evala.py

"""
+M
+P
+C
+A

nur t%cod bis zum naechsten *
sd3: 20, pro user *JAN, *MUT, *LEN

danach + oder |

sd3 ist eine utterance mit 3 verschiedenen semcons. wieviele sd3s, sd4s, sd2s, etc gibt es?

-----------------------------

pro sem. inhaltszeichen: in wievielen utterances kommt es min. 1x vor

----------

pro ZEILE = von $ bis }

+V-X: im V-feld min ein +, aber in den E-felder darf kein + sein. 
+V+X: wenn in V min ein + und in min einem der E-felder min. 1 + ist.
-V+X: im V-feld kein +, aber in min. 1 E-feld min. ein + sein

"""

import sys
import csv

def utteranceDensity(line):
   """
      how many sem cons are there in this line
   """
   occurances = 0;
   for semConVar in ['P', 'C', 'M', 'A']:
      if informationFocus(line, semConVar):
         occurances += 1
   return occurances

def informationFocus(line, semConVar):
   """
      is there a semcon in this line?
   """
   return line.find('+' + semConVar + '+') > -1 or line.find('+' + semConVar + '|') > 1

def informationLocus(line): 
   """
      return weither there is a V and/or one of the semcons in this line
   """
   #parts = line.split('\t')
   fields = line.split(':')
   hasV = (fields[2][0] == 'V') and fields[2].find('+') > 1
   hasX = False
   eFields = ':'.join(fields[5:9])
   for semConVar in ['P', 'C', 'M', 'A']:
      if informationFocus(eFields, semConVar):
         hasX = True
   return (hasV, hasX)


def anal_utterancesWithSemCon():
   semCons = {
      'P': 0,
      'C': 0,
      'L': 0,
      'M': 0,
      'A': 0
   }
   for [user, utter] in utterances():      
      for semConVar in semCons.keys():
         if informationFocus(utter, semConVar):
            semCons[semConVar] += 1
   return semCons

def anal_semConDensity():
   densities = {
      0: 0,
      1: 0,
      2: 0,
      3: 0,
      4: 0
   }
   for [user, utter] in utterances():
      count = utteranceDensity(utter)
      densities[count] += 1
   return densities

def anal_informationLocus():
   locs = {
      (True, False): 0,
      (True, True): 0,
      (False, True): 0,
      (False, False): 0
   
   }
   for [user, utterLine] in lines():
      if not utterLine: break
      
      locData = informationLocus(utterLine)
      locs[locData] += 1
   return locs


def lines():
   inCod = False   
   for line in file.readlines():
      print line
      if line[0] == '*' or line[0:4].lower() == '@end':
         if inCod:
            inCod = False
         user = line[1:4]
      elif line[0:5].lower() == '%cod:':
         inCod = True
         parts = line.split('\t')
         if len(parts) > 1 and len(parts[1]) > 2:
            yield [user, parts[1]]
      elif line[0:1] == '%':
         inCod = False
      elif inCod:
         yield [user, line]
#      elif line[0] == '%':
      else:
         pass
   yield [None, None];

def utterances():
   utterance = ''
   user = None
   inCod = False
   for line in file.readlines():
      if line[0] == '*' or line[0:4].lower() == '@end':
         if inCod:
            inCod = False
            yield [user, utterance]
         user = line[1:4]
      elif line[0:5].lower() == '%cod:':
         inCod = True
         parts = line.split('\t')
         if len(parts) > 1:
            utterance = parts[1]
      elif line[0:1] == '%':
         inCod = False
      elif inCod:
         utterance += line
#      elif line[0] == '%':
      else:
#         print 'pass '+ line[0:5]
         pass
   yield [user, utterance]


paths = sys.argv[1:]
for COD_PATH in paths:
   file = open(COD_PATH)
   print '--------------'
   print 'DATEI: ' + file.name
   print 'How many utterances have a certain sem con'
   semAndCount = anal_utterancesWithSemCon()
   for semCon in semAndCount.keys():
         print semCon + ', ' + str(semAndCount[semCon])
   file.close()
   file = open(COD_PATH)
   print 'Information Locus'
   print 'V, X, Anzahl'
   infoLocus = anal_informationLocus()
   for table in infoLocus.keys():
      plusV = '+V' if table[0] == True else '-V'
      plusX = '+X' if table[1] == True else  '-X'
      print plusV + '' + plusX + ', ' + str(infoLocus[table])
   file.close()
   file = open(COD_PATH)
   print 'Number of Utterances per sem con Density'
   print 'number of semCons, number of utterances'
   density = anal_semConDensity()
   for numberOf in density.keys():
      print str(numberOf) + ', ' + str(density[numberOf])
   file.close()
   print '--------------'