new-words

annotate new-words.py @ 64:c3adf6452eda

find changesets by author, revision, files, or words in the commit message
lingvo-en-ru.pl script moved to misc/
author Igor Chubin <igor@chub.in>
date Sat Nov 12 14:03:54 2011 +0100 (2011-11-12)
parents 3682038403ad
children 5a003076eb11
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@63 32 'fr' : 'french',
igor@38 33 'en' : 'english',
igor@51 34 'es' : 'spanish',
igor@38 35 'ru' : 'russian',
igor@51 36 'it' : 'italian',
igor@38 37 'uk' : 'ukrainian',
igor@38 38 }
igor@38 39 self.stemmer = Stemmer.Stemmer(stemmer_algorithm[language])
igor@38 40 self.linked_words = linked_words
igor@38 41
igor@38 42 def normalize(self, word):
igor@38 43 word_chain = []
igor@38 44 while word in self.linked_words and not word in word_chain:
igor@38 45 word_chain.append(word)
igor@38 46 word = self.linked_words[word]
igor@38 47 return self.stemmer.stemWord(word.lower())
igor@37 48
igor@47 49 def best_word_from_group(self, wordpairs_group):
igor@47 50 """Returns the word that is the most relevant to the wordpairs_group.
igor@47 51
igor@47 52 At the moment: returns the word with minimal length"""
igor@49 53
igor@49 54 def f(x, y):
igor@49 55 return difflib.SequenceMatcher(
igor@49 56 None,
igor@49 57 #(x[-2:] == 'en' and x[:-2].lower() or x.lower()),
igor@49 58 x.lower(),
igor@49 59 y.lower()).ratio()
igor@47 60
igor@47 61 minimal_length = min(len(pair[1]) for pair in wordpairs_group)
igor@49 62 best_match = list(x[1] for x in sorted(
igor@47 63 (x for x in wordpairs_group if len(x[1]) == minimal_length),
igor@47 64 key=lambda x:x[0],
igor@47 65 reverse=True))[0]
igor@47 66
igor@51 67 return best_match
igor@51 68
igor@49 69 suggestions = self.dictionary_suggestions(best_match)
igor@49 70 if len(suggestions) == 1:
igor@49 71 return best_match
igor@49 72
igor@49 73 verb = False
igor@49 74 corrected_best_match = best_match
igor@49 75 if best_match[-2:] == 'et':
igor@49 76 word = best_match[:-1]+"n"
igor@49 77 sugg = self.dictionary_suggestions(word)
igor@49 78 if len(sugg) == 1:
igor@49 79 return word
igor@49 80 suggestions += sugg
igor@49 81 corrected_best_match = word
igor@49 82 corrected_best_match = best_match[:-2]
igor@49 83 verb = True
igor@49 84
igor@49 85 if best_match[-1] == 't':
igor@49 86 word = best_match[:-1]+"en"
igor@49 87 sugg = self.dictionary_suggestions(word)
igor@49 88 if len(sugg) == 1:
igor@49 89 return word
igor@49 90 suggestions += sugg
igor@49 91 corrected_best_match = best_match[:-1]
igor@49 92 verb = True
igor@49 93
igor@49 94 if corrected_best_match[0].lower() == corrected_best_match[0]:
igor@49 95 suggestions = [ x for x in suggestions
igor@49 96 if x[0].lower() == x[0] ]
igor@49 97
igor@49 98 if suggestions == []:
igor@49 99 return best_match+"_"
igor@49 100 return best_match+" "+(" ".join(
igor@49 101 sorted(
igor@49 102 suggestions,
igor@49 103 key = lambda x: f(x, corrected_best_match),
igor@49 104 reverse = True
igor@49 105 )
igor@49 106 )
igor@49 107 )
igor@49 108
igor@49 109 def dictionary_suggestions(self, word):
igor@49 110 return [
igor@49 111 x.decode('utf-8').rstrip('\n')
igor@49 112 for x
igor@49 113 in subprocess.Popen(
igor@49 114 ["de-variants", word],
igor@49 115 stdout=subprocess.PIPE
igor@49 116 ).stdout.readlines() ]
igor@49 117
igor@49 118
igor@37 119 parser = optparse.OptionParser()
