def readDocument(path_sentences, path_trees, path_output, max_length):
''' Performs CKY and Viterbi '''
file_sentences = open(path_sentences, 'r+')
file_trees = open(path_trees, 'r+')
timenow = str(time.time())
correct_trees = open("GOLD_OUTPUT_" + timenow + ".txt", 'w+')
test_trees = open(path_output, 'w+')
SentenceParser = Parser()
parse_information = SentenceParser.load_database( 'parse_information_unknown.p' )
Algorithm = CKY( parse_information )
i = 0
while True:
sentence = file_sentences.readline()
correct_tree = file_trees.readline()
words = sentence.split(' ')
size = len(words) - 2
if size > max_length:
print "-- Skip sentence too long --"
i += 1
continue
print "CKY starts parsing sentence:", i, "of size:", size
i += 1
if sentence == '':
break
# Run CKY and Viterbi
tree = Algorithm.run( words[:size] )
if tree:
test_trees.write(tree)
correct_trees.write(correct_tree)
correct_trees.close()
test_trees.close()
