en/AdvancedNlpCourse2015/InformationExtraction: demo.py

#!/usr/bin/python

import sys, re

clauses = []
clause = {}

for ll in sys.stdin:
    line = ll.strip()
    if line.startswith('<clause>'):
        if clause: clauses.append(clause)
        clause = {'text': line[8:], 'vp': [], 'phr': []}
    elif line.startswith('<vp>'):
        tag, text = line[6:].split('): ', 1)
        clause['vp'].append({'text': text, 'tag': tag})
    elif line.startswith('<vplemma>'):
        clause['vp'][-1]['lemma'] = line.split('): ', 1)[1]
    elif line.startswith('<phr>'):
        tag, text = line[7:].split('): ', 1)
        clause['phr'].append({'text': text, 'tag': tag})
    elif line.startswith('<phrlemma>'):
        clause['phr'][-1]['lemma'] = line.split('): ', 1)[1]

for clause in clauses: # just the example print of the data structure
    print clause['text']
    for vp in clause['vp']:
        print '  ', 'vp:', vp['text'], '[%s]' % vp['tag'], '[%s]' % vp['lemma']
    for p in clause['phr']:
        print '  ', 'phr:', p['text'], '[%s]' % p['tag'], '[%s]' % p['lemma']
    print
print
print

for clause in clauses: # "intelligence"
    if len(clause['vp']) == 0: continue
    if len(clause['phr']) < 2: continue
    phr1 = clause['phr'][0]
    for phr2 in clause['phr'][1:]:
        print phr1['text'], ':: was ::', phr2['text']
   
print