igor@37 120
igor@37 121 parser.add_option(
igor@37 122 "-a", "--no-marks",
igor@55 123 help="don't add marks (and don't save marks added by user) [NOT IMPLEMENTED YET]",
igor@37 124 action="store_true",
igor@37 125 dest="no_marks")
igor@37 126
igor@37 127 parser.add_option(
igor@37 128 "-c", "--compressed",
igor@37 129 help="show compressed wordlist: one word per group",
igor@37 130 action="store_true",
igor@37 131 dest="compressed")
igor@37 132
igor@37 133 parser.add_option(
igor@37 134 "-k", "--known-words",
igor@37 135 help="put higher words that are similar to the known words (only for English)",
igor@37 136 action="store_true",
igor@37 137 dest="compressed")
igor@37 138
igor@37 139 parser.add_option(
igor@37 140 "-l", "--language",
igor@37 141 help="specify language of text",
igor@37 142 action="store",
igor@37 143 dest="language")
igor@37 144
igor@37 145 parser.add_option(
igor@54 146 "-f", "--allowed-words",
igor@54 147 help="file with list of allowed words (words that will be shown in the output)",
igor@54 148 action="store",
igor@54 149 dest="allowed_words")
igor@54 150
igor@54 151 parser.add_option(
igor@55 152 "-G", "--words-grouping",
igor@55 153 help="turn off word grouping",
igor@55 154 action="store_true",
igor@55 155 dest="no_words_grouping")
igor@55 156
igor@55 157 parser.add_option(
igor@54 158 "-X", "--function",
igor@38 159 help="filter through subsystem [INTERNAL]",
igor@38 160 action="store",
igor@38 161 dest="function")
igor@38 162
igor@38 163 parser.add_option(
igor@37 164 "-m", "--merge-tag",
igor@55 165 help="merge words tagged with specified tag into the main vocabulary [NOT IMPLEMENTED YET]",
igor@37 166 action="store",
igor@37 167 dest="merge_tag")
igor@37 168
igor@37 169 parser.add_option(
igor@37 170 "-M", "--merge-tagged",
igor@55 171 help="merge words tagged with ANY tag into the main vocabulary [NOT IMPLEMENTED YET]",
igor@37 172 action="store_true",
igor@37 173 dest="merge_tagged")
igor@37 174
igor@37 175 parser.add_option(
igor@37 176 "-n", "--non-interactive",
igor@37 177 help="non-interactive mode (don't run vi)",
igor@37 178 action="store_true",
igor@37 179 dest="non_interactive")
igor@37 180
igor@37 181 parser.add_option(
igor@37 182 "-N", "--no-filter",
igor@37 183 help="switch off known words filtering",
igor@37 184 action="store_true",
igor@37 185 dest="no_filter")
igor@37 186
igor@37 187 parser.add_option(
igor@37 188 "-p", "--pages",
igor@37 189 help="work with specified pages only (pages = start-stop/total )",
igor@37 190 action="store",
igor@37 191 dest="pages")
igor@37 192
igor@37 193 parser.add_option(
igor@48 194 "-d", "--delete-tag",
igor@48 195 help="delete subvocabulary of specified tag",
igor@37 196 action="store",
igor@48 197 dest="delete_tag")
igor@37 198
igor@37 199 parser.add_option(
igor@55 200 "-r", "--show-range",
igor@55 201 help="show only words specified number of words",
igor@55 202 action="store",
igor@55 203 dest="show_range")
igor@55 204
igor@55 205 parser.add_option(
igor@54 206 "-R", "--show-range-percentage",
igor@54 207 help="show only words that cover specified percentage of the text, skip the rest",
igor@54 208 action="store",
igor@54 209 dest="show_range_percentage")
igor@54 210
igor@54 211 parser.add_option(
igor@37 212 "-s", "--text-stats",
