Last active
August 29, 2015 14:07
-
-
Save he7d3r/cbdffa019659f6f437ef to your computer and use it in GitHub Desktop.
Prints out words (found in a dump.xml) which corresponds to some rule in a file.
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
#!/usr/bin/env python | |
# -*- coding: UTF-8 -*- | |
# Copyright © 2014 He7d3r | |
# License: http://he7d3r.mit-license.org/ | |
""" | |
Prints out words (found in a dump.xml) which corresponds to some rule in a file. | |
Example: | |
python WordsMatchingSalebotRules.py salebot.txt words.txt dump1.xml dump2.xml | |
""" | |
from mw import xml_dump | |
from collections import Counter | |
import sys | |
import re | |
import time | |
from datetime import timedelta | |
globRules = set() | |
startTime = time.time() | |
def page_info(dump, path): | |
global globRules | |
for page in dump: | |
count = Counter() | |
for revision in page: | |
for rule in globRules: | |
m = re.search(rule, revision.text) | |
# Ignore wikitext syntax, n-grams, words in other languages, etc | |
if m and not re.search('[^a-zA-ZáàâãçéêíóôõúüÁÀÂÃÇÉÊÍÓÔÕÚ]', m.group(0)): | |
# TODO: Compute the stem too, while we are at it | |
# TODO: Conver to lower case? | |
count[m.group(0)] += 1 | |
yield page.title, count | |
def run(fileIn, fileOut, dumps): | |
global globRules | |
total = Counter() | |
log = open(fileOut, 'w') | |
for line in open(fileIn).read().splitlines(): | |
# Ignore comments | |
if re.search('^\\s*#', line): | |
continue | |
# Ignore lines containing estrange syntax and lines with a positive score (good words) | |
m = re.search('^\\s*(-\\d+)\\s+\\/(.+)\\/', line) | |
# Ignore regexes which match 2+ words or which do not have letters from A-Z | |
# FIXME: This will ignore rules like "(foo|bar baz)" even if "foo" does not have spaces | |
if m and not re.search(' \w', m.group(2)) and re.search('[a-zA-ZáàâãçéêíóôõúüÁÀÂÃÇÉÊÍÓÔÕÚ]', m.group(2)): | |
# Reduce infinite repetitions to finite ones | |
rule = re.sub('\{(\d+),\}', '{\\1}', m.group(2)) | |
rule = rule.replace( '+', '{1,5}').replace( '*', '{0,5}') | |
globRules.add(re.compile(rule)) | |
#totalPages = 768 # ocwikibooks | |
totalPages = 3617245 # ptwiki | |
p = 0 | |
for title, words in xml_dump.map(dumps, page_info): | |
p += 1 | |
print( '%5.2f%% (+%s) %s' % ( 100*p/totalPages, | |
timedelta(seconds= (time.time() - startTime)*(totalPages - p)/p ), | |
title ) ) | |
for w in words: | |
total[w] += words[w] | |
for s in total.most_common(): | |
# TODO: Ignore words with a single occurrence in the whole history? | |
print('%s\t%s' % s, end='\n', file=log) | |
print('Done in %s.' % timedelta(seconds=time.time() - startTime)) | |
if __name__ == "__main__": | |
if len(sys.argv) < 3: | |
print('Please provide 3+ file names: in, out and dump(s).') | |
sys.exit(1) | |
fileIn = sys.argv[1] | |
fileOut = sys.argv[2] | |
dumps = sys.argv[3:] | |
run(fileIn, fileOut, dumps) |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment