#!/usr/bin/env python3
"""
Note: This only runs with Python3!
Demo: Find unlinked or unknown words.
These demo is extremely simplified.
It can only work with link-grammar library version >= 5.3.10.
Input: English sentences, one per line.
Output: If there are any []-marked words in the linkage results,
the output contains unique combinations of the input sentence with
these works marked. No attempt is done to handle the walls.
Spell guesses are not handled in this demo.
Example:
This is a the test of bfgiuing and xxxvfrg
Output:
Sentence has 1 unlinked word:
1: LEFT-WALL this.p is.v [a] the test.n of bfgiuing[!].g and.j-n xxxvfrg[?].n RIGHT-WALL
2: LEFT-WALL this.p is.v a [the] test.n of bfgiuing[!].g and.j-n xxxvfrg[?].n RIGHT-WALL
3: LEFT-WALL this.p is.v [a] the test.n of bfgiuing[!].g and.j-n xxxvfrg[?].a RIGHT-WALL
4: LEFT-WALL this.p is.v a [the] test.n of bfgiuing[!].g and.j-n xxxvfrg[?].a RIGHT-WALL
"""
import sys
import re
import argparse
from linkgrammar import (Sentence, ParseOptions, Dictionary,
LG_Error, LG_TimerExhausted, Clinkgrammar as clg)
def nsuffix(q):
return '' if q == 1 else 's'
class Formatter(argparse.HelpFormatter):
""" Display the "lang" argument as a first one, as in link-parser. """
def _format_usage(self, usage, actions, groups, prefix):
usage_message = super(Formatter, self)._format_usage(usage, actions, groups, prefix)
return re.sub(r'(usage: \S+) (.*) \[lang]', r'\1 [lang] \2', str(usage_message))
#-----------------------------------------------------------------------------#
is_stdin_atty = sys.stdin.isatty()
PROMPT = "sentence-check: " if is_stdin_atty else ""
DISPLAY_GUESSES = True # Display regex and POS guesses
BATCH_LABELS = '*: '
print("Version:", clg.linkgrammar_get_version())
args = argparse.ArgumentParser(formatter_class=Formatter)
args.add_argument('lang', nargs='?', default='en',
help="language or dictionary location")
args.add_argument("-v", "--verbosity", type=int,default=0,
choices=range(0,199), metavar='[0-199]',
help= "1: Basic verbosity; 2-4: Trace; >5: Debug")
args.add_argument("-p", "--position", action="store_true",
help="show word sentence position")
args.add_argument("-nm", "--no-morphology", dest='morphology', action='store_false',
help="do not display morphology")
args.add_argument("-i", "--interactive", action="store_true",
help="interactive mode after each result")
arg = args.parse_args()
try:
lgdict = Dictionary(arg.lang)
except LG_Error:
# The default error handler will print the error message
args.print_usage()
sys.exit(2)
po = ParseOptions(verbosity=arg.verbosity)
po.max_null_count = 999 # > allowed maximum number of words
po.linkage_limit = 10000 # maximum number of linkages to generate
po.max_parse_time = 10 # actual parse timeout may be about twice bigger
po.spell_guess = True if DISPLAY_GUESSES else False
po.display_morphology = arg.morphology
while True:
try:
sentence_text = input(PROMPT)
except EOFError:
print("EOF")
exit(0)
if not is_stdin_atty and sentence_text:
if sentence_text[0] == '%':
continue
if sentence_text[0] == '!': # ignore user-settings for now
continue
if sentence_text[0] in BATCH_LABELS:
sentence_text = sentence_text[1:]
if sentence_text.strip() == '':
continue
if not is_stdin_atty:
print("\n" + sentence_text)
sent = Sentence(str(sentence_text), lgdict, po)
try:
linkages = sent.parse()
except LG_TimerExhausted:
print('Sentence too complex for parsing in ~{} second{}.'.format(
po.max_parse_time,nsuffix(po.max_parse_time)))
continue
if not linkages:
print('Error occurred - sentence ignored.')
continue
if len(linkages) <= 0:
print('Cannot parse the input sentence')
continue
null_count = sent.null_count()
if arg.position:
print(' ' * len(PROMPT), end='')
for p in range (0, len(sentence_text)):
print(p%10, end="")
print()
if null_count == 0:
print("Sentence parsed OK", end='')
linkages = list(linkages)
correction_found = False
# search for correction suggestions
for l in linkages:
for word in l.words():
if word.find(r'.#') > 0:
correction_found = True
break
if correction_found:
break
if correction_found:
print(" - with correction", end='')
if null_count == 0:
print(".")
guess_found = False
if DISPLAY_GUESSES:
# Check the first linkage for regexed/unknown words
for word in linkages[0].words():
# search for something[x]
if re.search(r'\S+\[[^]]+]', word):
guess_found = True
break
# Show results with unlinked words or guesses
if arg.position or guess_found or correction_found or null_count != 0:
print('Sentence has {} unlinked word{}:'.format(
null_count, nsuffix(null_count)))
result_no = 0
uniqe_parse = {}
for linkage in linkages:
words = list(linkage.words())
if str(words) in uniqe_parse:
continue
result_no += 1
uniqe_parse[str(words)] = True
if arg.position:
words_char = []
words_byte = []
for wi, w in enumerate(words):
words_char.append(w + str((linkage.word_char_start(wi), linkage.word_char_end(wi))))
words_byte.append(w + str((linkage.word_byte_start(wi), linkage.word_byte_end(wi))))
print(u"{}: {}".format(result_no, ' '.join(words_char)))
print(u"{}: {}".format(result_no, ' '.join(words_byte)))
else:
print("{}: {}".format(result_no, ' '.join(words)))
if arg.interactive:
print("Interactive session (^D to end):")
import code
code.interact(local=locals())