and start again
nicholas@2524:         headerrow = list(storage[page_name]['header'])  # Extract the element IDs
nicholas@2524:         headerrow.insert(0,'file_keys')
nicholas@2524:         with open(file_name, 'w') as writefile:
b@2264:             filewriter = csv.writer(writefile, delimiter=',')
nicholas@2524:             filewriter.writerow(headerrow)
nicholas@2524:         
nicholas@2524:         # open file to write the page
nicholas@2524:         writefile = open(file_name, 'a')
nicholas@2524:         filewriter = csv.writer(writefile, delimiter=',')
nicholas@2524:         
nicholas@2524:         for subject_id in storage[page_name]['axis'][axis_name]:
nicholas@2524:             entry = [subject_id]
nicholas@2524:             for value in storage[page_name]['axis'][axis_name][subject_id]:
nicholas@2524:                 entry.append(value)
nicholas@2524:             filewriter.writerow(entry)
epdamskagg@2947:         writefile.close()