Created
February 21, 2012 03:23
-
-
Save vls/1873402 to your computer and use it in GitHub Desktop.
test script for re2 and re2
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 | |
#encoding=utf-8 | |
import time | |
import os, sys | |
import re | |
import pickle | |
import re2 | |
re2.set_fallback_notification(re2.FALLBACK_WARNING) | |
FLAG_RE2 = False | |
FORCE_UNICODE = True | |
REGEX_DICT = {} | |
REGEX_DICT['fengting1'] = [] | |
REGEX_DICT['fengting2'] = [] | |
REGEX_DICT['geli'] = [] | |
REGEX_DICT['jinyan'] = [] | |
REGEX_DICT['nick'] = [] | |
REGEX_DICT['warning'] = [] | |
REGEX_DICT['exc'] = [] | |
global_arr = [ | |
# ('regex/fengting1.txt', 'fengting1', u'封停'), | |
# ('regex/fengting2.txt', 'fengting2', u'封停'), | |
# ('regex/geli.txt', 'geli', u'隔离'), | |
# ['regex/clear_jinyan.txt', 'jinyan', u'禁言'], | |
['regex/clear_warning.txt', 'warning', u'警告'], | |
# ('regex/nickname.txt', 'nick', u'昵称'), | |
#('regex/exception.txt', 'exc', u'例外'), | |
] | |
REG_DUMPS = 'regex.dumps' | |
def load(): | |
global REGEX_DICT | |
mod_re = re2 if FLAG_RE2 else re | |
print "Module = %s" % (mod_re) | |
count_re = 0 | |
count_re2 = 0 | |
for fname, key, _ in global_arr: | |
l = REGEX_DICT[key] | |
for line in open(fname): | |
line = line.strip() | |
if FORCE_UNICODE: | |
line = line.decode('utf-8') | |
regex = mod_re.compile(line) | |
l.append((regex, line)) | |
if isinstance(regex, re2.Pattern): | |
count_re2 += 1 | |
else: | |
count_re += 1 | |
assert l | |
print "count re = %d, re2 = %d" % (count_re, count_re2) | |
total_time = 0 | |
def match(nick, msg): | |
global REGEX_DICT | |
global total_time | |
for fname, key, desc in global_arr: | |
#flag_nick = 'nickname' in fname | |
l = REGEX_DICT[key] | |
for regex, pat in l: | |
# if not isinstance(regex, re2.Pattern): | |
# # Python re module, need unicode | |
# if not isinstance(nick, unicode): | |
# nick = nick.decode('utf-8') | |
# | |
# if not isinstance(msg, unicode): | |
# msg = msg.decode('utf-8') | |
#if flag_nick: | |
# if regex.search(nick): | |
# return True, desc, pat | |
# continue | |
if regex.search(msg): | |
return True, desc, pat | |
return False, '', '' | |
REGEX_FILTER = re.compile(r'#c[0-9a-fA-F]{6}|#[RGBKYWbunr#]|#\d{1,3}') | |
OUTPUT_DIR = 'output' | |
try: | |
os.makedirs(OUTPUT_DIR) | |
except: | |
pass | |
def export_fname(fname): | |
line_dict = {} | |
servername = os.path.basename(os.path.dirname(fname)) | |
if not servername: | |
servername = 'NA' | |
basename = os.path.basename(fname) | |
count = 0 | |
match_count = 0 | |
oname = 'plain-%s-%s.result.txt' % (servername, basename) | |
oname = os.path.join(OUTPUT_DIR, oname) | |
with open(oname, 'w') as wf: | |
for line in open(fname): | |
count += 1 | |
line = line.strip() | |
if FORCE_UNICODE: | |
line =line.decode('utf-8') | |
arr = line.split(None, 7) | |
chan = arr[3] | |
#if u'INFO|CHAT' not in chan: | |
# continue | |
#if u'世界' in chan: | |
# continue | |
try: | |
nick = arr[5] | |
msg = arr[-1] | |
msg = REGEX_FILTER.sub('', msg) | |
flag, desc, pat = match(nick, msg) | |
if flag: | |
ret_arr = [line, desc, pat ] | |
us = '\t'.join(ret_arr) | |
if FORCE_UNICODE: | |
us = us.encode('utf-8') | |
wf.write(us) | |
wf.write('\n') | |
match_count += 1 | |
except IndexError: | |
continue | |
print "match count = %s" % (match_count) | |
def run_re2_test(): | |
sys.argv = sys.argv[:1] | |
sys.argv.extend(['test-utf-8.log', '1']) | |
main() | |
def run_re_test(): | |
sys.argv = sys.argv[:1] | |
sys.argv.extend(['test-utf-8.log']) | |
main() | |
def main(): | |
if len(sys.argv) <= 1: | |
print 'Usage: %s <file>' % (sys.argv[0]) | |
sys.exit(0) | |
if len(sys.argv) >= 3: | |
print 'using re2!!!' | |
global FLAG_RE2, FORCE_UNICODE | |
FLAG_RE2 = True | |
FORCE_UNICODE = False | |
global global_arr | |
for i, (_, _, name) in enumerate(global_arr): | |
if not FORCE_UNICODE: | |
global_arr[i][2] = name.encode('utf-8') | |
t0 = time.time() | |
load() | |
print >> sys.stderr, 'load complete. time = %s' % (time.time() - t0) | |
t0 = time.time() | |
if os.path.isdir(sys.argv[1]): | |
import glob | |
import multiprocessing | |
fnames = glob.glob("%s/*/*.log" % sys.argv[1]) | |
pool = multiprocessing.Pool(multiprocessing.cpu_count() - 1) | |
pool.map_async(export_fname, fnames) | |
pool.close() | |
pool.join() | |
else: | |
export_fname(sys.argv[1]) | |
print >> sys.stderr, 'process time = %s' % (time.time() - t0) | |
print >> sys.stderr, 'total process time = %s' % (total_time) | |
if __name__ == '__main__': | |
main() |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment