new-words

annotate new-words.py @ 54:e25de9ea9184

find changesets by author, revision, files, or words in the commit message
new-words.py is almost ready
author Igor Chubin <igor@chub.in>
date Tue Nov 01 20:19:18 2011 +0100 (2011-11-01)
parents f583256b7ab1
children 2a1a25e61872
rev   line source
igor@37 1 #!/usr/bin/env python
igor@38 2 # -*- coding: utf-8 -*-
igor@37 3
igor@40 4 from __future__ import with_statement
igor@38 5 import codecs
igor@49 6 import difflib
igor@38 7 import logging
igor@38 8 import os
igor@37 9 import optparse
igor@38 10 import re
igor@38 11 import subprocess
igor@38 12 import sys
igor@38 13 import Stemmer
igor@54 14 import tempfile
igor@42 15 try:
igor@42 16 import psyco
igor@42 17 psyco.full()
igor@42 18 except:
igor@42 19 pass
igor@38 20
igor@38 21 config = {
igor@38 22 'config_directory': os.environ['HOME'] + '/.new-words',
igor@38 23 'language': 'en',
igor@38 24 }
igor@38 25
igor@38 26 logging.basicConfig(filename='/tmp/new-words-py.log', level=logging.DEBUG)
igor@38 27
igor@38 28 class Normalizator:
igor@38 29 def __init__(self, language, linked_words={}):
igor@38 30 stemmer_algorithm = {
igor@38 31 'de' : 'german',
igor@38 32 'en' : 'english',
igor@51 33 'es' : 'spanish',
igor@38 34 'ru' : 'russian',
igor@51 35 'it' : 'italian',
igor@38 36 'uk' : 'ukrainian',
igor@38 37 }
igor@38 38 self.stemmer = Stemmer.Stemmer(stemmer_algorithm[language])
igor@38 39 self.linked_words = linked_words
igor@38 40
igor@38 41 def normalize(self, word):
igor@38 42 word_chain = []
igor@38 43 while word in self.linked_words and not word in word_chain:
igor@38 44 word_chain.append(word)
igor@38 45 word = self.linked_words[word]
igor@38 46 return self.stemmer.stemWord(word.lower())
igor@37 47
igor@47 48 def best_word_from_group(self, wordpairs_group):
igor@47 49 """Returns the word that is the most relevant to the wordpairs_group.
igor@47 50
igor@47 51 At the moment: returns the word with minimal length"""
igor@49 52
igor@49 53 def f(x, y):
igor@49 54 return difflib.SequenceMatcher(
igor@49 55 None,
igor@49 56 #(x[-2:] == 'en' and x[:-2].lower() or x.lower()),
igor@49 57 x.lower(),
igor@49 58 y.lower()).ratio()
igor@47 59
igor@47 60 minimal_length = min(len(pair[1]) for pair in wordpairs_group)
igor@49 61 best_match = list(x[1] for x in sorted(
igor@47 62 (x for x in wordpairs_group if len(x[1]) == minimal_length),
igor@47 63 key=lambda x:x[0],
igor@47 64 reverse=True))[0]
igor@47 65
igor@51 66 return best_match
igor@51 67
igor@49 68 suggestions = self.dictionary_suggestions(best_match)
igor@49 69 if len(suggestions) == 1:
igor@49 70 return best_match
igor@49 71
igor@49 72 verb = False
igor@49 73 corrected_best_match = best_match
igor@49 74 if best_match[-2:] == 'et':
igor@49 75 word = best_match[:-1]+"n"
igor@49 76 sugg = self.dictionary_suggestions(word)
igor@49 77 if len(sugg) == 1:
igor@49 78 return word
igor@49 79 suggestions += sugg
igor@49 80 corrected_best_match = word
igor@49 81 corrected_best_match = best_match[:-2]
igor@49 82 verb = True
igor@49 83
igor@49 84 if best_match[-1] == 't':
igor@49 85 word = best_match[:-1]+"en"
igor@49 86 sugg = self.dictionary_suggestions(word)
igor@49 87 if len(sugg) == 1:
igor@49 88 return word
igor@49 89 suggestions += sugg
igor@49 90 corrected_best_match = best_match[:-1]
igor@49 91 verb = True
igor@49 92
igor@49 93 if corrected_best_match[0].lower() == corrected_best_match[0]:
igor@49 94 suggestions = [ x for x in suggestions
igor@49 95 if x[0].lower() == x[0] ]
igor@49 96
igor@49 97 if suggestions == []:
igor@49 98 return best_match+"_"
igor@49 99 return best_match+" "+(" ".join(
igor@49 100 sorted(
igor@49 101 suggestions,
igor@49 102 key = lambda x: f(x, corrected_best_match),
igor@49 103 reverse = True
igor@49 104 )
igor@49 105 )
igor@49 106 )
igor@49 107
igor@49 108 def dictionary_suggestions(self, word):
igor@49 109 return [
igor@49 110 x.decode('utf-8').rstrip('\n')
igor@49 111 for x
igor@49 112 in subprocess.Popen(
igor@49 113 ["de-variants", word],
igor@49 114 stdout=subprocess.PIPE
igor@49 115 ).stdout.readlines() ]
igor@49 116
igor@49 117
igor@37 118 parser = optparse.OptionParser()
igor@37 119
igor@37 120 parser.add_option(
igor@37 121 "-a", "--no-marks",
igor@37 122 help="don't add marks (and don't save marks added by user)",
igor@37 123 action="store_true",
igor@37 124 dest="no_marks")
igor@37 125
igor@37 126 parser.add_option(
igor@37 127 "-c", "--compressed",
igor@37 128 help="show compressed wordlist: one word per group",
igor@37 129 action="store_true",
igor@37 130 dest="compressed")
igor@37 131
igor@37 132 parser.add_option(
igor@37 133 "-k", "--known-words",
igor@37 134 help="put higher words that are similar to the known words (only for English)",
igor@37 135 action="store_true",
igor@37 136 dest="compressed")
igor@37 137
igor@37 138 parser.add_option(
igor@37 139 "-l", "--language",
igor@37 140 help="specify language of text",
igor@37 141 action="store",
igor@37 142 dest="language")
igor@37 143
igor@37 144 parser.add_option(
igor@54 145 "-f", "--allowed-words",
igor@54 146 help="file with list of allowed words (words that will be shown in the output)",
igor@54 147 action="store",
igor@54 148 dest="allowed_words")
igor@54 149
igor@54 150 parser.add_option(
igor@54 151 "-X", "--function",
igor@38 152 help="filter through subsystem [INTERNAL]",
igor@38 153 action="store",
igor@38 154 dest="function")
igor@38 155
igor@38 156 parser.add_option(
igor@37 157 "-m", "--merge-tag",
igor@37 158 help="merge words tagged with specified tag into the main vocabulary",
igor@37 159 action="store",
igor@37 160 dest="merge_tag")
igor@37 161
igor@37 162 parser.add_option(
igor@37 163 "-M", "--merge-tagged",
igor@37 164 help="merge words tagged with ANY tag into the main vocabulary",
igor@37 165 action="store_true",
igor@37 166 dest="merge_tagged")
igor@37 167
igor@37 168 parser.add_option(
igor@37 169 "-n", "--non-interactive",
igor@37 170 help="non-interactive mode (don't run vi)",
igor@37 171 action="store_true",
igor@37 172 dest="non_interactive")
igor@37 173
igor@37 174 parser.add_option(
igor@37 175 "-N", "--no-filter",
igor@37 176 help="switch off known words filtering",
igor@37 177 action="store_true",
igor@37 178 dest="no_filter")
igor@37 179
igor@37 180 parser.add_option(
igor@37 181 "-p", "--pages",
igor@37 182 help="work with specified pages only (pages = start-stop/total )",
igor@37 183 action="store",
igor@37 184 dest="pages")
igor@37 185
igor@37 186 parser.add_option(
igor@48 187 "-d", "--delete-tag",
igor@48 188 help="delete subvocabulary of specified tag",
igor@37 189 action="store",
igor@48 190 dest="delete_tag")
