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
|
#!/usr/bin/env python
#Copyright 2007 Sebastian Hagen
# This file is part of wordnet_tools.
# wordnet_tools is free software; you can redistribute it and/or modify
# it under the terms of the GNU General Public License version 2
# as published by the Free Software Foundation
# wordnet_tools is distributed in the hope that it will be useful,
# but WITHOUT ANY WARRANTY; without even the implied warranty of
# MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
# GNU General Public License for more details.
# You should have received a copy of the GNU General Public License
# along with wordnet_tools; if not, write to the Free Software
# Foundation, Inc., 51 Franklin St, Fifth Floor, Boston, MA 02110-1301 USA
# This program requires python >= 2.4.
# This program converts wordnet index/data file pairs into dict index/data
# files usable by dictd.
# This is basically a reimplementation of the wnfilter program by Rik Faith,
# which unfortunately doesn't work correctly for wordnet files in the newer
# formats. This version of wordnet_structures whould parse wordnet 2.1 files
# correctly, and create output very similar to what wnfilter would have
# written.
import datetime
from textwrap import TextWrapper
CAT_ADJECTIVE = 0
CAT_ADVERB = 1
CAT_NOUN = 2
CAT_VERB = 3
category_map = {
'n': CAT_NOUN,
'v': CAT_VERB,
'a': CAT_ADJECTIVE,
's': CAT_ADJECTIVE,
'r': CAT_ADVERB
}
class WordIndex:
def __init__(self, lemma, category, ptrs, synsets, tagsense_count):
self.lemma = lemma
self.category = category
self.ptrs = ptrs
self.synsets = synsets
self.tagsense_count = tagsense_count
@classmethod
def build_from_line(cls, line_data, synset_map):
line_split = line_data.split()
lemma = line_split[0]
category = category_map[line_split[1]]
synset_count = int(line_split[2],10)
ptr_count = int(line_split[3],10)
ptrs = [line_split[i] for i in range(3, 3+ptr_count)]
tagsense_count = int(line_split[5 + ptr_count],10)
synsets = [synset_map[int(line_split[i],10)] for i in range(6 + ptr_count, 6 + ptr_count + synset_count)]
return cls(lemma, category, ptrs, synsets, tagsense_count)
@classmethod
def build_from_file(cls, f, synset_map, rv_base=None):
if (rv_base is None):
rv = {}
else:
rv = rv_base
for line in f:
if (line.startswith(' ')):
continue
wi = cls.build_from_line(line, synset_map)
word = wi.lemma.lower()
if not (word in rv):
rv[word] = []
rv[word].append(wi)
return rv
def __repr__(self):
return '%s%s' % (self.__class__.__name__, (self.lemma, self.category, self.ptrs, self.synsets, self.tagsense_count))
class WordIndexDictFormatter(WordIndex):
category_map_rev = {
CAT_NOUN: 'n',
CAT_VERB: 'v',
CAT_ADJECTIVE: 'adj',
CAT_ADVERB: 'adv'
}
linesep = '\n'
LINE_WIDTH_MAX = 68
prefix_fmtf_line_first = '%5s 1: '
prefix_fmtn_line_first = ' '
prefix_fmtf_line_nonfirst = '%5d: '
prefix_fmtn_line_nonfirst = ' '
def dict_str(self):
tw = TextWrapper(width=self.LINE_WIDTH_MAX,
initial_indent=(self.prefix_fmtf_line_first % self.category_map_rev[self.category]),
subsequent_indent=self.prefix_fmtn_line_first)
lines = (tw.wrap(self.synsets[0].dict_str()))
i = 2
for synset in self.synsets[1:]:
tw = TextWrapper(width=self.LINE_WIDTH_MAX,
initial_indent=(self.prefix_fmtf_line_nonfirst % i),
subsequent_indent=self.prefix_fmtn_line_nonfirst)
lines.extend(tw.wrap(synset.dict_str()))
i += 1
return self.linesep.join(lines)
class Synset:
def __init__(self, offset, ss_type, words, ptrs, gloss, frames=()):
self.offset = offset
self.type = ss_type
self.words = words
self.ptrs = ptrs
self.gloss = gloss
self.frames = frames
self.comments = []
@classmethod
def build_from_line(cls, line_data):
line_split = line_data.split()
synset_offset = int(line_split[0],10)
ss_type = category_map[line_split[2]]
word_count = int(line_split[3],16)
words = [line_split[i] for i in range(4, 4 + word_count*2,2)]
ptr_count = int(line_split[4 + word_count*2],10)
