-
Notifications
You must be signed in to change notification settings - Fork 97
/
__init__.py
398 lines (339 loc) · 14.9 KB
/
__init__.py
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
388
389
390
391
392
393
394
395
396
397
398
# pyenchant
#
# Copyright (C) 2004-2008, Ryan Kelly
#
# This library is free software; you can redistribute it and/or
# modify it under the terms of the GNU Lesser General Public
# License as published by the Free Software Foundation; either
# version 2.1 of the License, or (at your option) any later version.
#
# This library 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
# Lesser General Public License for more details.
#
# You should have received a copy of the GNU Lesser General Public
# License along with this library; if not, write to the
# Free Software Foundation, Inc., 59 Temple Place - Suite 330,
# Boston, MA 02111-1307, USA.
#
# In addition, as a special exception, you are
# given permission to link the code of this program with
# non-LGPL Spelling Provider libraries (eg: a MSFT Office
# spell checker backend) and distribute linked combinations including
# the two. You must obey the GNU Lesser General Public License in all
# respects for all of the code used other than said providers. If you modify
# this file, you may extend this exception to your version of the
# file, but you are not obligated to do so. If you do not wish to
# do so, delete this exception statement from your version.
#
"""
enchant.checker: High-level spellchecking functionality
========================================================
This package is designed to host higher-level spellchecking functionality
than is available in the base enchant package. It should make writing
applications that follow common usage idioms significantly easier.
The most useful class is :py:class:`SpellChecker`, which implements a spellchecking
loop over a block of text. It is capable of modifying the text in-place
if given an array of characters to work with.
This package also contains several interfaces to the SpellChecker class,
such as a wxPython GUI dialog and a command-line interface.
"""
import array
import warnings
from typing import List, Optional, Type, Union
import enchant
from enchant import Dict
from enchant.errors import * # noqa F401,F403
from enchant.errors import (
DefaultLanguageNotFoundError,
DictNotFoundError,
TokenizerNotFoundError,
)
from enchant.tokenize import Chunker, Filter, get_tokenizer, tokenize
from enchant.utils import get_default_language
class SpellChecker:
"""Class implementing stateful spellchecking behaviour.
This class is designed to implement a spell-checking loop over
a block of text, correcting/ignoring/replacing words as required.
This loop is implemented using an iterator paradigm so it can be
embedded inside other loops of control.
The `SpellChecker` object is stateful, and the appropriate methods
must be called to alter its state and affect the progress of
the spell checking session. At any point during the checking
session, the attribute :py:attr:`word` will hold the current erroneously
spelled word under consideration. The action to take on this word
is determined by calling methods such as :py:meth:`replace`, :py:meth:`replace_always`
and :py:meth:`ignore_always`. Once this is done, calling :py:meth:`next` advances
to the next misspelled word.
As a quick (and rather silly) example, the following code replaces
each misspelled word with the string "SPAM":
>>> text = "This is sme text with a fw speling errors in it."
>>> chkr = SpellChecker("en_US",text)
>>> for err in chkr:
... err.replace("SPAM")
...
>>> chkr.get_text()
'This is SPAM text with a SPAM SPAM errors in it.'
>>>
Internally, the `SpellChecker` always works with arrays of (possibly
unicode) character elements. This allows the in-place modification
of the string as it is checked, and is the closest thing Python has
to a mutable string. The text can be set as any of a normal string,
unicode string, character array or unicode character array. The
:py:meth:`get_text` method will return the modified array object if an
array is used, or a new string object if a string it used.
Words input to the `SpellChecker` may be either plain strings or
unicode objects. They will be converted to the same type as the
text being checked, using python's default encoding/decoding
settings.
If using an array of characters with this object and the
array is modified outside of the spellchecking loop, use the
method :py:meth:`set_offset` to reposition the internal loop pointer
to make sure it doesn't skip any words.