igor@37 213 help="show the text statistics (percentage of known words and so on) and exit",
igor@37 214 action="store_true",
igor@37 215 dest="text_stats")
igor@37 216
igor@37 217 parser.add_option(
igor@37 218 "-S", "--voc-stats",
igor@55 219 help="show your vocabulary statistics (number of words and word groups) [NOT IMPLEMENTED YET]",
igor@37 220 action="store_true",
igor@37 221 dest="voc_stats")
igor@37 222
igor@37 223 parser.add_option(
igor@37 224 "-t", "--tag",
igor@37 225 help="tag known words with tag",
igor@37 226 action="store",
igor@37 227 dest="tag")
igor@37 228
igor@37 229 parser.add_option(
igor@37 230 "-T", "--show-tags",
igor@37 231 help="tag known words with tag",
igor@37 232 action="store_true",
igor@37 233 dest="show_tags")
igor@37 234
igor@37 235 parser.add_option(
igor@63 236 "-v", "--vocabulary-filename",
igor@63 237 help="use specified file as a vocabulary",
igor@63 238 action="store",
igor@63 239 dest="vocabulary_filename")
igor@63 240
igor@63 241 parser.add_option(
igor@37 242 "-2", "--two-words",
igor@37 243 help="find 2 words' sequences",
igor@37 244 action="store_true",
igor@37 245 dest="two_words")
igor@37 246
igor@37 247 parser.add_option(
igor@37 248 "-3", "--three-words",
igor@37 249 help="find 3 words' sequences",
igor@37 250 action="store_true",
igor@37 251 dest="three_words")
igor@37 252
igor@38 253 def readlines_from_file(filename):
igor@38 254 res = []
igor@38 255 with codecs.open(filename, "r", "utf-8") as f:
igor@38 256 for line in f.readlines():
igor@38 257 res += [line]
igor@38 258 return res
igor@38 259
igor@54 260 def readlines_from_url(url):
igor@54 261 return [x.decode('utf-8') for x in
igor@54 262 subprocess.Popen(
igor@54 263 "lynx -dump '{url}' | perl -p -e 's@http://[a-zA-Z&_.:/0-9%?=,#+()\[\]~-]*@@'".format(url=url),
igor@54 264 shell = True,
igor@54 265 stdout = subprocess.PIPE,
igor@54 266 stderr = subprocess.STDOUT
igor@54 267 ).communicate()[0].split('\n')
igor@54 268 ]
igor@54 269
igor@38 270 def readlines_from_stdin():
igor@38 271 return codecs.getreader("utf-8")(sys.stdin).readlines()
igor@38 272
igor@38 273 def words_from_line(line):
igor@38 274 line = line.rstrip('\n')
igor@38 275 #return re.split('(?:\s|[*\r,.:#@()+=<>$;"?!|\[\]^%&~{}«»–])+', line)
igor@38 276 #return re.split('[^a-zA-ZäöëüßÄËÖÜß]+', line)
igor@44 277 return re.compile("(?!['_])(?:\W)+", flags=re.UNICODE).split(line)
igor@38 278
igor@44 279 def get_words(lines, group_by=[1]):
igor@38 280 """
igor@38 281 Returns hash of words in a file
igor@38 282 word => number
igor@38 283 """
igor@38 284 result = {}
igor@44 285 (a, b, c) = ("", "", "")
igor@38 286 for line in lines:
igor@38 287 words = words_from_line(line)
igor@38 288 for word in words:
igor@41 289 if re.match('[0-9]*$', word):
igor@41 290 continue
igor@38 291 result.setdefault(word, 0)
igor@38 292 result[word] += 1
igor@44 293 if 2 in group_by and a != "" and b != "":
igor@44 294 w = "%s_%s" % (a,b)
igor@44 295 result.setdefault(w, 0)
igor@44 296 result[w] += 1
igor@44 297 if 3 in group_by and not "" in [a,b,c]:
igor@44 298 w = "%s_%s_%s" % (a,b,c)
igor@44 299 result.setdefault(w, 0)
igor@44 300 result[w] += 1
igor@44 301 (a,b,c) = (b, c, word)
igor@44 302
igor@44 303 logging.debug(result)
igor@38 304 return result
igor@38 305
igor@54 306 def voc_filename():
