import_resztki.py 31.5 KB
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 91 92 93 94 95 96 97 98 99 100 101 102 103 104 105 106 107 108 109 110 111 112 113 114 115 116 117 118 119 120 121 122 123 124 125 126 127 128 129 130 131 132 133 134 135 136 137 138 139 140 141 142 143 144 145 146 147 148 149 150 151 152 153 154 155 156 157 158 159 160 161 162 163 164 165 166 167 168 169 170 171 172 173 174 175 176 177 178 179 180 181 182 183 184 185 186 187 188 189 190 191 192 193 194 195 196 197 198 199 200 201 202 203 204 205 206 207 208 209 210 211 212 213 214 215 216 217 218 219 220 221 222 223 224 225 226 227 228 229 230 231 232 233 234 235 236 237 238 239 240 241 242 243 244 245 246 247 248 249 250 251 252 253 254 255 256 257 258 259 260 261 262 263 264 265 266 267 268 269 270 271 272 273 274 275 276 277 278 279 280 281 282 283 284 285 286 287 288 289 290 291 292 293 294 295 296 297 298 299 300 301 302 303 304 305 306 307 308 309 310 311 312 313 314 315 316 317 318 319 320 321 322 323 324 325 326 327 328 329 330 331 332 333 334 335 336 337 338 339 340 341 342 343 344 345 346 347 348 349 350 351 352 353 354 355 356 357 358 359 360 361 362 363 364 365 366 367 368 369 370 371 372 373 374 375 376 377 378 379 380 381 382 383 384 385 386 387 388 389 390 391 392 393 394 395 396 397 398 399 400 401 402 403 404 405 406 407 408 409 410 411 412 413 414 415 416 417 418 419 420 421 422 423 424 425 426 427 428 429 430 431 432 433 434 435 436 437 438 439 440 441 442 443 444 445 446 447 448 449 450 451 452 453 454 455 456 457 458 459 460 461 462 463 464 465 466 467 468 469 470 471 472 473 474 475 476 477 478 479 480 481 482 483 484 485 486 487 488 489 490 491 492 493 494 495 496 497 498 499 500 501 502 503 504 505 506 507 508 509 510 511 512 513 514 515 516 517 518 519 520 521 522 523 524 525 526 527 528 529 530 531 532 533 534 535 536 537 538 539 540 541 542 543 544 545 546 547 548 549 550 551 552 553 554 555 556 557 558 559 560 561 562 563 564 565 566 567 568 569 570 571 572 573 574 575 576 577 578 579 580 581 582 583 584 585 586 587 588 589 590 591 592 593 594 595 596 597 598 599 600 601 602 603 604 605 606 607 608 609 610 611 612 613 614 615 616 617 618 619 620 621 622 623 624 625 626 627 628 629 630 631 632 633 634 635 636 637 638 639 640 641 642 643 644 645 646 647 648 649 650 651 652 653 654 655 656 657 658 659 660 661 662 663 664 665 666 667 668 669 670 671 672 673 674 675 676 677 678 679 680 681 682 683 684 685 686 687 688 689 690 691 692 693 694 695 696 697 698 699 700 701 702 703 704 705 706 707 708 709 710 711 712 713 714 715 716 717 718 719 720 721 722 723 724 725 726 727 728 729 730 731 732 733 734 735 736 737 738 739 740 741 742 743 744 745 746 747 748 749 750 751 752 753 754 755 756 757 758 759 760 761 762 763 764 765 766 767 768 769 770 771 772 773 774 775 776 777 778 779 780 781 782 783 784 785 786 787 788 789 790 791 792 793 794 795 796 797 798 799 800 801 802 803 804 805 806 807 808 809 810 811 812 813 814 815 816 817 818 819 820 821 822 823 824 825 826 827 828 829 830 831 832 833
#-*- coding:utf-8 -*-

import json

from django.core.management.base import BaseCommand

from common.util import suffixes, cut_end, debug, GroupDict
from dictionary.models import Pattern, Lexeme, InflectionCharacteristic, \
    Ending, LexicalClass, BaseFormLabel
from dictionary.pattern_blacklist import blacklist
from dictionary.management.commands.import_morfologik import join, join_many, \
    relevant_subst, relevant_adj, find_minimal_sets


class Command(BaseCommand):
    args = '<nazwa pliku wejściowego>'

    def handle(self, input_file, **options):
        import_resztki(open(input_file))