"""
_DOC_ERRORS = ["sme", "fw", "speling", "chkr", "chkr", "chkr"]
def __init__(
self,
lang: Union[Dict, str] = None,
text: Optional[str] = None,
tokenize: Union[Type[tokenize], Filter] = None,
chunkers: List[Chunker] = None,
filters: List[Filter] = None,
) -> None:
"""Constructor for the `SpellChecker` class.
`SpellChecker` objects can be created in two ways, depending on
the nature of the first argument. If it is a string, it
specifies a language tag from which a dictionary is created.
Otherwise, it must be an :py:class:`enchant.Dict` object to be used.
Optional keyword arguments are:
:param text: to set the text to be checked at creation time
:param tokenize: a custom tokenization function to use
:param chunkers: a list of chunkers to apply during tokenization
:param filters: a list of filters to apply during tokenization
If `tokenize` is not given and the first argument is a :py:class:`Dict`,
its `tag` attribute must be a language tag so that a tokenization
function can be created automatically. If this attribute is missing
the user's default language will be used.
"""
if lang is None:
lang = get_default_language()
if isinstance(lang, (str, bytes)):
try:
dict = enchant.Dict(lang)
except DictNotFoundError:
raise DefaultLanguageNotFoundError(lang) from None
else:
dict = lang
try:
lang = dict.tag
except AttributeError:
lang = get_default_language()
if lang is None:
raise DefaultLanguageNotFoundError from None
self.lang = lang
self.dict = dict
if tokenize is None:
try:
tokenize = get_tokenizer(lang, chunkers, filters)
except TokenizerNotFoundError:
# Fall back to default tokenization if no match for 'lang'
tokenize = get_tokenizer(None, chunkers, filters)
self._tokenize = tokenize
self.word = None
self.wordpos = None
self._ignore_words = {}
self._replace_words = {}
# Default to the empty string as the text to be checked
self._text = array.array("u")
self._use_tostring = False
self._tokens = iter([])
if text is not None:
self.set_text(text)
def __iter__(self):
"""Each SpellChecker object is its own iterator"""
return self
def set_text(self, text: str) -> None:
"""Set the text to be spell-checked.
This method must be called, or the `text` argument supplied
to the constructor, before calling the method :py:meth:`next()`.
"""
# Convert to an array object if necessary
if isinstance(text, (str, bytes)):
if type(text) is str:
self._text = array.array("u", text)
else:
self._text = array.array("c", text)
self._use_tostring = True
else:
self._text = text
self._use_tostring = False
self._tokens = self._tokenize(self._text)
def get_text(self) -> str:
"""Return the spell-checked text."""
if self._use_tostring:
return self._array_to_string(self._text)
return self._text
def _array_to_string(self, text):
"""Format an internal array as a standard string."""
if text.typecode == "u":
return text.tounicode()
return text.tostring()
def wants_unicode(self) -> bool:
"""Check whether the checker wants unicode strings.
This method will return `True` if the checker wants unicode strings
as input, `False` if it wants normal strings. It's important to
provide the correct type of string to the checker.
"""
return self._text.typecode == "u"
def coerce_string(self, text: str, enc: Optional[str] = None) -> str:
"""Coerce string into the required type.
This method can be used to automatically ensure that strings
are of the correct type required by this checker - either unicode
or standard. If there is a mismatch, conversion is done using
python's default encoding unless another encoding is specified.
"""
if self.wants_unicode():
if not isinstance(text, str):
if enc is None:
return text.decode()
else:
return text.decode(enc)
return text
if not isinstance(text, bytes):
if enc is None:
return text.encode()
else:
return text.encode(enc)
return text
def __next__(self):
return self.next()
def next(self) -> "SpellChecker":
"""Process text up to the next spelling error.
This method is designed to support the iterator protocol.
Each time it is called, it will advance the :py:attr:`word` attribute
to the next spelling error in the text. When no more errors
are found, it will raise :py:exc:`StopIteration`.