igor@63 307 if 'vocabulary_filename' in config:
igor@63 308 return config['vocabulary_filename']
igor@54 309 return "%s/%s.txt"%(config['config_directory'], config['language'])
igor@54 310
igor@38 311 def load_vocabulary():
igor@54 312 return get_words(readlines_from_file(voc_filename()))
igor@38 313
igor@38 314 def notes_filenames():
igor@38 315 return ["%s/notes-%s.txt"%(config['config_directory'], config['language'])]
igor@38 316
igor@38 317 def load_notes(files):
igor@38 318 notes = {}
igor@38 319 for filename in files:
igor@39 320 with codecs.open(filename, "r", "utf-8") as f:
igor@38 321 for line in f.readlines():
igor@38 322 (word, note) = re.split('\s+', line.rstrip('\n'), maxsplit=1)
igor@38 323 notes.setdefault(word, {})
igor@38 324 notes[word][filename] = note
igor@38 325 return notes
igor@38 326
igor@39 327 def add_notes(lines, notes):
igor@39 328 notes_filename = notes_filenames()[0]
igor@39 329 result = []
igor@39 330 for line in lines:
igor@39 331 if line.startswith('#'):
igor@39 332 result += [line]
igor@39 333 else:
igor@39 334 match_object = re.search('^\s*\S+\s*(\S+)', line)
igor@39 335 if match_object:
igor@39 336 word = match_object.group(1)
igor@39 337 if word in notes:
igor@39 338 if notes_filename in notes[word]:
igor@39 339 line = line.rstrip('\n')
igor@39 340 line = "%-30s %s\n" % (line, notes[word][notes_filename])
igor@39 341 result += [line]
igor@39 342 else:
igor@39 343 result += [line]
igor@39 344 else:
igor@39 345 result += [line]
igor@39 346 return result
igor@39 347
igor@39 348 def remove_notes(lines, notes_group):
igor@39 349 notes_filename = notes_filenames()[0]
igor@39 350 notes = {}
igor@39 351 for k in notes_group.keys():
igor@39 352 if notes_filename in notes_group[k]:
igor@39 353 notes[k] = notes_group[k][notes_filename]
igor@39 354
igor@39 355 result = []
igor@39 356 for line in lines:
igor@39 357 line = line.rstrip('\n')
igor@39 358 match_object = re.match('(\s+)(\S+)(\s+)(\S+)(\s+)(.*)', line)
igor@39 359 if match_object:
igor@39 360 result.append("".join([
igor@39 361 match_object.group(1),
igor@39 362 match_object.group(2),
igor@39 363 match_object.group(3),
igor@39 364 match_object.group(4),
igor@39 365 "\n"
igor@39 366 ]))
igor@39 367 notes[match_object.group(4)] = match_object.group(6)
igor@39 368 else:
igor@39 369 result.append(line+"\n")
igor@39 370
igor@39 371 save_notes(notes_filename, notes)
igor@39 372 return result
igor@39 373
igor@39 374 def save_notes(filename, notes):
igor@39 375 lines = []
igor@39 376 saved_words = []
igor@39 377 with codecs.open(filename, "r", "utf-8") as f:
igor@39 378 for line in f.readlines():
igor@39 379 (word, note) = re.split('\s+', line.rstrip('\n'), maxsplit=1)
igor@39 380 if word in notes:
igor@39 381 line = "%-29s %s\n" % (word, notes[word])
igor@39 382 saved_words.append(word)
igor@39 383 lines.append(line)
igor@39 384 for word in [x for x in notes.keys() if not x in saved_words]:
igor@39 385 line = "%-29s %s\n" % (word, notes[word])
igor@39 386 lines.append(line)
igor@39 387
igor@39 388 with codecs.open(filename, "w", "utf-8") as f:
igor@39 389 for line in lines:
igor@39 390 f.write(line)
igor@39 391
igor@39 392
igor@38 393 def substract_dictionary(dict1, dict2):
igor@38 394 """
igor@38 395 returns dict1 - dict2
igor@38 396 """
igor@38 397 result = {}
igor@38 398 for (k,v) in dict1.items():
