nltk /nltk/ccg/lexicon.py

Language Python Lines 245
MD5 Hash c581f43b7bd16ec24977bb7f2a5c8621
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
# Natural Language Toolkit: Combinatory Categorial Grammar
#
# Copyright (C) 2001-2013 NLTK Project
# Author: Graeme Gange <ggange@csse.unimelb.edu.au>
# URL: <http://www.nltk.org/>
# For license information, see LICENSE.TXT
from __future__ import unicode_literals

import re
from collections import defaultdict

from nltk.ccg.api import PrimitiveCategory, Direction, CCGVar, FunctionalCategory
from nltk.compat import python_2_unicode_compatible

#------------
# Regular expressions used for parsing components of the lexicon
#------------

# Parses a primitive category and subscripts
rePrim = re.compile(r'''([A-Za-z]+)(\[[A-Za-z,]+\])?''')

# Separates the next primitive category from the remainder of the
# string
reNextPrim = re.compile(r'''([A-Za-z]+(?:\[[A-Za-z,]+\])?)(.*)''')

# Separates the next application operator from the remainder
reApp = re.compile(r'''([\\/])([.,]?)([.,]?)(.*)''')

# Parses the definition of the category of either a word or a family
reLex = re.compile(r'''([A-Za-z_]+)\s*(::|[-=]+>)\s*(.+)''')

# Strips comments from a line
reComm = re.compile('''([^#]*)(?:#.*)?''')

#----------
# Lexicons
#----------
@python_2_unicode_compatible
class CCGLexicon(object):
    '''
    Class representing a lexicon for CCG grammars.
    primitives - The list of primitive categories for the lexicon
    families - Families of categories
    entries - A mapping of words to possible categories
    '''
    def __init__(self,start,primitives,families,entries):
        self._start = PrimitiveCategory(start)
        self._primitives = primitives
        self._families = families
        self._entries = entries

    # Returns all the possible categories for a word
    def categories(self,word):
        return self._entries[word]

    # Returns the target category for the parser
    def start(self):
        return self._start

    # String representation of the lexicon
    # Used for debugging
    def __str__(self):
        st = ""
        first = True
        for ident in self._entries:
            if not first:
                st = st + "\n"
            st = st + ident + " => "

            first = True
            for cat in self._entries[ident]:
                if not first:
                    st = st + " | "
                else:
                    first = False
                st = st + "%s" % cat
        return st


#-----------
# Parsing lexicons
#-----------

# Separates the contents matching the first set of brackets
# from the rest of the input.
def matchBrackets(string):
    rest = string[1:]
    inside = "("

    while rest != "" and not rest.startswith(')'):
        if rest.startswith('('):
            (part,rest) = matchBrackets(rest)
            inside = inside + part
        else:
            inside = inside + rest[0]
            rest = rest[1:]
    if rest.startswith(')'):
        return (inside + ')',rest[1:])
    raise AssertionError('Unmatched bracket in string \'' + string + '\'')

# Separates the string for the next portion of the category
# from the rest of the string
def nextCategory(string):
    if string.startswith('('):
        return matchBrackets(string)
    return reNextPrim.match(string).groups()

# Parses an application operator
def parseApplication(app):
    return Direction(app[0],app[1:])

# Parses the subscripts for a primitive category
def parseSubscripts(subscr):
    if subscr:
        return subscr[1:-1].split(',')
    return []

# Parse a primitive category
def parsePrimitiveCategory(chunks,primitives,families,var):
    # If the primitive is the special category 'var',
    # replace it with the correct CCGVar
    if chunks[0] == "var":
        if chunks[1] is None:
            if var is None:
                var = CCGVar()
            return (var,var)

    catstr = chunks[0]
    if catstr in families:
        (cat, cvar) = families[catstr]
        if var is None:
            var = cvar
        else:
            cat = cat.substitute([(cvar,var)])
        return (cat,var)

    if catstr in primitives:
        subscrs = parseSubscripts(chunks[1])
        return (PrimitiveCategory(catstr,subscrs),var)
    raise AssertionError('String \'' + catstr + '\' is neither a family nor primitive category.')

# parseCategory drops the 'var' from the tuple
def parseCategory(line,primitives,families):
    return augParseCategory(line,primitives,families)[0]

# Parses a string representing a category, and returns
# a tuple with (possibly) the CCG variable for the category
def augParseCategory(line,primitives,families,var = None):
    (str,rest) = nextCategory(line)

    if str.startswith('('):
        (res,var) = augParseCategory(str[1:-1],primitives,families,var)

    else:
#        print rePrim.match(str).groups()
        (res,var) = parsePrimitiveCategory(rePrim.match(str).groups(),primitives,families,var)

    while rest != "":
        app = reApp.match(rest).groups()
        dir = parseApplication(app[0:3])
        rest = app[3]

        (str,rest) = nextCategory(rest)
        if str.startswith('('):
            (arg,var) = augParseCategory(str[1:-1],primitives,families,var)
        else:
            (arg,var) = parsePrimitiveCategory(rePrim.match(str).groups(),primitives,families,var)
        res = FunctionalCategory(res,arg,dir)

    return (res,var)

# Takes an input string, and converts it into a lexicon for CCGs.
def parseLexicon(lex_str):
    primitives = []
    families = {}
    entries = defaultdict(list)
    for line in lex_str.splitlines():
        # Strip comments and leading/trailing whitespace.
        line = reComm.match(line).groups()[0].strip()
        if line == "":
            continue

        if line.startswith(':-'):
            # A line of primitive categories.
            # The first line is the target category
            # ie, :- S, N, NP, VP
            primitives = primitives + [ prim.strip() for prim in line[2:].strip().split(',') ]
        else:
            # Either a family definition, or a word definition
            (ident, sep, catstr) = reLex.match(line).groups()
            (cat,var) = augParseCategory(catstr,primitives,families)
            if sep == '::':
                # Family definition
                # ie, Det :: NP/N
                families[ident] = (cat,var)
            else:
                # Word definition
                # ie, which => (N\N)/(S/NP)
                entries[ident].append(cat)
    return CCGLexicon(primitives[0],primitives,families,entries)


openccg_tinytiny = parseLexicon('''
    # Rather minimal lexicon based on the openccg `tinytiny' grammar.
    # Only incorporates a subset of the morphological subcategories, however.
    :- S,NP,N                    # Primitive categories
    Det :: NP/N                  # Determiners
    Pro :: NP
    IntransVsg :: S\\NP[sg]    # Tensed intransitive verbs (singular)
    IntransVpl :: S\\NP[pl]    # Plural
    TransVsg :: S\\NP[sg]/NP   # Tensed transitive verbs (singular)
    TransVpl :: S\\NP[pl]/NP   # Plural

    the => NP[sg]/N[sg]
    the => NP[pl]/N[pl]

    I => Pro
    me => Pro
    we => Pro
    us => Pro

    book => N[sg]
    books => N[pl]

    peach => N[sg]
    peaches => N[pl]

    policeman => N[sg]
    policemen => N[pl]

    boy => N[sg]
    boys => N[pl]

    sleep => IntransVsg
    sleep => IntransVpl

    eat => IntransVpl
    eat => TransVpl
    eats => IntransVsg
    eats => TransVsg

    see => TransVpl
    sees => TransVsg
    ''')
Back to Top