The method will always return `self`, so that it can be used
sensibly in common idioms such as::
for err in checker:
err.do_something()
"""
# Find the next spelling error.
# The uncaught StopIteration from next(self._tokens)
# will provide the StopIteration for this method
while True:
(word, pos) = next(self._tokens)
# decode back to a regular string
word = self._array_to_string(word)
if self.dict.check(word):
continue
if word in self._ignore_words:
continue
self.word = word
self.wordpos = pos
if word in self._replace_words:
self.replace(self._replace_words[word])
continue
break
return self
def replace(self, repl: str) -> None:
"""Replace the current erroneous word with the given string."""
repl = self.coerce_string(repl)
a_repl = array.array(self._text.typecode, repl)
if repl:
self.dict.store_replacement(self.word, repl)
self._text[self.wordpos : self.wordpos + len(self.word)] = a_repl
incr = len(repl) - len(self.word)
self._tokens.set_offset(self._tokens.offset + incr, replaced=True)
def replace_always(self, word: str, repl: Optional[str] = None) -> None:
"""Always replace given word with given replacement.
If a single argument is given, this is used to replace the
current erroneous word. If two arguments are given, that
combination is added to the list for future use.
"""
if repl is None:
repl = word
word = self.word
repl = self.coerce_string(repl)
word = self.coerce_string(word)
self._replace_words[word] = repl
if self.word == word:
self.replace(repl)
def ignore_always(self, word: Optional[str] = None) -> None:
"""Add given word to list of words to ignore.
If no word is given, the current erroneous word is added.
"""
if word is None:
word = self.word
word = self.coerce_string(word)
if word not in self._ignore_words:
self._ignore_words[word] = True
def add_to_personal(self, word: Optional[str] = None) -> None:
"""Add given word to the personal word list.
If no word is given, the current erroneous word is added.
"""
warnings.warn(
"SpellChecker.add_to_personal is deprecated, "
"please use SpellChecker.add",
category=DeprecationWarning,
stacklevel=2,
)
self.add(word)
def add(self, word: Optional[str] = None) -> None:
"""Add given word to the personal word list.
If no word is given, the current erroneous word is added.
"""
if word is None:
word = self.word
self.dict.add(word)
def suggest(self, word: Optional[str] = None) -> List[str]:
"""Return suggested spellings for the given word.
If no word is given, the current erroneous word is used.
"""
if word is None:
word = self.word
suggs = self.dict.suggest(word)
return suggs
def check(self, word: str) -> bool:
"""Check correctness of the given word."""
return self.dict.check(word)
def set_offset(self, off: int, whence: int = 0) -> None:
"""Set the offset of the tokenization routine.
For more details on the purpose of the tokenization offset,
see the documentation of the module :py:mod:`enchant.tokenize`.
The optional argument `whence` indicates the method by
which to change the offset:
* 0 (the default) treats `off` as an increment
* 1 treats `off` as a distance from the start
* 2 treats `off` as a distance from the end
"""
if whence == 0:
self._tokens.set_offset(self._tokens.offset + off)
elif whence == 1:
assert off > 0
self._tokens.set_offset(off)
elif whence == 2:
assert off > 0
self._tokens.set_offset(len(self._text) - 1 - off)
else:
raise ValueError("Invalid value for whence: %s" % (whence,))
def leading_context(self, chars: int) -> str:
"""Get `chars` characters of leading context.
This method returns up to `chars` characters of leading
context - the text that occurs in the string immediately
before the current erroneous word.
"""
start = max(self.wordpos - chars, 0)
context = self._text[start : self.wordpos]
return self._array_to_string(context)
def trailing_context(self, chars: int) -> str:
"""Get `chars` characters of trailing context.
This method returns up to `chars` characters of trailing
context - the text that occurs in the string immediately
after the current erroneous word.
"""
start = self.wordpos + len(self.word)
end = min(start + chars, len(self._text))
context = self._text[start:end]
return self._array_to_string(context)