DEBUG = False

GENDERS = ('m1', 'm2', 'm3', 'm', 'f', 'n1', 'n2', 'p1', 'p2', 'p3')

#morf = Vocabulary.objects.get(id='Morfologik').owned_lexemes.all()
sgjp = Lexeme.objects.exclude(source='Morfologik')


def get_basic_endings(parts_of_speech, genders=None):
    ics = InflectionCharacteristic.objects.filter(
        part_of_speech__in=parts_of_speech)
    if genders:
        ics = ics.filter(symbol__in=genders)
    basic_form_labels = ics.values_list('basic_form_label',
        flat=True).distinct()
    return Ending.objects.filter(base_form_label__pk__in=basic_form_labels,
        pattern__type__lexical_class=lexical_class)


def expand_gender(gender):
    if gender == 'm':
        return ['m1', 'm2', 'm3']
    else:
        return [gender]


basic_form_endings_dict = {}
for lexical_class in LexicalClass.objects.all():
    parts_of_speech = lexical_class.partofspeech_set.all()
    if lexical_class.symbol == 'subst':
        for gender in GENDERS:
            basic_form_endings_dict[
                (lexical_class, gender)] = get_basic_endings(
                parts_of_speech, expand_gender(gender))
    else:
        basic_form_endings_dict[lexical_class] = get_basic_endings(
            parts_of_speech)


def tantum_a_posteriori(form_set, patterns):
    tantum = None
    for pattern, root in patterns:
        tantum_forms = {
            'sg': set(root + e for e in
                pattern.endings.filter(
                    base_form_label__symbol__startswith='sg')
                .values_list('string', flat=True)),
            'pl': set(root + e for e in
                pattern.endings.filter(
                    base_form_label__symbol__startswith='pl')
                .values_list('string', flat=True)),
        }
        for num in ('sg', 'pl'):
            if form_set.issubset(tantum_forms[num]):
                tantum = num
        if tantum:
            return tantum
    if not patterns:
        return 'sg'
    return None


def relevant(lexical_class, ending, **extra):
    if lexical_class.symbol == 'subst':
        return relevant_subst(ending, **extra)
    elif lexical_class.symbol == 'adj':
        return relevant_adj(ending)


from itertools import chain, combinations


def powerset(iterable):
    """powerset([1,2,3]) --> () (1,) (2,) (3,) (1,2) (1,3) (2,3) (1,2,3)"""
    s = list(iterable)
    return chain.from_iterable(
        combinations(s, r) for r in xrange(min(len(s) + 1, 5)))


sure_bfls_sg = tuple(
    BaseFormLabel.objects.filter(
        symbol__in=['sg:dat', 'sg:gen', 'sg:inst']).values_list('pk',
        flat=True))
sure_bfls_pl = tuple(
    BaseFormLabel.objects.filter(
        symbol__in=['pl:dat', 'pl:inst', 'pl:loc']).values_list('pk',
        flat=True))


def basic_form_endings(lexical_class, basic_form, form_set, **extra):
    if 'gender' in extra:
        key = (lexical_class, extra['gender'])
    else:
        key = lexical_class
    if lexical_class.symbol != 'subst':
        return basic_form_endings_dict[key].filter(
            string__in=suffixes(basic_form))
    else:
        # karkołomne, ale trochę przyśpiesza
        endings = basic_form_endings_dict[key]
        new_endings = Ending.objects.none()
        for suf in suffixes(basic_form):
            root = cut_end(basic_form, suf)
            n = len(root)
            ending_strings = tuple(
                form[n:] for form in form_set if form.startswith(root))
            endings_part = endings.filter(string=suf)
            pattern_pks = endings_part.values_list('pattern', flat=True)
            patterns = Pattern.objects.filter(pk__in=pattern_pks).extra(
                where=["(w_id = '0000' or not exists "
                       "(select id from zakonczenia where w_id = wzory.id "
                       "and zak not in %s and efobaz in %s) or not exists "
                       "(select id from zakonczenia where w_id = wzory.id "
                       "and zak not in %s and efobaz in %s))"],
                params=[ending_strings, sure_bfls_sg, ending_strings,
                    sure_bfls_pl])
            new_endings = new_endings | endings_part.filter(
                pattern__in=patterns)
        return new_endings


