nltk /nltk/corpus/reader/verbnet.py

Language Python Lines 390
MD5 Hash f371dccb15962bb5ac29273d1ac43162
Repository https://github.com/BrucePHill/nltk.git View Raw File
  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
# Natural Language Toolkit: Verbnet Corpus Reader
#
# Copyright (C) 2001-2013 NLTK Project
# Author: Edward Loper <edloper@gradient.cis.upenn.edu>
# URL: <http://www.nltk.org/>
# For license information, see LICENSE.TXT
from __future__ import unicode_literals

import re
import textwrap
from collections import defaultdict

from nltk import compat
from .xmldocs import XMLCorpusReader

class VerbnetCorpusReader(XMLCorpusReader):

    # No unicode encoding param, since the data files are all XML.
    def __init__(self, root, fileids, wrap_etree=False):
        XMLCorpusReader.__init__(self, root, fileids, wrap_etree)

        self._lemma_to_class = defaultdict(list)
        """A dictionary mapping from verb lemma strings to lists of
        verbnet class identifiers."""

        self._wordnet_to_class = defaultdict(list)
        """A dictionary mapping from wordnet identifier strings to
        lists of verbnet class identifiers."""

        self._class_to_fileid = {}
        """A dictionary mapping from class identifiers to
        corresponding file identifiers.  The keys of this dictionary
        provide a complete list of all classes and subclasses."""

        self._shortid_to_longid = {}

        # Initialize the dictionaries.  Use the quick (regexp-based)
        # method instead of the slow (xml-based) method, because it
        # runs 2-30 times faster.
        self._quick_index()

    _LONGID_RE = re.compile(r'([^\-\.]*)-([\d+.\-]+)$')
    """Regular expression that matches (and decomposes) longids"""

    _SHORTID_RE = re.compile(r'[\d+.\-]+$')
    """Regular expression that matches shortids"""

    _INDEX_RE = re.compile(r'<MEMBER name="\??([^"]+)" wn="([^"]*)"[^>]+>|'
                           r'<VNSUBCLASS ID="([^"]+)"/?>')
    """Regular expression used by ``_index()`` to quickly scan the corpus
       for basic information."""

    def lemmas(self, classid=None):
        """
        Return a list of all verb lemmas that appear in any class, or
        in the ``classid`` if specified.
        """
        if classid is None:
            return sorted(self._lemma_to_class.keys())
        else:
            # [xx] should this include subclass members?
            vnclass = self.vnclass(classid)
            return [member.get('name') for member in
                    vnclass.findall('MEMBERS/MEMBER')]

    def wordnetids(self, classid=None):
        """
        Return a list of all wordnet identifiers that appear in any
        class, or in ``classid`` if specified.
        """
        if classid is None:
            return sorted(self._wordnet_to_class.keys())
        else:
            # [xx] should this include subclass members?
            vnclass = self.vnclass(classid)
            return sum([member.get('wn','').split() for member in
                        vnclass.findall('MEMBERS/MEMBER')], [])

    def classids(self, lemma=None, wordnetid=None, fileid=None, classid=None):
        """
        Return a list of the verbnet class identifiers.  If a file
        identifier is specified, then return only the verbnet class
        identifiers for classes (and subclasses) defined by that file.
        If a lemma is specified, then return only verbnet class
        identifiers for classes that contain that lemma as a member.
        If a wordnetid is specified, then return only identifiers for
        classes that contain that wordnetid as a member.  If a classid
        is specified, then return only identifiers for subclasses of
        the specified verbnet class.
        """
        if len([x for x in [lemma, wordnetid, fileid, classid]
                if x is not None]) > 1:
            raise ValueError('Specify at most one of: fileid, wordnetid, '
                             'fileid, classid')
        if fileid is not None:
            return [c for (c,f) in self._class_to_fileid.items()
                    if f == fileid]
        elif lemma is not None:
            return self._lemma_to_class[lemma]
        elif wordnetid is not None:
            return self._wordnet_to_class[wordnetid]
        elif classid is not None:
            xmltree = self.vnclass(classid)
            return [subclass.get('ID') for subclass in
                    xmltree.findall('SUBCLASSES/VNSUBCLASS')]
        else:
            return sorted(self._class_to_fileid.keys())

    def vnclass(self, fileid_or_classid):
        """
        Return an ElementTree containing the xml for the specified
        verbnet class.

