aboutsummaryrefslogtreecommitdiffstats
path: root/contrib/python/charset-normalizer/charset_normalizer/cd.py
blob: 4ea6760c45bce5773bfe4b46d7b3c07c2c139d49 (plain) (blame)
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
import importlib
from codecs import IncrementalDecoder
from collections import Counter
from functools import lru_cache
from typing import Counter as TypeCounter, Dict, List, Optional, Tuple

from .constant import (
    FREQUENCIES,
    KO_NAMES,
    LANGUAGE_SUPPORTED_COUNT,
    TOO_SMALL_SEQUENCE,
    ZH_NAMES,
)
from .md import is_suspiciously_successive_range
from .models import CoherenceMatches
from .utils import (
    is_accentuated,
    is_latin,
    is_multi_byte_encoding,
    is_unicode_range_secondary,
    unicode_range,
)


def encoding_unicode_range(iana_name: str) -> List[str]:
    """
    Return associated unicode ranges in a single byte code page.
    """
    if is_multi_byte_encoding(iana_name):
        raise IOError("Function not supported on multi-byte code page")

    decoder = importlib.import_module(
        "encodings.{}".format(iana_name)
    ).IncrementalDecoder

    p: IncrementalDecoder = decoder(errors="ignore")
    seen_ranges: Dict[str, int] = {}
    character_count: int = 0

    for i in range(0x40, 0xFF):
        chunk: str = p.decode(bytes([i]))

        if chunk:
            character_range: Optional[str] = unicode_range(chunk)

            if character_range is None:
                continue

            if is_unicode_range_secondary(character_range) is False:
                if character_range not in seen_ranges:
                    seen_ranges[character_range] = 0
                seen_ranges[character_range] += 1
                character_count += 1

    return sorted(
        [
            character_range
            for character_range in seen_ranges
            if seen_ranges[character_range] / character_count >= 0.15
        ]
    )


def unicode_range_languages(primary_range: str) -> List[str]:
    """
    Return inferred languages used with a unicode range.
    """
    languages: List[str] = []

    for language, characters in FREQUENCIES.items():
        for character in characters:
            if unicode_range(character) == primary_range:
                languages.append(language)
                break

    return languages


@lru_cache()
def encoding_languages(iana_name: str) -> List[str]:
    """
    Single-byte encoding language association. Some code page are heavily linked to particular language(s).
    This function does the correspondence.
    """
    unicode_ranges: List[str] = encoding_unicode_range(iana_name)
    primary_range: Optional[str] = None

    for specified_range in unicode_ranges:
        if "Latin" not in specified_range:
            primary_range = specified_range
            break

    if primary_range is None:
        return ["Latin Based"]

    return unicode_range_languages(primary_range)


@lru_cache()
def mb_encoding_languages(iana_name: str) -> List[str]:
    """
    Multi-byte encoding language association. Some code page are heavily linked to particular language(s).
    This function does the correspondence.
    """
    if (
        iana_name.startswith("shift_")
        or iana_name.startswith("iso2022_jp")
        or iana_name.startswith("euc_j")
        or iana_name == "cp932"
    ):
        return ["Japanese"]
    if iana_name.startswith("gb") or iana_name in ZH_NAMES:
        return ["Chinese"]
    if iana_name.startswith("iso2022_kr") or iana_name in KO_NAMES:
        return ["Korean"]

    return []