memoized_good_endings = {}


def good_ending_set_subst(pattern, root, tantum, gender):
    if (pattern, tantum, gender) in memoized_good_endings:
        good_endings = memoized_good_endings[(pattern, tantum, gender)]
        return set(root + e for e in good_endings)
    endings = pattern.endings
    if tantum:
        endings = endings.filter(base_form_label__symbol__startswith=tantum)
    if gender not in ('m1', 'p1'):
        endings = endings.exclude(base_form_label__symbol='pl:nom:mo')
    if gender[0] != 'p':
        for g in list(set('mfn') - set(gender[0])):
            endings = endings.exclude(
                base_form_label__symbol__startswith='pl:gen:' + g)
    if gender == 'p3':
        if pattern.type.symbol == 'f':
            endings = endings.exclude(base_form_label__symbol='pl:gen:m')
        if pattern.type.symbol == 'n':
            endings = endings.exclude(base_form_label__symbol='pl:gen:n')
    good_endings = list(endings.values_list('string', flat=True))
    memoized_good_endings[(pattern, tantum, gender)] = good_endings
    return set(root + e for e in good_endings)


def good_ending_set(lexical_class, pattern, root='', **extra):
    if lexical_class.symbol != 'subst':
        return pattern.ending_set(root)
    else:
        return good_ending_set_subst(pattern, root, **extra)


memoized_pattern_ics = {}


def bad_pattern_subst(pattern, gender, tantum):
    if (pattern, gender) in memoized_pattern_ics:
        return memoized_pattern_ics[(pattern, gender)]
    ics = expand_gender(gender)
    if gender == 'p1':
        ics.append('m1')
    if gender in ('p2', 'p3'):
        ics += ['m2', 'm3', 'f', 'n1', 'n2']
    if not pattern.lexemeinflectionpattern_set.filter(
            inflection_characteristic__symbol__in=ics).filter(
            lexeme__pk__in=sgjp):
        ret = True
    elif pattern.type.symbol in 'mn' and gender == 'f':
        ret = True
    elif pattern.type.symbol in 'fm' and gender[0] == 'n':
        ret = True
    else:
        ret = False
    memoized_pattern_ics[(pattern, gender)] = ret
    return ret


def find_patterns(lexical_class, basic_form, forms, **extra):
    #patterns = Pattern.objects.filter(type__lexical_class=lexical_class)
    # znaleźć wszystkie zawarte i zawierające wzory
    form_set = set(forms)
    ending_sets = {}
    included_patterns = set()
    including_patterns = set()
    matching_patterns = set()
    base_forms_changed = False
    for basic_ending in basic_form_endings(
            lexical_class, basic_form, form_set, **extra):
        pattern = basic_ending.pattern
        if lexical_class.symbol == 'subst' and bad_pattern_subst(pattern,
                **extra):
            #print 'odpadł:', pattern
            continue # olewamy komentarze że formy odrzucone przez charfle?
        root = basic_form[:len(basic_form) - len(basic_ending.string)]
        ending_sets[pattern] = good_ending_set(
            lexical_class, pattern, root, **extra)
        including = form_set.issubset(ending_sets[pattern])
        extra_base_forms = []
        bad_forms = set()
        for ending in pattern.endings.all():
            if relevant(lexical_class, ending, **extra):
                if root + ending.string not in form_set:
                    if DEBUG:
                        bfl = ending.base_form_label.symbol
                        #print pattern.name, root, ending.string, bfl
                    bad_forms.add(root + ending.string)
        if not bad_forms:
            if extra_base_forms:
                extra['base_forms'] += extra_base_forms
                base_forms_changed = True
            included_patterns.add((pattern, root))
            if including:
                matching_patterns.add((pattern, root))
        elif including:
            including_patterns.add(((pattern, root), tuple(bad_forms)))