igor@37 191
igor@37 192 parser.add_option(
igor@54 193 "-R", "--show-range-percentage",
igor@54 194 help="show only words that cover specified percentage of the text, skip the rest",
igor@54 195 action="store",
igor@54 196 dest="show_range_percentage")
igor@54 197
igor@54 198 parser.add_option(
igor@37 199 "-s", "--text-stats",
igor@37 200 help="show the text statistics (percentage of known words and so on) and exit",
igor@37 201 action="store_true",
igor@37 202 dest="text_stats")
igor@37 203
igor@37 204 parser.add_option(
igor@37 205 "-S", "--voc-stats",
igor@37 206 help="show your vocabulary statistics (number of words and word groups)",
igor@37 207 action="store_true",
igor@37 208 dest="voc_stats")
igor@37 209
igor@37 210 parser.add_option(
igor@37 211 "-t", "--tag",
igor@37 212 help="tag known words with tag",
igor@37 213 action="store",
igor@37 214 dest="tag")
igor@37 215
igor@37 216 parser.add_option(
igor@37 217 "-T", "--show-tags",
igor@37 218 help="tag known words with tag",
igor@37 219 action="store_true",
igor@37 220 dest="show_tags")
igor@37 221
igor@37 222 parser.add_option(
igor@37 223 "-2", "--two-words",
igor@37 224 help="find 2 words' sequences",
igor@37 225 action="store_true",
igor@37 226 dest="two_words")
igor@37 227
igor@37 228 parser.add_option(
igor@37 229 "-3", "--three-words",
igor@37 230 help="find 3 words' sequences",
igor@37 231 action="store_true",
igor@37 232 dest="three_words")
igor@37 233
igor@38 234 def readlines_from_file(filename):
igor@38 235 res = []
igor@38 236 with codecs.open(filename, "r", "utf-8") as f:
igor@38 237 for line in f.readlines():
igor@38 238 res += [line]
igor@38 239 return res
igor@38 240
igor@54 241 def readlines_from_url(url):
igor@54 242 return [x.decode('utf-8') for x in
igor@54 243 subprocess.Popen(
igor@54 244 "lynx -dump '{url}' | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'".format(url=url),
igor@54 245 shell = True,
igor@54 246 stdout = subprocess.PIPE,
igor@54 247 stderr = subprocess.STDOUT
igor@54 248 ).communicate()[0].split('\n')
igor@54 249 ]
igor@54 250
igor@38 251 def readlines_from_stdin():
igor@38 252 return codecs.getreader("utf-8")(sys.stdin).readlines()
igor@38 253
igor@38 254 def words_from_line(line):
igor@38 255 line = line.rstrip('\n')
igor@38 256 #return re.split('(?:\s|[*\r,.:#@()+=<>$;"?!|\[\]^%&~{}«»–])+', line)
igor@38 257 #return re.split('[^a-zA-ZäöëüßÄËÖÜß]+', line)
igor@44 258 return re.compile("(?!['_])(?:\W)+", flags=re.UNICODE).split(line)
igor@38 259
igor@44 260 def get_words(lines, group_by=[1]):
igor@38 261 """
igor@38 262 Returns hash of words in a file
igor@38 263 word => number
igor@38 264 """
igor@38 265 result = {}
igor@44 266 (a, b, c) = ("", "", "")
igor@38 267 for line in lines:
igor@38 268 words = words_from_line(line)
igor@38 269 for word in words:
igor@41 270 if re.match('[0-9]*$', word):
igor@41 271 continue
igor@38 272 result.setdefault(word, 0)
igor@38 273 result[word] += 1
igor@44 274 if 2 in group_by and a != "" and b != "":
igor@44 275 w = "%s_%s" % (a,b)
igor@44 276 result.setdefault(w, 0)
igor@44 277 result[w] += 1
igor@44 278 if 3 in group_by and not "" in [a,b,c]:
igor@44 279 w = "%s_%s_%s" % (a,b,c)
igor@44 280 result.setdefault(w, 0)
igor@44 281 result[w] += 1
igor@44 282 (a,b,c) = (b, c, word)
igor@44 283