igor@38 399 if not k in dict2:
igor@38 400 result[k] = v
igor@38 401 return result
igor@38 402
igor@38 403 def dump_words(words, filename):
igor@38 404 with codecs.open(filename, "w+", "utf-8") as f:
igor@38 405 for word in words.keys():
igor@38 406 f.write(("%s\n"%word)*words[word])
igor@38 407
igor@38 408 def error_message(text):
igor@38 409 print text
igor@38 410
igor@40 411 def find_wordgroups_weights(word_pairs, normalizator):
igor@38 412 weight = {}
igor@40 413 for (num, word) in word_pairs:
igor@38 414 normalized = normalizator.normalize(word)
igor@38 415 weight.setdefault(normalized, 0)
igor@40 416 weight[normalized] += num
igor@38 417 return weight
igor@38 418
igor@38 419 def find_linked_words(notes):
igor@38 420 linked_words = {}
igor@38 421 for word in notes.keys():
igor@38 422 for note in notes[word].values():
igor@38 423 if "@" in note:
igor@38 424 result = re.search(r'\@(\S*)', note)
igor@38 425 if result:
igor@38 426 main_word = result.group(1)
igor@38 427 if main_word:
igor@38 428 linked_words[word] = main_word
igor@38 429 return linked_words
igor@38 430
igor@40 431 def compare_word_pairs(pair1, pair2, wgw, normalizator, linked_words):
igor@40 432 (num1, word1) = pair1
igor@40 433 (num2, word2) = pair2
igor@38 434
igor@38 435 normalized_word1 = normalizator.normalize(word1)
igor@38 436 normalized_word2 = normalizator.normalize(word2)
igor@38 437
igor@38 438 cmp_res = cmp(wgw[normalized_word1], wgw[normalized_word2])
igor@38 439 if cmp_res != 0:
igor@38 440 return cmp_res
igor@38 441 else:
igor@38 442 cmp_res = cmp(normalized_word1, normalized_word2)
igor@38 443 if cmp_res != 0:
igor@38 444 return cmp_res
igor@38 445 else:
igor@38 446 return cmp(int(num1), int(num2))
igor@38 447
igor@47 448
igor@48 449 def print_words_sorted(
igor@48 450 word_pairs,
igor@48 451 stats,
igor@48 452 normalizator,
igor@48 453 print_stats=True,
igor@48 454 stats_only=False,
igor@48 455 compressed_wordlist=False,
igor@48 456 show_range=0,
igor@48 457 show_range_percentage=0,
igor@48 458 ):
igor@54 459 result = []
igor@40 460 if stats_only:
igor@54 461 #codecs.getwriter("utf-8")(sys.stdout).write(
igor@54 462 result.append(
igor@43 463 " ".join([
igor@43 464 "%-10s" % x for x in [
igor@43 465 "LANG",
igor@43 466 "KNOWN%",
igor@43 467 "UNKNOWN%",
igor@43 468 "KNOWN",
igor@43 469 "TOTAL",
igor@43 470 "WPS",
igor@43 471 "UWPS*10"
igor@43 472 ]]) + "\n")
igor@54 473 result.append(
igor@43 474 " ".join([
igor@43 475 "%(language)-10s",
igor@43 476 "%(percentage)-10.2f",
igor@43 477 "%(percentage_unknown)-10.2f",
igor@43 478 "%(total_known)-11d"
igor@43 479 "%(total)-11d"
igor@43 480 "%(wps)-11d"
igor@43 481 "%(uwps)-11d"
igor@43 482 ]) % stats + "\n")
igor@54 483 return "".join(result)
igor@38 484
igor@40 485 if print_stats:
igor@54 486 result.append(
igor@43 487 "# %(language)s, %(percentage)-7.2f, <%(total_known)s/%(total)s>, <%(groups)s/%(words)s>\n" % stats)
igor@38 488
igor@40 489 level_lines = range(int(float(stats['percentage']))/5*5+5,95,5)+range(90,102)
igor@40 490 known = int(stats['total_known'])
igor@40 491 total = int(stats['total'])
igor@40 492 current_level = 0
igor@47 493 old_normalized_word = None
igor@47 494 words_of_this_group = []
igor@48 495 printed_words = 0
igor@40 496 for word_pair in word_pairs:
igor@47 497