    if base_forms_changed:
        #print extra['base_forms']
        return find_patterns(lexical_class, basic_form, forms, **extra)
        # nie wiem, czy to potrzebne, ale na wszelki wypadek
    included_patterns = list(included_patterns)
    including_patterns = list(including_patterns)
    matching_patterns = list(matching_patterns)
    if len(matching_patterns) > 0:
        if DEBUG:
            print u'dokładne wzory: %s' % join(matching_patterns)
        return 'match', matching_patterns, included_patterns, including_patterns
        # nic nie pasuje albo trzeba wybrać wiele wzorów
    if DEBUG and len(including_patterns) > 0:
        print u'zawierające: %s' % join(p for p, b_f in including_patterns)
    if DEBUG and len(included_patterns) > 0:
        print u'zawarte: %s' % join(included_patterns)
    return find_many_patterns(
        lexical_class, form_set, basic_form, included_patterns, ending_sets,
        **extra) + (included_patterns, including_patterns)


def find_many_patterns(lexical_class, form_set, basic_form, included_patterns,
                       ending_sets, **extra):
    necessary_patterns = set()
    missing_form = None
    for form in form_set:
        having = []
        for pattern, root in included_patterns:
            if form in ending_sets[pattern]:
                having.append((pattern, root))
        if len(having) == 1:
            necessary_patterns.add(having[0])
        if not having:
            missing_form = form
            break
    if missing_form:
        if DEBUG:
            print u"brak formy: %s" % missing_form
        return 'none', []
    covered_forms = set()
    for pattern, root in necessary_patterns:
        covered_forms |= ending_sets[pattern]
    if form_set.issubset(covered_forms):
        if DEBUG:
            print u"pokryte koniecznymi wzorami: %s" % join(necessary_patterns)
        return 'many', [list(necessary_patterns)]
    else:
        #for pattern, root in included_patterns:
        #  print pattern, ending_sets[pattern]
        minimal_sets = find_minimal_sets(
            form_set, covered_forms, necessary_patterns, included_patterns,
            ending_sets)
        return 'many', minimal_sets


def check_sgjp(lc_sym, entry, form_set, **extra):
    if lc_sym != 'adj':
        lexemes = Lexeme.objects.distinct().filter(
            entry=entry, part_of_speech__lexical_class__symbol=lc_sym)
    else:
        lexemes = Lexeme.objects.distinct().filter(
            entry=entry, part_of_speech__symbol__in=('adj', 'appas'))
    lexemes = lexemes.filter(pk__in=sgjp)
    matched_lexemes = []
    for lexeme in lexemes:
        if lc_sym == 'adj' and lexeme.refs_to.filter(type='nieadj'):
            continue
        if lc_sym == 'subst' and extra['tantum'] == 'sg':
            sgjp_forms = lexeme.all_forms(affixes=False, label_filter=r'sg:')
        elif lexeme.part_of_speech.symbol == 'appas':
            sgjp_forms = lexeme.all_forms(affixes=True)
        else:
            sgjp_forms = lexeme.all_forms(affixes=False)
        if sgjp_forms == form_set:
            matched_lexemes.append(lexeme)
            continue
        diff = sgjp_forms - form_set
        exceptions = []
        if lc_sym == 'subst':
            if lexeme.lexemeinflectionpattern_set.filter(
                    inflection_characteristic__symbol__in=(
                        'm1', 'p1')).exists():
                # depr
                exceptions = lexeme.all_forms(affixes=False,
                    label_filter=r'^pl:nom$')
        elif lc_sym == 'adj':
            # -o
            exceptions = lexeme.all_forms(affixes=False, label_filter=r'^0$')
        if form_set.issubset(sgjp_forms) and diff.issubset(exceptions):
            matched_lexemes.append(lexeme)
    if len(matched_lexemes) > 1:
        if lc_sym == 'subst' and entry.endswith(u'ość'):
            matched_lexemes_subst = [
                l for l in matched_lexemes if
                l.part_of_speech.symbol == 'subst']
            if matched_lexemes_subst:
                matched_lexemes = matched_lexemes_subst
        if len(matched_lexemes) > 1:
            debug(entry, u'niejednoznaczność dopasowanych leksemów')
    if len(matched_lexemes) > 0:
        return matched_lexemes[0]
    return False