igor@44 284 logging.debug(result)
igor@38 285 return result
igor@38 286
igor@54 287 def voc_filename():
igor@54 288 return "%s/%s.txt"%(config['config_directory'], config['language'])
igor@54 289
igor@38 290 def load_vocabulary():
igor@54 291 return get_words(readlines_from_file(voc_filename()))
igor@38 292
igor@38 293 def notes_filenames():
igor@38 294 return ["%s/notes-%s.txt"%(config['config_directory'], config['language'])]
igor@38 295
igor@38 296 def load_notes(files):
igor@38 297 notes = {}
igor@38 298 for filename in files:
igor@39 299 with codecs.open(filename, "r", "utf-8") as f:
igor@38 300 for line in f.readlines():
igor@38 301 (word, note) = re.split('\s+', line.rstrip('\n'), maxsplit=1)
igor@38 302 notes.setdefault(word, {})
igor@38 303 notes[word][filename] = note
igor@38 304 return notes
igor@38 305
igor@39 306 def add_notes(lines, notes):
igor@39 307 notes_filename = notes_filenames()[0]
igor@39 308 result = []
igor@39 309 for line in lines:
igor@39 310 if line.startswith('#'):
igor@39 311 result += [line]
igor@39 312 else:
igor@39 313 match_object = re.search('^\s*\S+\s*(\S+)', line)
igor@39 314 if match_object:
igor@39 315 word = match_object.group(1)
igor@39 316 if word in notes:
igor@39 317 if notes_filename in notes[word]:
igor@39 318 line = line.rstrip('\n')
igor@39 319 line = "%-30s %s\n" % (line, notes[word][notes_filename])
igor@39 320 result += [line]
igor@39 321 else:
igor@39 322 result += [line]
igor@39 323 else:
igor@39 324 result += [line]
igor@39 325 return result
igor@39 326
igor@39 327 def remove_notes(lines, notes_group):
igor@39 328 notes_filename = notes_filenames()[0]
igor@39 329 notes = {}
igor@39 330 for k in notes_group.keys():
igor@39 331 if notes_filename in notes_group[k]:
igor@39 332 notes[k] = notes_group[k][notes_filename]
igor@39 333
igor@39 334 result = []
igor@39 335 for line in lines:
igor@39 336 line = line.rstrip('\n')
igor@39 337 match_object = re.match('(\s+)(\S+)(\s+)(\S+)(\s+)(.*)', line)
igor@39 338 if match_object:
igor@39 339 result.append("".join([
igor@39 340 match_object.group(1),
igor@39 341 match_object.group(2),
igor@39 342 match_object.group(3),
igor@39 343 match_object.group(4),
igor@39 344 "\n"
igor@39 345 ]))
igor@39 346 notes[match_object.group(4)] = match_object.group(6)
igor@39 347 else:
igor@39 348 result.append(line+"\n")
igor@39 349
igor@39 350 save_notes(notes_filename, notes)
igor@39 351 return result
igor@39 352
igor@39 353 def save_notes(filename, notes):
igor@39 354 lines = []
igor@39 355 saved_words = []
igor@39 356 with codecs.open(filename, "r", "utf-8") as f:
igor@39 357 for line in f.readlines():
igor@39 358 (word, note) = re.split('\s+', line.rstrip('\n'), maxsplit=1)
igor@39 359 if word in notes:
igor@39 360 line = "%-29s %s\n" % (word, notes[word])
igor@39 361 saved_words.append(word)
igor@39 362 lines.append(line)
igor@39 363 for word in [x for x in notes.keys() if not x in saved_words]:
igor@39 364 line = "%-29s %s\n" % (word, notes[word])
igor@39 365 lines.append(line)
igor@39 366
igor@39 367 with codecs.open(filename, "w", "utf-8") as f:
igor@39 368 for line in lines:
igor@39 369 f.write(line)
igor@39 370
igor@39 371
igor@38 372 def substract_dictionary(dict1, dict2):
igor@38 373 """
igor@38 374 returns dict1 - dict2
igor@38 375 """
igor@38 376 result = {}