igor@47 498 normalized_word = normalizator.normalize(word_pair[1])
igor@47 499 if old_normalized_word and old_normalized_word != normalized_word:
igor@47 500 if compressed_wordlist:
igor@49 501 compressed_word_pair = (
igor@49 502 sum(x[0] for x in words_of_this_group),
igor@49 503 normalizator.best_word_from_group(words_of_this_group)
igor@49 504 )
igor@54 505 result.append("%10s %s\n" % compressed_word_pair)
igor@48 506 printed_words += 1
igor@47 507 words_of_this_group = []
igor@47 508
igor@47 509 old_normalized_word = normalized_word
igor@47 510 words_of_this_group.append(word_pair)
igor@47 511
igor@47 512 if not compressed_wordlist:
igor@54 513 result.append("%10s %s\n" % word_pair)
igor@48 514 printed_words += 1
igor@47 515
igor@47 516
igor@40 517 known += word_pair[0]
igor@40 518 if 100.0*known/total >= level_lines[0]:
igor@40 519 current_level = level_lines[0]
igor@40 520 while 100.0*known/total > level_lines[0]:
igor@40 521 current_level = level_lines[0]
igor@40 522 level_lines = level_lines[1:]
igor@54 523 result.append("# %s\n" % current_level)
igor@38 524
igor@48 525 if show_range >0 and printed_words >= show_range:
igor@48 526 break
igor@48 527 if show_range_percentage >0 and 100.0*known/total >= show_range_percentage:
igor@48 528 break
igor@48 529
igor@54 530 return result
igor@39 531
igor@53 532 def parse_parts_description(parts_description):
igor@53 533 """
igor@53 534 Returns triad (start, stop, step)
igor@53 535 basing on parts_description string.
igor@53 536 from-to/step
igor@53 537 from+delta/step
igor@53 538 """
igor@53 539
igor@53 540 try:
igor@53 541 (a, step) = parts_description.split("/", 1)
igor@53 542 step = int(step)
igor@53 543 start = 0
igor@53 544 stop = 0
igor@53 545 if '-' in a:
igor@53 546 (start, stop) = a.split("-", 1)
igor@53 547 start = int(start)
igor@53 548 stop = int(stop)
igor@53 549 elif '+' in a:
igor@53 550 (start, stop) = a.split("+", 1)
igor@53 551 start = int(start)
igor@53 552 stop = int(stop)
igor@53 553 else:
igor@53 554 start = int(a)
igor@53 555 stop = start + 1
igor@53 556 return (start, stop, step)
igor@53 557
igor@53 558 except:
igor@54 559 raise ValueError("Parts description must be in format: num[[+-]num]/num; this [%s] is incorrect" % parts_description)
igor@53 560
igor@53 561
igor@53 562 def take_part(lines, part_description = None):
igor@55 563 if part_description == None or part_description == '':
igor@53 564 return lines
igor@53 565 (start, stop, step) = parse_parts_description(part_description)
igor@53 566 n = len(lines)
igor@53 567 part_size = (1.0*n) / step
igor@53 568 result = []
igor@53 569 for i in range(n):
igor@54 570 if i >= start * part_size and i <= stop * part_size:
igor@54 571 result += [lines[i]]
igor@53 572 return result
igor@53 573
igor@40 574 def filter_get_words_group_words_add_stat(args):
igor@40 575 vocabulary = load_vocabulary()
igor@40 576 notes = load_notes(notes_filenames())
igor@54 577
igor@54 578 if len(args) > 0:
igor@54 579 if 'http://' in args[0]:
igor@54 580 input_lines = readlines_from_url(args[0])
igor@54 581 else:
igor@54 582 input_lines = readlines_from_file(args[0])
igor@54 583 else:
igor@54 584 input_lines = readlines_from_stdin()
igor@54 585
igor@54 586 if len(input_lines) == 0:
igor@54 587 print >> sys.stderr, "Nothing to do, standard input is empty, exiting."