def closest_lexeme_subst(entry, gender, patterns, included=None):
    lexemes = Lexeme.objects.filter(
        part_of_speech__lexical_class__symbol='subst')
    lexemes = lexemes.distinct()
    # ten sam rodzaj
    genders = expand_gender(gender)
    lexemes = lexemes.filter(
        lexemeinflectionpattern__inflection_characteristic__symbol__in=
        genders)
    if not included:
        # posiada wzór zawierający się w pasujących
        lexemes = lexemes.filter(lexemeinflectionpattern__pattern__in=patterns)
    else:
        #print patterns, included
        new_lexemes = Lexeme.objects.none()
        # posiada wszystkie wzory z któregoś zestawu
        for pattern_set in patterns:
            part = lexemes
            for pattern, root in pattern_set:
                part = part.filter(lexemeinflectionpattern__pattern=pattern)
            new_lexemes |= part
        lexemes = new_lexemes.distinct()
        # nie posiada wzorów niezawierających się w pasujących, dobra wielkość
    uppercase = entry[0].isupper()
    good_lexemes = []
    for lexeme in lexemes:
        if lexeme.entry[0].isupper() == uppercase:
            for lip in lexeme.lexemeinflectionpattern_set.all():
                if not included:
                    if lip.pattern not in patterns:
                        break
                else:
                    if lip.pattern not in included:
                        break
            else:
                good_lexemes.append(lexeme)
                # najdłuższe wspólne zakończenie
    best = (-1, None)
    for lexeme in good_lexemes:
        common_suffix = 0
        for char1, char2 in zip(entry[::-1], lexeme.entry[::-1]):
            if char1 == char2:
                common_suffix += 1
            else:
                break
        if common_suffix > best[0]:
            best = (common_suffix, lexeme)
    return best[1]


def blacklist_filter(patterns):
    return [(pattern, root) for (pattern, root) in patterns
        if pattern.name not in blacklist]


def filter_patterns(filter, action_name, type, patterns, included, including,
                    lexical_class, form_set, entry, **extra):
    old_patterns = patterns
    old_included = included
    bad_patterns = False
    if type == 'many':
        if any(pattern_set != filter(pattern_set) for pattern_set in patterns):
            included = filter(included)
            ending_sets = {}
            for pattern, root in included:
                ending_sets[pattern] = good_ending_set(
                    lexical_class, pattern, root, **extra)
            type, patterns = find_many_patterns(
                lexical_class, form_set, entry, included, ending_sets, **extra)
            if type != 'many':
                debug(entry, u'mnogie dopasowanie zepsute przez %s (%s)' %
                             (action_name, join_many(old_patterns)))
                type = 'many'
                patterns, included = old_patterns, old_included
                bad_patterns = True
    elif type == 'none':
        including_dict = dict(including)
        including = [(key, including_dict[key]) for key in
            filter(including_dict)]
    else: # type == 'match'
        patterns = filter(patterns)
        including_dict = dict(including)
        including = [(key, including_dict[key]) for key in
            filter(including_dict)]
        included = filter(included)
        if old_patterns and not patterns:
            ending_sets = {}
            for pattern, root in included:
                ending_sets[pattern] = good_ending_set(
                    lexical_class, pattern, root, **extra)
            type, patterns = find_many_patterns(
                lexical_class, form_set, entry, included, ending_sets, **extra)
            if type == 'none':
                debug(entry, u'znikły wzory przez %s (%s)' %
                             (action_name, join(old_patterns)))
                type = 'match'
                patterns = old_patterns
                bad_patterns = True
    return type, patterns, included, including, bad_patterns


def create_derived(pos, base_forms, forms, patterns):
    tab = {'ger': ('11', u'ie'), 'pact': ('3', u'cy'), 'ppas': ('10', u'y')}
    entries = GroupDict()
    for pattern, root in patterns:
        bfl = tab[pos][0]
        ending = pattern.endings.get(base_form_label__symbol=bfl)
        entry = root + ending.string + tab[pos][1]
        entries.add(entry, pattern.name)
    output = []
    for entry, patterns in entries.iteritems():
        if entry in forms:
            output.append((pos, entry, patterns))
    return output


def get_sgjp(lexeme):
    return {'source': 'sgjp', 'id': lexeme.pk, 'entry': lexeme.entry}


def create_lexeme(entry, part_of_speech, status, comment):
    return {
        'source': 'morfologik',
        'entry': entry,
        'part_of_speech': part_of_speech,
        'status': status,
        'comment': comment,
    }