igor@38 377 for (k,v) in dict1.items():
igor@38 378 if not k in dict2:
igor@38 379 result[k] = v
igor@38 380 return result
igor@38 381
igor@38 382 def dump_words(words, filename):
igor@38 383 with codecs.open(filename, "w+", "utf-8") as f:
igor@38 384 for word in words.keys():
igor@38 385 f.write(("%s\n"%word)*words[word])
igor@38 386
igor@38 387 def error_message(text):
igor@38 388 print text
igor@38 389
igor@40 390 def find_wordgroups_weights(word_pairs, normalizator):
igor@38 391 weight = {}
igor@40 392 for (num, word) in word_pairs:
igor@38 393 normalized = normalizator.normalize(word)
igor@38 394 weight.setdefault(normalized, 0)
igor@40 395 weight[normalized] += num
igor@38 396 return weight
igor@38 397
igor@38 398 def find_linked_words(notes):
igor@38 399 linked_words = {}
igor@38 400 for word in notes.keys():
igor@38 401 for note in notes[word].values():
igor@38 402 if "@" in note:
igor@38 403 result = re.search(r'\@(\S*)', note)
igor@38 404 if result:
igor@38 405 main_word = result.group(1)
igor@38 406 if main_word:
igor@38 407 linked_words[word] = main_word
igor@38 408 return linked_words
igor@38 409
igor@40 410 def compare_word_pairs(pair1, pair2, wgw, normalizator, linked_words):
igor@40 411 (num1, word1) = pair1
igor@40 412 (num2, word2) = pair2
igor@38 413
igor@38 414 normalized_word1 = normalizator.normalize(word1)
igor@38 415 normalized_word2 = normalizator.normalize(word2)
igor@38 416
igor@38 417 cmp_res = cmp(wgw[normalized_word1], wgw[normalized_word2])
igor@38 418 if cmp_res != 0:
igor@38 419 return cmp_res
igor@38 420 else:
igor@38 421 cmp_res = cmp(normalized_word1, normalized_word2)
igor@38 422 if cmp_res != 0:
igor@38 423 return cmp_res
igor@38 424 else:
igor@38 425 return cmp(int(num1), int(num2))
igor@38 426
igor@47 427
igor@48 428 def print_words_sorted(
igor@48 429 word_pairs,
igor@48 430 stats,
igor@48 431 normalizator,
igor@48 432 print_stats=True,
igor@48 433 stats_only=False,
igor@48 434 compressed_wordlist=False,
igor@48 435 show_range=0,
igor@48 436 show_range_percentage=0,
igor@48 437 ):
igor@54 438 result = []
igor@40 439 if stats_only:
igor@54 440 #codecs.getwriter("utf-8")(sys.stdout).write(
igor@54 441 result.append(
igor@43 442 " ".join([
igor@43 443 "%-10s" % x for x in [
igor@43 444 "LANG",
igor@43 445 "KNOWN%",
igor@43 446 "UNKNOWN%",
igor@43 447 "KNOWN",
igor@43 448 "TOTAL",
igor@43 449 "WPS",
igor@43 450 "UWPS*10"
igor@43 451 ]]) + "\n")
igor@54 452 result.append(
igor@43 453 " ".join([
igor@43 454 "%(language)-10s",
igor@43 455 "%(percentage)-10.2f",
igor@43 456 "%(percentage_unknown)-10.2f",
igor@43 457 "%(total_known)-11d"
igor@43 458 "%(total)-11d"
igor@43 459 "%(wps)-11d"
igor@43 460 "%(uwps)-11d"
igor@43 461 ]) % stats + "\n")
igor@54 462 return "".join(result)
igor@38 463
igor@40 464 if print_stats:
igor@54 465 result.append(
igor@43 466 "# %(language)s, %(percentage)-7.2f, <%(total_known)s/%(total)s>, <%(groups)s/%(words)s>\n" % stats)
igor@38 467
igor@40 468 level_lines = range(int(float(stats['percentage']))/5*5+5,95,5)+range(90,102)
igor@40 469 known = int(stats['total_known'])
igor@40 470 total = int(stats['total'])
igor@40 471 current_level = 0
igor@47 472 old_normalized_word = None
igor@47 473 words_of_this_group = []
igor@48 474 printed_words = 0
igor@40 475 for word_pair in word_pairs:
