export.py
6.75 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.db import connection
from common.util import debug, flatten
from dictionary.models import CrossReferenceType, ClassificationValue, LexemeAttributeValue
ADJPREDYKATYWNE = [
u'ciekaw',
u'godzien',
u'gotów',
u'łaskaw',
u'świadom',
u'winien',
u'zdrów',
# wątpliwe:
u'dłużen',
u'miłościw',
u'praw',
u'wesół',
u'żyw',
]
REFL_TRANSLATION = {
u'—': 'nonrefl',
u'się': 'refl',
u'sobie': 'refl',
u'się/sobie': 'refl',
u'(się)': 'refl.nonrefl',
u'(sobie)': 'refl.nonrefl',
}
def qualifier_clause(q_id):
return '''not exists (
select * from kwalifikatory_leksemow where lexeme_id = l.id and
qualifier_id = %(q)d) and not exists (
select * from kwalifikatory_odmieniasiow where qualifier_id = %(q)d and
lexemeinflectionpattern_id = o.id) and not exists (
select * from kwalifikatory_zakonczen where qualifier_id = %(q)d and
ending_id = z.id) and ''' % {'q': q_id}
def magic_qualifier_clause():
return '''and not (tag like %s and exists (
select kw.id
from kwalifikatory kw
join kwalifikatory_leksemow kwl on kw.id = kwl.qualifier_id
where kwl.lexeme_id = l.id and kw.id = %s)) '''
def export_lexemes(data=None, output_file=None):
if not data:
data = {
'vocabs': ['PoliMorf'],
'antivocabs': [],
'variant': 'Morfeusz',
'excluding_qualifiers': [],
'magic_qualifiers': [],
'refl': False,
'commonness': False,
}
if output_file is None:
output_file = sys.stdout
vocabs_placeholders = ', '.join('%s' for v in data['vocabs'])
if data['antivocabs']:
antivocabs_placeholders = ', '.join('%s' for v in data['antivocabs'])
antivocabs_clause = '''not exists (
select * from leksemy_w_slownikach ls2 where ls2.l_id = l.id
and ls2.slownik in (%s)) and''' % antivocabs_placeholders
else:
antivocabs_clause = ''
qualifier_clauses = ''.join(
qualifier_clause(q_id) for q_id in data['excluding_qualifiers'])
magic_qualifier_clauses = ''.join(
magic_qualifier_clause() for pattern, q_id in data['magic_qualifiers'])
crtypes = ['comadv', 'comadj', 'gerver', 'pactver', 'ppasver']
crtype_ids = CrossReferenceType.objects.filter(
symbol__in=crtypes).values_list('pk', flat=True)
cv_ids = ClassificationValue.objects.filter(
classification__name=u'pospolitość').values_list('pk', flat=True)
refls = dict(LexemeAttributeValue.objects.filter(
attribute__name=u'zwrotność').values_list('pk', 'value'))
cursor = connection.cursor()
query = """
select distinct haslo, prefiks||rdzen||zak||sufiks, l.pos, ch.charfl, tag,
l.id as leksem_id, refl.attribute_value_id %(clas_field)s
from leksemy l
join leksemy_w_slownikach ls on (ls.l_id = l.id)
left outer join dictionary_lexemeav refl
on (l.id = refl.lexeme_id and %(refl)s)
join odmieniasie o on (o.l_id = l.id)
join charfle ch on ch.id = o.charfl
join wzory w on (o.w_id = w.id)
join szablony_tabel s on (w.typ=s.wtyp and o.charfl=s.charfl)
join klatki k on k.st_id = s.id
join zakonczenia z on (o.w_id=z.w_id and k.efobaz=z.efobaz)
join efobazy e on (e.id = k.efobaz)
%(clas_join)s
where ls.slownik in (%(vocabs)s) and %(antivocabs)s %(x_qual)s
l.status<>'cand' and wariant=%%s and
l.pos in ('v','subst','osc','adj','adv', 'num','advndm','burk','comp',
'conj','interj','prep','qub','ppron','pred') and
l.usuniety = false %(magic)s
--and haslo < 'b'
union all
-- wymagające gniazdowania przy hasłowaniu: adjcom, advcom, derywaty:
select distinct g.haslo as haslo, prefiks||rdzen||zak||sufiks, l.pos,
ch.charfl, tag, l.id as leksem_id, refl.attribute_value_id
%(clas_field)s
from leksemy l
join leksemy_w_slownikach ls on (ls.l_id = l.id)
join odsylacze on l.id=l_id_od
join leksemy g on (l_id_do=g.id and g.usuniety = false)
left outer join dictionary_lexemeav refl
on (g.id = refl.lexeme_id and %(refl)s)
join odmieniasie o on l.id=o.l_id
join charfle ch on ch.id = o.charfl
join wzory w on (o.w_id = w.id)
join szablony_tabel s on (w.typ=s.wtyp and o.charfl=s.charfl)
join klatki k on k.st_id = s.id
join zakonczenia z on (o.w_id=z.w_id and k.efobaz=z.efobaz)
%(clas_join)s
where ls.slownik in (%(vocabs)s) and %(antivocabs)s %(x_qual)s
typods_id in (%(crtype_ids)s) and
l.pos in ('adjcom','advcom','ger','pact','ppas','appas') and
wariant=%%s and l.status<>'cand' and l.usuniety = false
--and g.haslo < 'b'
order by haslo, leksem_id
""" % {
'vocabs': vocabs_placeholders,
'antivocabs': antivocabs_clause,
'x_qual': qualifier_clauses,
'magic': magic_qualifier_clauses,
'crtype_ids': ', '.join(str(pk) for pk in crtype_ids), # brzydko, oj tam
'clas_field': ', classification_value_id' if data['commonness'] else '',
'clas_join':
'left outer join dictionary_lexemecv wkl '
'on (wkl.lexeme_id=l.id and wkl.classification_value_id in (%s))'
% ', '.join(str(pk) for pk in cv_ids) if data['commonness'] else '',
'refl':
'refl.attribute_value_id in (%s)'
% ', '.join(str(pk) for pk in refls),
}
params_part = (list(data['vocabs']) + list(data['antivocabs']) +
[data['variant']])
params = params_part + flatten(data['magic_qualifiers']) + params_part
cursor.execute(query, params)
refl = data['refl']
cv_table = dict(ClassificationValue.objects.values_list('id', 'label'))
for row in cursor:
if data['commonness']:
entry, form, pos, _ic, tag, _id, refl_id, cv_id = row
else:
entry, form, pos, _ic, tag, _id, refl_id = row
form = form.lstrip('+') # odmienne postfiksy
tags = None
if tag == 'adja':
form = form.rstrip('+')
if tag == 'adjc':
if form not in ADJPREDYKATYWNE:
tags = ["adj:sg:nom:m1.m2.m3:pos", "adj:sg:acc:m3:pos"]
if refl and pos in ('v', 'pact', 'ger'):
if refl_id in refls:
tag += ':' + REFL_TRANSLATION[refls[refl_id]]
else:
debug(entry, u'Nieznana zwrotność: %s' % refl_id)
tags = tags or [tag]
for tag in tags:
if data['commonness']:
cv = cv_table[cv_id] if cv_id else ''
output_file.write(
(u'%s\t%s\t%s\t%s\n' % (form, entry, tag, cv)).encode('utf-8'))
else:
output_file.write(
(u'%s\t%s\t%s\n' % (form, entry, tag)).encode('utf-8'))