def create_lip(pattern, root, i, ic, part_of_speech):
    output = {
        'pattern': pattern if isinstance(pattern, basestring) else pattern.name,
        'ind': i,
        'ic': (ic, part_of_speech),
    }
    if root:
        output['root'] = {'type': 'string', 'root': root}
    else:
        output['root'] = {'type': 'compute'}
    return output


alternative_gender = {
    'p1': 'p3',
    'p3': 'p1',
    'm1': 'm2/m3',
    'm': 'm1',
}

alternative_gender2 = {
    'p1': 'p3',
    'p3': 'p1',
    'm1': 'm',
    'm': 'm1',
}


def lexeme_creation(lc_sym, entry, ic, forms, type, patterns, fitting,
                    bad_patterns, included, other_result, tantum=None,
                    gender=None, negated=None, base_forms=None, derived=None):
    status = 'desc' if type != 'none' else 'cand'
    comments = [u'Z importu resztek']
    copy_lips = False
    if lc_sym == 'subst':
        part_of_speech = 'subst' # co z osc i skrs?
        if ic in ('m2', 'm3'):
            sure = False
            if type != 'none' and len(fitting) == 1:
                for pattern, root in patterns:
                    for e in patterns[0][0].endings.filter(
                            base_form_label__symbol='sg:gen'):
                        if not e.string.endswith('u'):
                            break
                    else:
                        continue
                    break
                else: # wszystkie sg:gen kończą się na 'u'
                    ic = 'm3'
                    sure = True
                for pattern, root in patterns:
                    if pattern.type.symbol == 'f':
                        ic = 'm2'
                        sure = True
                        break
            if not sure:
                status = 'cand'
        if tantum is None and ic == 'm1':
            for pattern, root in patterns:
                nmo_endings = pattern.endings.filter(
                    base_form_label__symbol='pl:nom')
                for e in nmo_endings:
                    nmo_form = root + e.string
                    if nmo_form not in forms:
                        comments.append(u'Dodano formę depr')
                        break
                else:
                    continue
                break
        if ic == 'p1':
            for pattern, root in patterns:
                nmo_endings = pattern.endings.filter(
                    base_form_label__symbol='pl:nom')
                other_endings = pattern.endings.exclude(
                    base_form_label__symbol='pl:nom')
                other_strings = other_endings.values_list('string', flat=True)
                nmo_strings = [e.string for e in nmo_endings if
                    e.string not in other_strings]
                nmo_forms = set(root + s for s in nmo_strings)
                if nmo_forms & set(forms):
                    comments.append(
                        u'Usunięto formę depr: %s' % ', '.join(list(nmo_forms)))
                    break
        if tantum == 'sg' and type != 'none':
            if type == 'match':
                search_patterns = [pattern for pattern, root in fitting]
                l = closest_lexeme_subst(entry, gender, search_patterns)
            else:
                included_patterns = [pattern for pattern, root in included]
                l = closest_lexeme_subst(entry, gender, fitting,
                    included_patterns)
            if l:
                copy_lips = l.lexemeinflectionpattern_set.all()
                #print l
                comments.append(u'Automatycznie rozszerzone singulare tantum')
            else:
                if type == 'match':
                    p = join(fitting)
                else:
                    p = join_many(fitting)
                debug(entry,
                    u'nie ma pasujących leksemów dla rozszerzenia sgtant '
                    u'dla wzorów %s' % p)
                comments.append(u'Nie udało się rozszerzyć singulare tantum')
                #status = 'cand'
                # dodać kwalifikator [po imporcie jednak]
    if bad_patterns:
        comments.append(u'Wzory z czarnej listy!')
        status = 'cand'
    if len(fitting) > 1 or (type == 'none' and fitting):
        status = 'cand'
        if type == 'none':
            comments.append(u'Zawierające wzory:')
            for (pattern, root), bad_forms in fitting:
                comments.append('%s: %s' % (pattern.name, ', '.join(bad_forms)))
        elif type != 'many':
            comments.append(u'Pasujące wzory: %s' % join(fitting))
        else:
            comments.append(u'Pasujące zestawy wzorów: %s' % join_many(fitting))
    if other_result:
        status = 'cand'
        type2, patterns2, included2, including2 = other_result
        comments.append(u'Alternatywny rodzaj: %s' % alternative_gender[gender])
        if type2 == 'match':
            comments.append(u'Pasujące wzory: %s' % join(patterns2))
        elif type2 == 'many':
            comments.append(
                u'Pasujące zestawy wzorów: %s' % join_many(patterns2))
            # hm?
    if ic is None and type != 'none':
        comments.append(u'Dopasowane wzory: %s' % join(patterns))
        # zbieramy wzory do porównania [fuj, copypasta!]
    if len(fitting) > 1:
        if type == 'none':
            all_patterns = set(p for p, b_f in fitting)
        elif type != 'many':
            all_patterns = set(fitting)
        else:
            all_patterns = set()
            for pattern_set in fitting:
                all_patterns |= set(pattern_set)
    if other_result and len(patterns2) > 1:
        if type2 != 'many':
            other_patterns = set(patterns2)
        else:
            other_patterns = set()
            for pattern_set in patterns2:
                other_patterns |= set(pattern_set)
    comment = '\n'.join(comments)
    output = {
        'lexeme': create_lexeme(entry, part_of_speech, status, comment)
    }
    lips = []
    if ic is not None:
        if not copy_lips:
            for i, (pattern, root) in enumerate(patterns):
                lips.append(
                    create_lip(pattern, root, i + 1, ic, part_of_speech))
        else:
            for lip in copy_lips:
                ic = lip.inflection_characteristic.symbol
                lips.append(
                    create_lip(lip.pattern, None, lip.index, ic,
                        part_of_speech))
    output['lips'] = lips
    if lc_sym == 'adj' and negated:
        output['negated'] = True
    if lc_sym == 'v':
        derived_data = []
        for pos in derived:
            # wypadałoby informować, jeśli wyszło puste... (?)
            derived_data += create_derived(pos, base_forms, forms, patterns)
        output['derived'] = derived_data
    return output