ptrs = [(line_split[i], line_split[i+1], line_split[i+2], line_split[i+3]) for i in range(5 + word_count*2,4 + word_count*2 + ptr_count*4,4)]
tok = line_split[5 + word_count*2 + ptr_count*4]
base = 6 + word_count*2 + ptr_count*4
if (tok != '|'):
frame_count = int(tok, 10)
frames = [(int(line_split[i+1],10), int(line_split[i+2],16)) for i in range(base, base + frame_count*3, 3)]
base += frame_count*3 + 1
else:
frames = []
line_split2 = line_data.split(None, base)
if (len(line_split2) < base):
gloss = None
else:
gloss = line_split2[-1]
return cls(synset_offset, ss_type, words, ptrs, gloss, frames)
@classmethod
def build_from_file(cls, f):
rv = {}
comments = []
for line in f:
if (line.startswith(' ')):
line_s = line.lstrip().rstrip('\n')
line_elements = line_s.split(None,1)
try:
int(line_elements[0])
except ValueError:
continue
if (len(line_elements) == 1):
line_elements.append('')
comments.append(line_elements[1])
continue
synset = cls.build_from_line(line.rstrip())
rv[synset.offset] = synset
return (rv, comments)
def dict_str(self):
rv = self.gloss
if (len(self.words) > 1):
rv += ' [syn: %s]' % (', '.join([('{%s}' % word) for word in self.words]))
return rv
def __repr__(self):
return '%s%s' % (self.__class__.__name__, (self.offset, self.type, self.words, self.ptrs, self.gloss, self.frames))
class WordnetDict:
db_info_fmt = '''This file was converted from the original database on:
%(conversion_datetime)s
The original data is available from:
%(wn_url)s
The original data was distributed with the notice shown below. No
additional restrictions are claimed. Please redistribute this changed
version under the same conditions and restriction that apply to the
original version.\n\n
%(wn_license)s'''
datetime_fmt = '%Y-%m-%dT%H:%M:%S'
base64_map = 'ABCDEFGHIJKLMNOPQRSTUVWXYZabcdefghijklmnopqrstuvwxyz0123456789+/'
def __init__(self, wn_url, desc_short, desc_long):
self.word_data = {}
self.wn_url = wn_url
self.desc_short = desc_short
self.desc_long = desc_long
self.wn_license = None
def wn_dict_add(self, file_index, file_data):
file_data.seek(0)
file_index.seek(0)
(synsets, license_lines) = Synset.build_from_file(file_data)
WordIndexDictFormatter.build_from_file(file_index, synsets, self.word_data)
if (license_lines):
self.wn_license = '\n'.join(license_lines) + '\n'
@classmethod
def base64_encode(cls, i):
"""Encode a non-negative integer into a dictd compatible base64 string"""
if (i < 0):
raise ValueError('Value %r for i is negative' % (i,))
r = 63
e = 1
while (r < i):
e += 1
r = 64**e - 1
rv = ''
while (e > 0):
e -= 1
d = (i / 64**e)
rv += cls.base64_map[d]
i = i % (64**e)
return rv
@classmethod
def dict_entry_write(cls, file_index, file_data, key, entry, linesep='\n'):
"""Write a single dict entry for <key> to index and data files"""
entry_start = file_data.tell()
file_data.write(entry)
entry_len = len(entry)
file_index.write('%s\t%s\t%s%s' % (key, cls.base64_encode(entry_start),
cls.base64_encode(entry_len), linesep))
def dict_generate(self, file_index, file_data):
file_index.seek(0)
file_data.seek(0)
# The dictd file format is fairly iffy on the subject of special
# headwords: either dictd is buggy, or the manpage doesn't tell the whole
# story about the format.
# The upshot is that order of these entries in the index *matters*.
# Putting them at the beginning and in alphabetic order is afaict ok.
# Some other orders completely and quietly break the ability to look
# those headwords up.
# -- problem encountered with 1.10.2, at 2007-08-05.
file_data.write('\n')
wn_url = self.wn_url
conversion_datetime = datetime.datetime.now().strftime(self.datetime_fmt)
wn_license = self.wn_license
self.dict_entry_write(file_index, file_data, '00-database-info', '00-database-info\n%s\n' % (self.db_info_fmt % vars()))
self.dict_entry_write(file_index, file_data, '00-database-long', '00-database-long\n%s\n' % self.desc_long)
self.dict_entry_write(file_index, file_data, '00-database-short', '00-database-short\n%s\n' % self.desc_short)
self.
|