import nltk
import sys
from parser import CKYParser
if __name__ == "__main__":
if (len(sys.argv) == 2):
sent = sys.argv[0]
cnf = sys.argv[1]
else:
sent = "sentences.txt"
cnf = "grammar.cnf"
grammar = nltk.data.load("file:grammar.cnf")
sentences = open(sent, 'r')
parser = CKYParser(grammar)
# initialize the counting
sentCount = 0;
parsCount = 0;
totalParsCount = 0;
for sentence in sentences: # reads each sentence
tokens = nltk.tokenize.wordpunct_tokenize(sentence) # makes each sentence splitted
sentCount += 1 # counts the number of sentences in this example
parses = parser.nbest_parse(tokens) # parses the given sentence to get parse trees that represent possible structures for the given sentence
for parse in parses: # prints the parse trees
print str(parse) + "\n"
parsCount += 1 # counts the number of parses for the sentence
sentence_path = sys.argv[1]
except IndexError:
exit("Please give a path to a file of sentences.")
try:
grammar_path = sys.argv[2]
except IndexError:
exit("Please give a path to a file with a grammar.")
with open(sentence_path) as sentence_file:
sentence_data = sentence_file.read()
sentences = sentence_data.strip().split("\n")
sentences = [nltk.wordpunct_tokenize(sentence) for sentence in sentences]
parser = CKYParser(grammar_path)
delimiter = 40 * "-"
metadata = ["Kathryn Nichols", "Stefan Behr", "LING 571", "Project 1"]
print "\n".join("# {0}".format(datum) for datum in metadata)
print delimiter
# iterate over tokenized sentences, parse each
# and output parses
for sentence in sentences:
sentence = ["'{0}'".format(token) for token in sentence
] # make sure tokens are single-quoted
parses = parser.get_parses(sentence)
sys.stdout.write(print_parses(parses))
print len(parses)
sentence_path = sys.argv[1]
except IndexError:
exit("Please give a path to a file of sentences.")
try:
grammar_path = sys.argv[2]
except IndexError:
exit("Please give a path to a file with a grammar.")
with open(sentence_path) as sentence_file:
sentence_data = sentence_file.read()
sentences = sentence_data.strip().split("\n")
sentences = [nltk.wordpunct_tokenize(sentence) for sentence in sentences]
parser = CKYParser(grammar_path)
delimiter = 40 * "-"
metadata = ["Kathryn Nichols", "Stefan Behr", "LING 571", "Project 1"]
print "\n".join("# {0}".format(datum) for datum in metadata)
print delimiter
# iterate over tokenized sentences, parse each
# and output parses
for sentence in sentences:
sentence = ["'{0}'".format(token) for token in sentence] # make sure tokens are single-quoted
parses = parser.get_parses(sentence)
sys.stdout.write(print_parses(parses))
print len(parses)
print delimiter