extra_crs.py
7.43 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
#-*- coding:utf-8 -*-
import sys
from django.core.management.base import BaseCommand, CommandError
from common.util import no_history
from dictionary.models import Lexeme, Vocabulary, CrossReference
# UWAGA: zepsute! (po rewolucji w odsyłaczach)
# Nie wiem, czy warto to poprawiać...
class Command(BaseCommand):
args = '<type> <input file>'
help = 'Adds extra cross-references'
def handle(self, type, input_file, **options):
return
add_crs(type, input_file)
def debug(entry, text):
print>>sys.stderr, (u'%s: %s' % (entry, text)).encode('utf-8')
def connect_real(from_l, to_l, cr_type):
created1 = created2 = False
if from_l.source == 'Morfologik':
cr1, created1 = CrossReference.objects.get_or_create(
from_lexeme=from_l, to_lexeme=to_l, type=cr_type[0])
if to_l.source == 'Morfologik':
cr2, created2 = CrossReference.objects.get_or_create(
from_lexeme=to_l, to_lexeme=from_l, type=cr_type[1])
if created1 or created2:
debug(from_l.entry, u'Powiązano z %s' % to_l.entry)
def connect_dummy(from_l, to_l, cr_type):
created1 = not CrossReference.objects.filter(
from_lexeme=from_l, to_lexeme=to_l, type=cr_type[0])
created2 = not CrossReference.objects.filter(
from_lexeme=to_l, to_lexeme=from_l, type=cr_type[1])
if created1 or created2:
debug(from_l.entry, u'Powiązanoby z %s' % to_l.entry)
if 'Morfologik' not in (from_l.source, to_l.source):
debug(from_l.entry, u'Powiązanoby leksemy spoza Morfologika!')
connect = connect_dummy
def make_detail(qs, desc, morf, details):
if qs.count() > 1:
morf_count = qs.filter(pk__in=morf).count()
details.append(
u'%s: %s, w tym z Morfologika: %s' % (desc, qs.count(), morf_count))
def which_lacks(qs1, desc1, qs2, desc2):
if qs2.count() > 0:
return desc2
elif qs1.count() > 0:
return desc1
else:
return u'obu'
# dużo copypasty, ale chyba nie warto refaktoryzować
def add_crs(type, path):
no_history()
morfologik = Vocabulary.objects.get(id='Morfologik')
morf = morfologik.owned_lexemes_pk()
lexemes = Lexeme.objects.filter(
lexemeassociation__vocabulary__id='Morfologik')
adv = lexemes.filter(part_of_speech__symbol__in=('adv', 'advndm'))
adj = lexemes.filter(part_of_speech__symbol='adj')
advcom = lexemes.filter(part_of_speech__symbol='advcom')
with open(path) as file:
if type == 'advnie':
cr_type = ('nieadj', 'adjnie')
for line in file:
advneg_e, base_e = line.strip().decode('utf-8').split()
advnegs = adv.filter(entry=advneg_e)
if adv.filter(entry=base_e):
advs = adv.filter(entry=base_e)
if advnegs.count() > 1 or advs.count() > 1:
details = []
make_detail(advnegs, u'zanegowane', morf, details)
make_detail(advs, u'niezanegowane', morf, details)
debug(advneg_e, u'Niejednoznaczność: %s' % '; '.join(details))
elif advnegs.count() == 0 or advs.count() == 0:
lack = which_lacks(
advnegs, u'zanegowanego', advs, u'niezanegowanego')
debug(advneg_e, u'Brak %s' % lack)
else:
connect(advnegs[0], advs[0], cr_type)
elif adj.filter(entry=base_e):
