nicholas@2586: #!/usr/bin/python
nicholas@2586: 
nicholas@2586: import xml.etree.ElementTree as ET
nicholas@2586: import os
nicholas@2586: import sys
nicholas@2586: import csv
nicholas@2586: 
nicholas@2586: # COMMAND LINE ARGUMENTS
nicholas@2586: 
nicholas@2586: assert len(sys.argv)<3, "score_parser takes at most 1 command line argument\n"+\
nicholas@2586:                         "Use: python score_parser.py [rating_folder_location]"
nicholas@2586: 
nicholas@2586: # XML results files location
nicholas@2586: if len(sys.argv) == 1:
nicholas@2586:     folder_name = "../saves"    # Looks in 'saves/' folder from 'scripts/' folder
nicholas@2586:     print("Use: python score_parser.py [rating_folder_location]")
nicholas@2586:     print("Using default path: " + folder_name)
nicholas@2586: elif len(sys.argv) == 2:
nicholas@2586:     folder_name = sys.argv[1]   # First command line argument is folder
nicholas@2586: 
nicholas@2586: # check if folder_name exists
nicholas@2586: if not os.path.exists(folder_name):
nicholas@2586:     #the file is not there
nicholas@2586:     print("Folder '"+folder_name+"' does not exist.")
nicholas@2586:     sys.exit() # terminate script execution
nicholas@2586: elif not os.access(os.path.dirname(folder_name), os.W_OK):
nicholas@2586:     #the file does exist but write privileges are not given
nicholas@2586:     print("No write privileges in folder '"+folder_name+"'.")
nicholas@2586:     
nicholas@2586: 
nicholas@2586: # CODE
nicholas@2586: 
nicholas@2586: storage = {}
nicholas@2586: 
nicholas@2586: # create folder 'ratings' if not yet created
nicholas@2586: if not os.path.exists(folder_name + '/ratings'):
nicholas@2586:     os.makedirs(folder_name + '/ratings')
nicholas@2586: 
nicholas@2586: # Get every XML file in the folder
nicholas@2586: for file_name in os.listdir(folder_name):
nicholas@2586:     if (file_name.endswith(".xml")):
nicholas@2586:         tree = ET.parse(folder_name + '/' + file_name)
nicholas@2586:         root = tree.getroot()
nicholas@2586:         
nicholas@2586:         subject_id = root.get('key');
nicholas@2586:         
nicholas@2586:         # get the list of the pages this subject evaluated
nicholas@2586:         for page in root.findall("./page"):     # iterate over pages
nicholas@2586:             page_name = page.get('ref') # get page ID
nicholas@2586:             
nicholas@2586:             if page_name is None: # ignore 'empty' audio_holders
nicholas@2586:                 print("WARNING: " + file_name + " contains empty audio holder. (score_parser.py)")
nicholas@2586:                 break
nicholas@2586:                 
nicholas@2586:             if page.get('state') != "complete":
nicholas@2586:                 print("WARNING: " + file_name + " contains incomplete page " +page_name+ ". (score_parser.py)")
nicholas@2586:                 break;
nicholas@2586:             
nicholas@2586:             # Check if page in the store
nicholas@2586:             if storage.get(page_name) == None:
nicholas@2586:                 storage[page_name] = {'header':[], 'axis':{}} # add to the store
nicholas@2586:             
nicholas@2586:             # Get the axis names
nicholas@2586:             pageConfig = root.find('./waet/page/[@id="'+page_name+'"]')
nicholas@2586:             for interface in pageConfig.findall('./interface'):    # Get the <interface> noeds
nicholas@2586:                 interfaceName = interface.get("name"); # Get the axis name
nicholas@2586:                 if interfaceName == None:
nicholas@2586:                     interfaceName = "default"   # If name not set, make name 'default'
nicholas@2586:                 if storage[page_name]['axis'].get(interfaceName) == None:
nicholas@2586:                     storage[page_name]['axis'][interfaceName] = {}  # If not in store for page, add empty dict
nicholas@2586:                 storage[page_name]['axis'][interfaceName][subject_id] = [] # Add the store for the session
nicholas@2586:                     
nicholas@2586:             # header: fragment IDs in 'alphabetical' order
nicholas@2586:             # go to fragment column, or create new column if it doesn't exist yet
nicholas@2586:             
nicholas@2586:             # get alphabetical array of fragment IDs from this subject's XML
nicholas@2586:             fragmentnamelist = []    # make empty list
nicholas@2586:             for audioelement in page.findall("./audioelement"): # iterate over all audioelements
nicholas@2586:                 fragmentnamelist.append(audioelement.get('ref')) # add to list
nicholas@2586:             
nicholas@2586:             fragmentnamelist = sorted(fragmentnamelist);    # Sort the list
nicholas@2586:             storage[page_name]['header'] = fragmentnamelist;
nicholas@2586:             
nicholas@2586:             for fragmentname in fragmentnamelist:
nicholas@2586:                 audioElement = page.find("./audioelement/[@ref='"+ fragmentname+ "']") # Get the element
nicholas@2586:                 for value in audioElement.findall('./value'):
nicholas@2586:                     axisName = value.get('interface-name')
nicholas@2586:                     if axisName == None:
nicholas@2586:                         axisName = 'default'
nicholas@2586:                     axisStore = storage[page_name]['axis'][axisName]
nicholas@2586:                     if hasattr(value, 'text'):
nicholas@2586:                         axisStore[subject_id].append(value.text)
nicholas@2586:                     else:
nicholas@2586:                         axisStore[subject_id].append('')
nicholas@2586: 
nicholas@2586: # Now create the individual files
nicholas@2586: for page_name in storage:
nicholas@2586:     for axis_name in storage[page_name]['axis']:
nicholas@2586:         
nicholas@2586:         file_name = folder_name+'/ratings/'+page_name+'-'+axis_name+'-ratings.csv' # score file name
nicholas@2586:         
nicholas@2586:         # I'm not as elegant, I say burn the files and start again
nicholas@2586:         headerrow = list(storage[page_name]['header'])  # Extract the element IDs
nicholas@2586:         headerrow.insert(0,'file_keys')
nicholas@2586:         with open(file_name, 'w') as writefile:
nicholas@2586:             filewriter = csv.writer(writefile, delimiter=',')
nicholas@2586:             filewriter.writerow(headerrow)
nicholas@2586:         
nicholas@2586:         # open file to write the page
nicholas@2586:         writefile = open(file_name, 'a')
nicholas@2586:         filewriter = csv.writer(writefile, delimiter=',')
nicholas@2586:         
nicholas@2586:         for subject_id in storage[page_name]['axis'][axis_name]:
nicholas@2586:             entry = [subject_id]
nicholas@2586:             for value in storage[page_name]['axis'][axis_name][subject_id]:
nicholas@2586:                 entry.append(value)
nicholas@2586:             filewriter.writerow(entry)
nicholas@2586:         writefile.close()