igor@47 476
igor@47 477 normalized_word = normalizator.normalize(word_pair[1])
igor@47 478 if old_normalized_word and old_normalized_word != normalized_word:
igor@47 479 if compressed_wordlist:
igor@49 480 compressed_word_pair = (
igor@49 481 sum(x[0] for x in words_of_this_group),
igor@49 482 normalizator.best_word_from_group(words_of_this_group)
igor@49 483 )
igor@54 484 result.append("%10s %s\n" % compressed_word_pair)
igor@48 485 printed_words += 1
igor@47 486 words_of_this_group = []
igor@47 487
igor@47 488 old_normalized_word = normalized_word
igor@47 489 words_of_this_group.append(word_pair)
igor@47 490
igor@47 491 if not compressed_wordlist:
igor@54 492 result.append("%10s %s\n" % word_pair)
igor@48 493 printed_words += 1
igor@47 494
igor@47 495
igor@40 496 known += word_pair[0]
igor@40 497 if 100.0*known/total >= level_lines[0]:
igor@40 498 current_level = level_lines[0]
igor@40 499 while 100.0*known/total > level_lines[0]:
igor@40 500 current_level = level_lines[0]
igor@40 501 level_lines = level_lines[1:]
igor@54 502 result.append("# %s\n" % current_level)
igor@38 503
igor@48 504 if show_range >0 and printed_words >= show_range:
igor@48 505 break
igor@48 506 if show_range_percentage >0 and 100.0*known/total >= show_range_percentage:
igor@48 507 break
igor@48 508
igor@54 509 return result
igor@39 510
igor@53 511 def parse_parts_description(parts_description):
igor@53 512 """
igor@53 513 Returns triad (start, stop, step)
igor@53 514 basing on parts_description string.
igor@53 515 from-to/step
igor@53 516 from+delta/step
igor@53 517 """
igor@53 518
igor@53 519 try:
igor@53 520 (a, step) = parts_description.split("/", 1)
igor@53 521 step = int(step)
igor@53 522 start = 0
igor@53 523 stop = 0
igor@53 524 if '-' in a:
igor@53 525 (start, stop) = a.split("-", 1)
igor@53 526 start = int(start)
igor@53 527 stop = int(stop)
igor@53 528 elif '+' in a:
igor@53 529 (start, stop) = a.split("+", 1)
igor@53 530 start = int(start)
igor@53 531 stop = int(stop)
igor@53 532 else:
igor@53 533 start = int(a)
igor@53 534 stop = start + 1
igor@53 535 return (start, stop, step)
igor@53 536
igor@53 537 except:
igor@54 538 raise ValueError("Parts description must be in format: num[[+-]num]/num; this [%s] is incorrect" % parts_description)
igor@53 539
igor@53 540
igor@53 541 def take_part(lines, part_description = None):
igor@53 542 if part_description == None:
igor@53 543 return lines
igor@53 544 (start, stop, step) = parse_parts_description(part_description)
igor@53 545 n = len(lines)
igor@53 546 part_size = (1.0*n) / step
igor@53 547 result = []
igor@53 548 for i in range(n):
igor@54 549 if i >= start * part_size and i <= stop * part_size:
igor@54 550 result += [lines[i]]
igor@53 551 return result
igor@53 552
igor@40 553 def filter_get_words_group_words_add_stat(args):
igor@40 554 vocabulary = load_vocabulary()
igor@40 555 notes = load_notes(notes_filenames())
igor@54 556
igor@54 557 if len(args) > 0:
igor@54 558 if 'http://' in args[0]:
igor@54 559 input_lines = readlines_from_url(args[0])
igor@54 560 else:
igor@54 561 input_lines = readlines_from_file(args[0])
igor@54 562 else:
igor@54 563 input_lines = readlines_from_stdin()
igor@54 564
igor@54 565 if len(input_lines) == 0:
igor@54 566 print >> sys.stderr, "Nothing to do, standard input is empty, exiting."