igor@54 588 sys.exit(1)
igor@54 589
igor@54 590 lines = take_part(input_lines, config.get('pages', ''))
igor@54 591
igor@54 592 (_, original_text_tempfile) = tempfile.mkstemp(prefix='new-word')
igor@54 593 with codecs.open(original_text_tempfile, "w", "utf-8") as f:
igor@54 594 f.write("".join(lines))
igor@54 595
igor@44 596 group_by = [1]
igor@48 597
igor@54 598 if 'two_words' in config:
igor@44 599 group_by.append(2)
igor@54 600 if 'three_words' in config:
igor@44 601 group_by.append(3)
igor@44 602 words = get_words(lines, group_by)
igor@43 603 stats_only = False
igor@54 604 if 'text_stats' in config:
igor@43 605 stats_only = True
igor@40 606
igor@47 607 compressed_wordlist = False
igor@54 608 if 'compressed' in config:
igor@47 609 compressed_wordlist = True
igor@47 610
igor@55 611 if 'show_range' in config:
igor@55 612 show_range = int(config['show_range'])
igor@48 613 else:
igor@48 614 show_range = 0
igor@54 615
igor@54 616 if 'show_range_percentage' in config:
igor@54 617 show_range_percentage = int(config['show_range_percentage'])
igor@48 618 else:
igor@48 619 show_range_percentage = 0
igor@48 620
igor@44 621
igor@40 622 stats = {}
igor@40 623 stats['total'] = sum(words[x] for x in words.keys())
igor@54 624 if not 'no_filter' in config:
igor@45 625 words = substract_dictionary(words, vocabulary)
igor@40 626
igor@40 627 stats['total_unknown'] = sum(words[x] for x in words.keys())
igor@40 628 stats['total_known'] = stats['total'] - stats['total_unknown']
igor@43 629 stats['percentage'] = 100.0*stats['total_known']/stats['total']
igor@43 630 stats['percentage_unknown'] = 100.0-100.0*stats['total_known']/stats['total']
igor@40 631 stats['groups'] = 0
igor@40 632 stats['words'] = len(words)
igor@43 633 stats['sentences'] = 0 #FIXME
igor@43 634 stats['wps'] = 0 #FIXME
igor@43 635 stats['uwps'] = 0 #FIXME
igor@40 636 stats['language'] = config['language']
igor@40 637
igor@40 638 linked_words = find_linked_words(notes)
igor@40 639 normalizator = Normalizator(config['language'], linked_words)
igor@40 640
igor@50 641 # filter words by allowed_words_filter
igor@54 642 if 'allowed_words' in config:
igor@54 643 allowed_words_filename = config['allowed_words']
igor@50 644 normalized_allowed_words = [
igor@50 645 normalizator.normalize(w.rstrip('\n'))
igor@50 646 for w in readlines_from_file(allowed_words_filename)
igor@50 647 ]
igor@50 648
igor@50 649 result = {}
igor@50 650 for w, wn in words.iteritems():
igor@50 651 if normalizator.normalize(w) in normalized_allowed_words:
igor@50 652 result[w] = wn
igor@50 653 words = result
igor@50 654
igor@44 655 words_with_freq = []
igor@40 656 for k in sorted(words.keys(), key=lambda k: words[k], reverse=True):
igor@44 657 words_with_freq.append((words[k], k))
igor@40 658
igor@44 659 wgw = find_wordgroups_weights(words_with_freq, normalizator)
igor@55 660 if not 'no_words_grouping' in config or not config['no_words_grouping']:
igor@45 661 words_with_freq = sorted(
igor@44 662 words_with_freq,
igor@40 663 cmp=lambda x,y:compare_word_pairs(x,y, wgw, normalizator, linked_words),
igor@40 664 reverse=True)
igor@40 665
igor@54 666 output = print_words_sorted(
igor@47 667 words_with_freq,
igor@47 668 stats,
igor@47 669 normalizator,
igor@47 670 stats_only=stats_only,
igor@48 671 compressed_wordlist=compressed_wordlist,
igor@48 672 show_range=show_range,
igor@48 673 show_range_percentage=show_range_percentage,
igor@47 674 )
igor@40 675
igor@54 676
igor@54 677 if ('non_interactive' in config or 'text_stats' in config):
igor@54 678 codecs.getwriter("utf-8")(sys.stdout).write("".join(output))