# najpierw trzeba odpalić advadj!
adjs = adj.filter(
entry=base_e, refs_to__type='adjadv',
refs_to__to_lexeme__deleted=False)
adjs = adjs | adj.filter(
entry=base_e, refs_from__type='advadj',
refs_from__from_lexeme__deleted=False)
adjs = adjs.distinct()
if advnegs.count() > 1 or adjs.count() > 1:
details = []
make_detail(advnegs, u'zanegowane', morf, details)
make_detail(adjs, u'przymiotniki', morf, details)
debug(advneg_e, u'Niejednoznaczność: %s' % '; '.join(details))
elif advnegs.count() == 0 or adjs.count() == 0:
lack = which_lacks(
advnegs, u'zanegowanego', adjs, u'przymiotnika')
debug(advneg_e, u'Brak %s' % lack)
else:
advs = [cr.to_lexeme.pk for cr
in adjs[0].refs_to.filter(type='adjadv')]
advs += [cr.from_lexeme.pk for cr
in adjs[0].refs_from.filter(type='advadj')]
advs = adv.filter(pk__in=advs).distinct()
if len(advs) > 1:
details = []
make_detail(advs, u'niezanegowane', morf, details)
debug(advneg_e, u'Niejednoznaczność: %s' % '; '.join(details))
elif len(advs) == 0:
debug(advneg_e, u'Brak niezanegowanego')
else:
connect(advnegs[0], advs[0], cr_type)
else:
debug(advneg_e, u'Brak drugiego leksemu [przymiotnik lub przysłówek]')
elif type == 'advadj':
cr_type = ('advadj', 'adjadv')
for line in file:
adv_e, adj_e = line.strip().decode('utf-8').split()
advs = adv.filter(entry=adv_e, part_of_speech__symbol='adv')
adjs = adj.filter(entry=adj_e, patterns__name__contains='').distinct()
if advs.count() > 1 or adjs.count() > 1:
details = []
make_detail(advs, u'przysłówki', morf, details)
make_detail(adjs, u'przymiotniki', morf, details)
debug(adv_e, u'Niejednoznaczność: %s' % '; '.join(details))
elif advs.count() == 0 or adjs.count() == 0:
lack = which_lacks(advs, u'przysłówka', adjs, u'przymiotnika')
debug(adv_e, u'Brak %s' % lack)
else:
connect(advs[0], adjs[0], cr_type)
elif type == 'advcom':
cr_type = ('advcom', 'comadv')
for line in file:
com_e, adv_e = line.strip().decode('utf-8').split()
advs = adv.filter(entry=adv_e)
coms = advcom.filter(entry=com_e)
if advs.count() > 1 or coms.count() > 1:
details = []
make_detail(advs, u'równe', morf, details)
make_detail(coms, u'wyższe', morf, details)
debug(adv_e, u'Niejednoznaczność: %s' % '; '.join(details))
elif advs.count() == 0 or coms.count() == 0:
lack = which_lacks(advs, u'równego', coms, u'wyższego')
debug(adv_e, u'Brak %s' % lack)
else:
connect(advs[0], coms[0], cr_type)
elif type == 'adjadvc': # do puszczenia na koniec
cr_type = ('adjadvc', 'advcadj')
# uch!
advs = Lexeme.objects.filter(
refs_to__type='advadj', refs_to__to_lexeme__deleted=False)
advs = advs | Lexeme.objects.filter(
refs_from__type='adjadv', refs_from__from_lexeme__deleted=False)
advs_with_com = advs.filter(
refs_to__type='advcom', refs_to__to_lexeme__deleted=False)
advs_with_com = advs_with_com | advs.filter(
refs_from__type='comadv', refs_from__from_lexeme__deleted=False)
advs = advs_with_com.distinct()
for adv in advs:
adjs = Lexeme.objects.filter(
refs_to__type='adjadv', refs_to__to_lexeme=adv,
refs_to__to_lexeme__deleted=False)
adjs = adjs | Lexeme.objects.filter(
refs_from__type='advadj', refs_from__from_lexeme=adv)
adjs = adjs.distinct()
advcoms = Lexeme.objects.filter(
refs_to__type='comadv', refs_to__to_lexeme=adv,
refs_to__to_lexeme__deleted=False)
advcoms = advcoms | Lexeme.objects.filter(
refs_from__type='advcom', refs_from__from_lexeme=adv)
for adj in adjs:
for advcom in advcoms:
if not adj.refs_to.filter(to_lexeme=advcom, type='adjadvc'):
connect(adj, advcom, cr_type)