        :param fileid_or_classid: An identifier specifying which class
            should be returned.  Can be a file identifier (such as
            ``'put-9.1.xml'``), or a verbnet class identifier (such as
            ``'put-9.1'``) or a short verbnet class identifier (such as
            ``'9.1'``).
        """
        # File identifier: just return the xml.
        if fileid_or_classid in self._fileids:
            return self.xml(fileid_or_classid)

        # Class identifier: get the xml, and find the right elt.
        classid = self.longid(fileid_or_classid)
        if classid in self._class_to_fileid:
            fileid = self._class_to_fileid[self.longid(classid)]
            tree = self.xml(fileid)
            if classid == tree.get('ID'):
                return tree
            else:
                for subclass in tree.findall('.//VNSUBCLASS'):
                    if classid == subclass.get('ID'):
                        return subclass
                else:
                    assert False # we saw it during _index()!

        else:
            raise ValueError('Unknown identifier %s' % fileid_or_classid)

    def fileids(self, vnclass_ids=None):
        """
        Return a list of fileids that make up this corpus.  If
        ``vnclass_ids`` is specified, then return the fileids that make
        up the specified verbnet class(es).
        """
        if vnclass_ids is None:
            return self._fileids
        elif isinstance(vnclass_ids, compat.string_types):
            return [self._class_to_fileid[self.longid(vnclass_ids)]]
        else:
            return [self._class_to_fileid[self.longid(vnclass_id)]
                    for vnclass_id in vnclass_ids]


    ######################################################################
    #{ Index Initialization
    ######################################################################

    def _index(self):
        """
        Initialize the indexes ``_lemma_to_class``,
        ``_wordnet_to_class``, and ``_class_to_fileid`` by scanning
        through the corpus fileids.  This is fast with cElementTree
        (<0.1 secs), but quite slow (>10 secs) with the python
        implementation of ElementTree.
        """
        for fileid in self._fileids:
            self._index_helper(self.xml(fileid), fileid)

    def _index_helper(self, xmltree, fileid):
        """Helper for ``_index()``"""
        vnclass = xmltree.get('ID')
        self._class_to_fileid[vnclass] = fileid
        self._shortid_to_longid[self.shortid(vnclass)] = vnclass
        for member in xmltree.findall('MEMBERS/MEMBER'):
            self._lemma_to_class[member.get('name')].append(vnclass)
            for wn in member.get('wn', '').split():
                self._wordnet_to_class[wn].append(vnclass)
        for subclass in xmltree.findall('SUBCLASSES/VNSUBCLASS'):
            self._index_helper(subclass, fileid)

    def _quick_index(self):
        """
        Initialize the indexes ``_lemma_to_class``,
        ``_wordnet_to_class``, and ``_class_to_fileid`` by scanning
        through the corpus fileids.  This doesn't do proper xml parsing,
        but is good enough to find everything in the standard verbnet
        corpus -- and it runs about 30 times faster than xml parsing
        (with the python ElementTree; only 2-3 times faster with
        cElementTree).
        """
        # nb: if we got rid of wordnet_to_class, this would run 2-3
        # times faster.
        for fileid in self._fileids:
            vnclass = fileid[:-4] # strip the '.xml'
            self._class_to_fileid[vnclass] = fileid
            self._shortid_to_longid[self.shortid(vnclass)] = vnclass
            for m in self._INDEX_RE.finditer(self.open(fileid).read()):
                groups = m.groups()
                if groups[0] is not None:
                    self._lemma_to_class[groups[0]].append(vnclass)
                    for wn in groups[1].split():
                        self._wordnet_to_class[wn].append(vnclass)
                elif groups[2] is not None:
                    self._class_to_fileid[groups[2]] = fileid
                    vnclass = groups[2] # for <MEMBER> elts.
                    self._shortid_to_longid[self.shortid(vnclass)] = vnclass
                else:
                    assert False, 'unexpected match condition'

    ######################################################################
    #{ Identifier conversion
    ######################################################################

    def longid(self, shortid):
        """Given a short verbnet class identifier (eg '37.10'), map it
        to a long id (eg 'confess-37.10').  If ``shortid`` is already a
        long id, then return it as-is"""
        if self._LONGID_RE.match(shortid):
            return shortid # it's already a longid.
        elif not self._SHORTID_RE.match(shortid):
            raise ValueError('vnclass identifier %r not found' % shortid)
        try:
            return self._shortid_to_longid[shortid]
        except KeyError:
            raise ValueError('vnclass identifier %r not found' % shortid)

    def shortid(self, longid):
        """Given a long verbnet class identifier (eg 'confess-37.10'),
        map it to a short id (eg '37.10').  If ``longid`` is already a
        short id, then return it as-is."""
        if self._SHORTID_RE.match(longid):
            return longid # it's already a shortid.
        m = self._LONGID_RE.match(longid)
        if m:
            return m.group(2)
        else:
            raise ValueError('vnclass identifier %r not found' % longid)

    ######################################################################
    #{ Pretty Printing
    ######################################################################

    def pprint(self, vnclass):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet class.