igor@54 679 else:
igor@54 680 (_, temp1) = tempfile.mkstemp(prefix='new-word')
igor@54 681 (_, temp2) = tempfile.mkstemp(prefix='new-word')
igor@54 682
igor@54 683 with codecs.open(temp1, "w", "utf-8") as f:
igor@54 684 f.write("".join(output))
igor@54 685 with codecs.open(temp2, "w", "utf-8") as f:
igor@54 686 f.write("".join(add_notes(output, notes)))
igor@54 687
igor@54 688 os.putenv('ORIGINAL_TEXT', original_text_tempfile)
igor@54 689 os.system((
igor@54 690 "vim"
igor@54 691 " -c 'setlocal spell spelllang={language}'"
igor@54 692 " -c 'set keywordprg={language}'"
igor@54 693 " -c 'set iskeyword=@,48-57,/,.,-,_,+,,,#,$,%,~,=,48-255'"
igor@54 694 " {filename}"
igor@54 695 " < /dev/tty > /dev/tty"
igor@54 696 ).format(language=config['language'], filename=temp2))
igor@54 697
igor@54 698 lines = remove_notes(readlines_from_file(temp2), notes)
igor@54 699
igor@54 700 # compare lines_before and lines_after and return deleted words
igor@54 701 lines_before = output
igor@54 702 lines_after = lines
igor@54 703 deleted_words = []
igor@54 704
igor@60 705 lines_after_set = set(lines_after)
igor@54 706 for line in lines_before:
igor@60 707 if line not in lines_after_set:
igor@54 708 line = line.strip()
igor@54 709 if ' ' in line:
igor@54 710 word = re.split('\s+', line, 1)[1]
igor@54 711 if ' ' in word:
igor@54 712 word = re.split('\s+', word, 1)[0]
igor@54 713 deleted_words.append(word)
igor@54 714
igor@54 715 with codecs.open(voc_filename(), "a", "utf-8") as f:
igor@54 716 f.write("\n".join(deleted_words + ['']))
igor@54 717
igor@54 718 os.unlink(temp1)
igor@54 719 os.unlink(temp2)
igor@54 720
igor@54 721 os.unlink(original_text_tempfile)
igor@54 722
igor@37 723 (options, args) = parser.parse_args()
igor@38 724 if options.language:
igor@38 725 config['language'] = options.language
igor@37 726
igor@54 727 if options.pages:
igor@54 728 config['pages'] = options.pages
igor@54 729 else:
igor@54 730 config['pages'] = ""
igor@54 731
igor@54 732 if options.allowed_words:
igor@54 733 config['allowed_words'] = options.allowed_words
igor@54 734
igor@55 735 if options.show_range:
igor@55 736 config['show_range'] = options.show_range
igor@55 737
igor@54 738 if options.show_range_percentage:
igor@54 739 config['show_range_percentage'] = options.show_range_percentage
igor@54 740
igor@54 741 if options.non_interactive:
igor@54 742 config['non_interactive'] = True
igor@54 743
igor@54 744 if options.text_stats:
igor@54 745 config['text_stats'] = True
igor@54 746
igor@54 747 if options.compressed:
igor@54 748 config['compressed'] = True
igor@54 749
igor@54 750 if options.no_filter:
igor@54 751 config['no_filter'] = True
igor@54 752
igor@54 753 if options.two_words:
igor@54 754 config['two_words'] = True
igor@54 755
igor@54 756 if options.three_words:
igor@54 757 config['three_words'] = True
igor@54 758
igor@55 759 if options.no_words_grouping:
igor@55 760 config['no_words_grouping'] = True
igor@37 761
igor@55 762 filter_get_words_group_words_add_stat(args)
igor@55 763
igor@55 764 #if options.function:
igor@55 765 # function_names = {
igor@55 766 # 'get_words_group_words_add_stat': ,
igor@55 767 # }
igor@55 768 # if options.function in function_names:
igor@55 769 # function_names[options.function](args)
igor@55 770 # else:
igor@55 771 # error_message("Unkown function %s.\nAvailable functions:\n%s" % (
igor@55 772 # options.function, "".join([" "+x for x in sorted(function_names.keys())])))
igor@55 773 # sys.exit(1)
igor@55 774 #
igor@37 775
igor@37 776
igor@37 777
igor@38 778 #os.system("vim")
igor@37 779