igor@54 567 sys.exit(1)
igor@54 568
igor@54 569 lines = take_part(input_lines, config.get('pages', ''))
igor@54 570
igor@54 571 (_, original_text_tempfile) = tempfile.mkstemp(prefix='new-word')
igor@54 572 with codecs.open(original_text_tempfile, "w", "utf-8") as f:
igor@54 573 f.write("".join(lines))
igor@54 574
igor@44 575 group_by = [1]
igor@48 576
igor@54 577 if 'two_words' in config:
igor@44 578 group_by.append(2)
igor@54 579 if 'three_words' in config:
igor@44 580 group_by.append(3)
igor@44 581 words = get_words(lines, group_by)
igor@43 582 stats_only = False
igor@54 583 if 'text_stats' in config:
igor@43 584 stats_only = True
igor@40 585
igor@47 586 compressed_wordlist = False
igor@54 587 if 'compressed' in config:
igor@47 588 compressed_wordlist = True
igor@47 589
igor@48 590 show_range = os.environ.get('SHOW_RANGE', '')
igor@48 591 if show_range != '':
igor@48 592 show_range = int(show_range)
igor@48 593 else:
igor@48 594 show_range = 0
igor@54 595
igor@54 596 if 'show_range_percentage' in config:
igor@54 597 show_range_percentage = int(config['show_range_percentage'])
igor@48 598 else:
igor@48 599 show_range_percentage = 0
igor@48 600
igor@44 601
igor@40 602 stats = {}
igor@40 603 stats['total'] = sum(words[x] for x in words.keys())
igor@54 604 if not 'no_filter' in config:
igor@45 605 words = substract_dictionary(words, vocabulary)
igor@40 606
igor@40 607 stats['total_unknown'] = sum(words[x] for x in words.keys())
igor@40 608 stats['total_known'] = stats['total'] - stats['total_unknown']
igor@43 609 stats['percentage'] = 100.0*stats['total_known']/stats['total']
igor@43 610 stats['percentage_unknown'] = 100.0-100.0*stats['total_known']/stats['total']
igor@40 611 stats['groups'] = 0
igor@40 612 stats['words'] = len(words)
igor@43 613 stats['sentences'] = 0 #FIXME
igor@43 614 stats['wps'] = 0 #FIXME
igor@43 615 stats['uwps'] = 0 #FIXME
igor@40 616 stats['language'] = config['language']
igor@40 617
igor@40 618 linked_words = find_linked_words(notes)
igor@40 619 normalizator = Normalizator(config['language'], linked_words)
igor@40 620
igor@50 621 # filter words by allowed_words_filter
igor@54 622 if 'allowed_words' in config:
igor@54 623 allowed_words_filename = config['allowed_words']
igor@50 624 normalized_allowed_words = [
igor@50 625 normalizator.normalize(w.rstrip('\n'))
igor@50 626 for w in readlines_from_file(allowed_words_filename)
igor@50 627 ]
igor@50 628
igor@50 629 result = {}
igor@50 630 for w, wn in words.iteritems():
igor@50 631 if normalizator.normalize(w) in normalized_allowed_words:
igor@50 632 result[w] = wn
igor@50 633 words = result
igor@50 634
igor@44 635 words_with_freq = []
igor@40 636 for k in sorted(words.keys(), key=lambda k: words[k], reverse=True):
igor@44 637 words_with_freq.append((words[k], k))
igor@40 638
igor@44 639 wgw = find_wordgroups_weights(words_with_freq, normalizator)
igor@45 640 if 'WORDS_GROUPING' in os.environ and os.environ['WORDS_GROUPING'] == 'YES':
igor@45 641 words_with_freq = sorted(
igor@44 642 words_with_freq,
igor@40 643 cmp=lambda x,y:compare_word_pairs(x,y, wgw, normalizator, linked_words),
igor@40 644 reverse=True)
igor@40 645
igor@54 646 output = print_words_sorted(
igor@47 647 words_with_freq,
igor@47 648 stats,
igor@47 649 normalizator,
igor@47 650 stats_only=stats_only,
igor@48 651 compressed_wordlist=compressed_wordlist,
igor@48 652 show_range=show_range,
igor@48 653 show_range_percentage=show_range_percentage,