def process_forms(entry, forms, lc_sym, **extra):
    lexical_class = LexicalClass.objects.get(symbol=lc_sym)
    other_result = None
    form_set = set(forms)
    check = check_sgjp(lc_sym, entry, form_set, **extra)
    if check and not DEBUG:
        # dopisz leksem do słownika
        data = {'lexeme': get_sgjp(check)}
        # TODO negacja przymiotnika
        print_data(data)
    else:
        if lc_sym == 'subst':
            ic = extra['gender']
        extra2 = dict(extra)
        # jeśli rzeczownik męski lub pltant, to puszczamy więcej razy
        if lc_sym == 'subst' and extra['gender'] in 'pm':
            if extra['gender'] == 'm':
                extra2['gender'] = 'm1'
                type1, patterns1, included1, including1 = find_patterns(
                    lexical_class, entry, forms, **extra2)
                extra2['gender'] = 'm'
                type2, patterns2, included2, including2 = find_patterns(
                    lexical_class, entry, forms, **extra)
            elif extra['gender'] == 'p':
                extra2['gender'] = 'p1'
                type1, patterns1, included1, including1 = find_patterns(
                    lexical_class, entry, forms, **extra2)
                extra2['gender'] = 'p3'
                type2, patterns2, included2, including2 = find_patterns(
                    lexical_class, entry, forms, **extra2)
            if type1 != 'none' and type2 == 'none':
                type, patterns, included, including = type1, patterns1, included1, including1
                if extra['gender'] == 'm':
                    ic = 'm1'
                else:
                    ic = 'p1'
            elif type1 == 'none' and type2 != 'none':
                type, patterns, included, including = type2, patterns2, included2, including2
                if extra['gender'] == 'm':
                    ic = 'm'
                else:
                    ic = 'p3'
            elif type1 == type2 == 'none':
                type = 'none'
                patterns = []
                included = list(set(included1) | set(included2))
                including = list(set(including1) | set(including2))
                # chyba warto coś ustawić
                if extra['gender'] == 'm':
                    ic = 'm'
                else:
                    ic = 'p3'
            else: # z obu coś wyszło
                type, patterns, included, including = type1, patterns1, included1, including1
                if extra['gender'] == 'm':
                    ic = 'm'
                else:
                    ic = 'p1'
                other_result = (type2, patterns2, included2, including2)
                if DEBUG:
                    print u"dwie możliwości na rodzaj"
        else:
            type, patterns, included, including = find_patterns(
                lexical_class, entry, forms, **extra)
        if type == 'none':
            if lc_sym == 'subst' and not extra['tantum']:
                extra['tantum'] = tantum_a_posteriori(
                    form_set, [p for p, b_f in including])
                if extra['tantum']:
                    if extra['tantum'] == 'pl':
                        extra['gender'] = 'p'
                    return process_forms(entry, forms, lc_sym, **extra)