        :param vnclass: A verbnet class identifier; or an ElementTree
        containing the xml contents of a verbnet class.
        """
        if isinstance(vnclass, compat.string_types):
            vnclass = self.vnclass(vnclass)

        s = vnclass.get('ID') + '\n'
        s += self.pprint_subclasses(vnclass, indent='  ') + '\n'
        s += self.pprint_members(vnclass, indent='  ') + '\n'
        s += '  Thematic roles:\n'
        s += self.pprint_themroles(vnclass, indent='    ') + '\n'
        s += '  Frames:\n'
        s += '\n'.join(self.pprint_frame(vnframe, indent='    ')
                       for vnframe in vnclass.findall('FRAMES/FRAME'))
        return s

    def pprint_subclasses(self, vnclass, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet class's subclasses.

        :param vnclass: A verbnet class identifier; or an ElementTree
            containing the xml contents of a verbnet class.
        """
        if isinstance(vnclass, compat.string_types):
            vnclass = self.vnclass(vnclass)

        subclasses = [subclass.get('ID') for subclass in
                      vnclass.findall('SUBCLASSES/VNSUBCLASS')]
        if not subclasses: subclasses = ['(none)']
        s = 'Subclasses: ' + ' '.join(subclasses)
        return textwrap.fill(s, 70, initial_indent=indent,
                             subsequent_indent=indent+'  ')

    def pprint_members(self, vnclass, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet class's member verbs.

        :param vnclass: A verbnet class identifier; or an ElementTree
            containing the xml contents of a verbnet class.
        """
        if isinstance(vnclass, compat.string_types):
            vnclass = self.vnclass(vnclass)

        members = [member.get('name') for member in
                   vnclass.findall('MEMBERS/MEMBER')]
        if not members: members = ['(none)']
        s = 'Members: ' + ' '.join(members)
        return textwrap.fill(s, 70, initial_indent=indent,
                             subsequent_indent=indent+'  ')

    def pprint_themroles(self, vnclass, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet class's thematic roles.

        :param vnclass: A verbnet class identifier; or an ElementTree
            containing the xml contents of a verbnet class.
        """
        if isinstance(vnclass, compat.string_types):
            vnclass = self.vnclass(vnclass)

        pieces = []
        for themrole in vnclass.findall('THEMROLES/THEMROLE'):
            piece = indent + '* ' + themrole.get('type')
            modifiers = ['%(Value)s%(type)s' % restr.attrib
                         for restr in themrole.findall('SELRESTRS/SELRESTR')]
            if modifiers:
                piece += '[%s]' % ' '.join(modifiers)
            pieces.append(piece)

        return '\n'.join(pieces)

    def pprint_frame(self, vnframe, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet frame.

        :param vnframe: An ElementTree containing the xml contents of
            a verbnet frame.
        """
        s = self.pprint_description(vnframe, indent) + '\n'
        s += self.pprint_syntax(vnframe, indent+'  Syntax: ') + '\n'
        s += indent + '  Semantics:\n'
        s += self.pprint_semantics(vnframe, indent+'    ')
        return s

    def pprint_description(self, vnframe, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet frame description.

        :param vnframe: An ElementTree containing the xml contents of
            a verbnet frame.
        """
        descr = vnframe.find('DESCRIPTION')
        s = indent + descr.attrib['primary']
        if descr.get('secondary', ''):
            s += ' (%s)' % descr.get('secondary')
        return s

    def pprint_syntax(self, vnframe, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet frame syntax.

        :param vnframe: An ElementTree containing the xml contents of
            a verbnet frame.
        """
        pieces = []
        for elt in vnframe.find('SYNTAX'):
            piece = elt.tag
            modifiers = []
            if 'value' in elt.attrib:
                modifiers.append(elt.get('value'))
            modifiers += ['%(Value)s%(type)s' % restr.attrib
                          for restr in (elt.findall('SELRESTRS/SELRESTR') +
                                        elt.findall('SYNRESTRS/SYNRESTR'))]
            if modifiers:
                piece += '[%s]' % ' '.join(modifiers)
            pieces.append(piece)

        return indent + ' '.join(pieces)

    def pprint_semantics(self, vnframe, indent=''):
        """
        Return a string containing a pretty-printed representation of
        the given verbnet frame semantics.

        :param vnframe: An ElementTree containing the xml contents of
            a verbnet frame.
        """
        pieces = []
        for pred in vnframe.findall('SEMANTICS/PRED'):
            args = [arg.get('value') for arg in pred.findall('ARGS/ARG')]
            pieces.append('%s(%s)' % (pred.get('value'), ', '.join(args)))
        return '\n'.join('%s* %s' % (indent, piece) for piece in pieces)
Back to Top