igor@47 654 )
igor@40 655
igor@54 656
igor@54 657 if ('non_interactive' in config or 'text_stats' in config):
igor@54 658 codecs.getwriter("utf-8")(sys.stdout).write("".join(output))
igor@54 659 else:
igor@54 660 (_, temp1) = tempfile.mkstemp(prefix='new-word')
igor@54 661 (_, temp2) = tempfile.mkstemp(prefix='new-word')
igor@54 662
igor@54 663 with codecs.open(temp1, "w", "utf-8") as f:
igor@54 664 f.write("".join(output))
igor@54 665 with codecs.open(temp2, "w", "utf-8") as f:
igor@54 666 f.write("".join(add_notes(output, notes)))
igor@54 667
igor@54 668 os.putenv('ORIGINAL_TEXT', original_text_tempfile)
igor@54 669 os.system((
igor@54 670 "vim"
igor@54 671 " -c 'setlocal spell spelllang={language}'"
igor@54 672 " -c 'set keywordprg={language}'"
igor@54 673 " -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=,48-255'"
igor@54 674 " {filename}"
igor@54 675 " < /dev/tty > /dev/tty"
igor@54 676 ).format(language=config['language'], filename=temp2))
igor@54 677
igor@54 678 lines = remove_notes(readlines_from_file(temp2), notes)
igor@54 679
igor@54 680 # compare lines_before and lines_after and return deleted words
igor@54 681 lines_before = output
igor@54 682 lines_after = lines
igor@54 683 deleted_words = []
igor@54 684
igor@54 685 for line in lines_before:
igor@54 686 if line not in lines_after:
igor@54 687 line = line.strip()
igor@54 688 if ' ' in line:
igor@54 689 word = re.split('\s+', line, 1)[1]
igor@54 690 if ' ' in word:
igor@54 691 word = re.split('\s+', word, 1)[0]
igor@54 692 deleted_words.append(word)
igor@54 693
igor@54 694 with codecs.open(voc_filename(), "a", "utf-8") as f:
igor@54 695 f.write("\n".join(deleted_words + ['']))
igor@54 696
igor@54 697 os.unlink(temp1)
igor@54 698 os.unlink(temp2)
igor@54 699
igor@54 700 os.unlink(original_text_tempfile)
igor@54 701
igor@37 702 (options, args) = parser.parse_args()
igor@38 703 if options.language:
igor@38 704 config['language'] = options.language
igor@37 705
igor@54 706 if options.pages:
igor@54 707 config['pages'] = options.pages
igor@54 708 else:
igor@54 709 config['pages'] = ""
igor@54 710
igor@54 711 if options.allowed_words:
igor@54 712 config['allowed_words'] = options.allowed_words
igor@54 713
igor@54 714 if options.show_range_percentage:
igor@54 715 config['show_range_percentage'] = options.show_range_percentage
igor@54 716
igor@54 717 if options.non_interactive:
igor@54 718 config['non_interactive'] = True
igor@54 719
igor@54 720 if options.text_stats:
igor@54 721 config['text_stats'] = True
igor@54 722
igor@54 723 if options.compressed:
igor@54 724 config['compressed'] = True
igor@54 725
igor@54 726 if options.no_filter:
igor@54 727 config['no_filter'] = True
igor@54 728
igor@54 729 if options.two_words:
igor@54 730 config['two_words'] = True
igor@54 731
igor@54 732 if options.three_words:
igor@54 733 config['three_words'] = True
igor@54 734
igor@38 735 if options.function:
igor@38 736 function_names = {
igor@40 737 'get_words_group_words_add_stat': filter_get_words_group_words_add_stat,
igor@38 738 }
igor@38 739 if options.function in function_names:
igor@38 740 function_names[options.function](args)
igor@38 741 else:
igor@38 742 error_message("Unkown function %s.\nAvailable functions:\n%s" % (
igor@38 743 options.function, "".join([" "+x for x in sorted(function_names.keys())])))
igor@38 744 sys.exit(1)
igor@37 745
igor@37 746
igor@37 747
igor@37 748
igor@38 749 #os.system("vim")
igor@37 750