        if lc_sym == 'subst':
            extra2['gender'] = ic
        type, patterns, included, including, bad_patterns = filter_patterns(
            blacklist_filter, u'czarną listę', type, patterns, included,
            including,
            lexical_class, form_set, entry, **extra2)
        if bad_patterns and other_result:
            type, patterns, included, including = other_result
            ic = extra2['gender'] = alternative_gender2[ic]
            type, patterns, included, including, bad_patterns = filter_patterns(
                blacklist_filter, u'czarną listę', type, patterns, included,
                including,
                lexical_class, form_set, entry, **extra2)
            other_result = None
        elif other_result:
            type2, patterns2, included2, including2 = other_result
            new_other_result = filter_patterns(
                blacklist_filter, u'czarną listę', type2, patterns2, included2,
                including2, lexical_class, form_set, entry, **extra2)
            if not new_other_result[4]:
                other_result = new_other_result[:4]
            else:
                other_result = None

        # wzory się już nie zmienią od tego miejsca
        if type == 'many':
            # albo patterns[0]...
            all_patterns = [p for pattern_set in patterns for p in pattern_set]
        else:
            all_patterns = patterns
        if type != 'none':
            # brzydko...
            if lexical_class.symbol == 'subst':
                extra['ic'] = ic
                del extra['ic']
        else:
            ic = None
        if lc_sym == 'subst':
            # poprawka dla m2/m3
            if ic in ('m2', 'm3') and patterns and not bad_patterns:
                new_ic = ''
                for pattern, root in all_patterns:
                    for ic2 in ('m2', 'm3'):
                        # jeśli wszystkie użycia tego wzoru są przy ic2
                        if not pattern.lexemeinflectionpattern_set.exclude(
                                inflection_characteristic__symbol=ic2).filter(
                                lexeme__pk__in=sgjp).exists():
                            if new_ic == '':
                                new_ic = ic2
                            elif new_ic != ic2:
                                new_ic = None
                if new_ic:
                    ic = new_ic

        if type == 'none':
            debug(entry, u'zawiera się w %s' % join(p for p, b_f in including))
            chosen = []
            fitting = including
        elif type == 'match':
            patterns.sort(key=lambda p: p[0].name)
            fitting = patterns
            chosen = patterns[:1]
        elif type == 'many':
            chosen = patterns[0]
            if DEBUG:
                print u'zestawy wielu wzorów: %s' % join_many(patterns)
            fitting = patterns
        if not DEBUG:
            data = lexeme_creation(
                lc_sym, entry, ic, forms, type, chosen, fitting, bad_patterns,
                included,
                other_result, **extra2)
            print_data(data)


def get_pos_ndm(tag):
    if tag[0] == 'adv':
        return 'adv' if tag[-1] != 'comp' else 'advcom'
    elif tag[0] == 'xxx':
        return 'burk'
    else:
        return tag[0]


def print_data(data):
    print json.dumps(data)


def import_resztki(input_file):
    for line in input_file:
        entry, rest = line.decode('utf-8').strip().replace("'", u'’').split(':')
        pos, ic, rest = rest.split(';')
        if ic in ('n2/m3', 'n2/m3/m2', 'm3/m2'):
            ic = 'm3'
        elif ic == 'p3/p2':
            ic = 'p3'
        elif ic in ('f/m3', 'f/n2'):
            ic = 'f'
        forms = rest.split(',')
        if entry not in forms:
            forms = [entry] + forms
        if ic[0] == 'p':
            tantum = 'pl'
        else:
            tantum = None
        if entry.endswith(u'stwo') and ic == 'p1':
            # zbędne formy mnogie
            root = entry[:-1]
            for suffix in ('', 'ami', 'ach', 'om'):
                forms.remove(root + suffix)
        process_forms(entry, forms, pos, gender=ic, tantum=tantum)