@lru_cache(maxsize=LANGUAGE_SUPPORTED_COUNT)
def get_target_features(language: str) -> Tuple[bool, bool]:
    """
    Determine main aspects from a supported language if it contains accents and if is pure Latin.
    """
    target_have_accents: bool = False
    target_pure_latin: bool = True

    for character in FREQUENCIES[language]:
        if not target_have_accents and is_accentuated(character):
            target_have_accents = True
        if target_pure_latin and is_latin(character) is False:
            target_pure_latin = False

    return target_have_accents, target_pure_latin


def alphabet_languages(
    characters: List[str], ignore_non_latin: bool = False
) -> List[str]:
    """
    Return associated languages associated to given characters.
    """
    languages: List[Tuple[str, float]] = []

    source_have_accents = any(is_accentuated(character) for character in characters)

    for language, language_characters in FREQUENCIES.items():
        target_have_accents, target_pure_latin = get_target_features(language)

        if ignore_non_latin and target_pure_latin is False:
            continue

        if target_have_accents is False and source_have_accents:
            continue

        character_count: int = len(language_characters)

        character_match_count: int = len(
            [c for c in language_characters if c in characters]
        )

        ratio: float = character_match_count / character_count

        if ratio >= 0.2:
            languages.append((language, ratio))

    languages = sorted(languages, key=lambda x: x[1], reverse=True)

    return [compatible_language[0] for compatible_language in languages]


def characters_popularity_compare(
    language: str, ordered_characters: List[str]
) -> float:
    """
    Determine if a ordered characters list (by occurrence from most appearance to rarest) match a particular language.
    The result is a ratio between 0. (absolutely no correspondence) and 1. (near perfect fit).
    Beware that is function is not strict on the match in order to ease the detection. (Meaning close match is 1.)
    """
    if language not in FREQUENCIES:
        raise ValueError("{} not available".format(language))

    character_approved_count: int = 0
    FREQUENCIES_language_set = set(FREQUENCIES[language])

    ordered_characters_count: int = len(ordered_characters)
    target_language_characters_count: int = len(FREQUENCIES[language])

    large_alphabet: bool = target_language_characters_count > 26

    for character, character_rank in zip(
        ordered_characters, range(0, ordered_characters_count)
    ):
        if character not in FREQUENCIES_language_set:
            continue

        character_rank_in_language: int = FREQUENCIES[language].index(character)
        expected_projection_ratio: float = (
            target_language_characters_count / ordered_characters_count
        )
        character_rank_projection: int = int(character_rank * expected_projection_ratio)

        if (
            large_alphabet is False
            and abs(character_rank_projection - character_rank_in_language) > 4
        ):
            continue

        if (
            large_alphabet is True
            and abs(character_rank_projection - character_rank_in_language)
            < target_language_characters_count / 3
        ):
            character_approved_count += 1
            continue

        characters_before_source: List[str] = FREQUENCIES[language][
            0:character_rank_in_language
        ]
        characters_after_source: List[str] = FREQUENCIES[language][
            character_rank_in_language:
        ]
        characters_before: List[str] = ordered_characters[0:character_rank]
        characters_after: List[str] = ordered_characters[character_rank:]

        before_match_count: int = len(
            set(characters_before) & set(characters_before_source)
        )

        after_match_count: int = len(
            set(characters_after) & set(characters_after_source)
        )

        if len(characters_before_source) == 0 and before_match_count <= 4:
            character_approved_count += 1
            continue

        if len(characters_after_source) == 0 and after_match_count <= 4:
            character_approved_count += 1
            continue

        if (
            before_match_count / len(characters_before_source) >= 0.4
            or after_match_count / len(characters_after_source) >= 0.4
        ):
            character_approved_count += 1
            continue

    return character_approved_count / len(ordered_characters)


def alpha_unicode_split(decoded_sequence: str) -> List[str]:
    """
    Given a decoded text sequence, return a list of str. Unicode range / alphabet separation.
    Ex. a text containing English/Latin with a bit a Hebrew will return two items in the resulting list;
    One containing the latin letters and the other hebrew.
    """
    layers: Dict[str, str] = {}

    for character in decoded_sequence:
        if character.isalpha() is False:
            continue

        character_range: Optional[str] = unicode_range(character)

        if character_range is None:
            continue

        layer_target_range: Optional[str] = None

        for discovered_range in layers:
            if (
                is_suspiciously_successive_range(discovered_range, character_range)
                is False
            ):
                layer_target_range = discovered_range
                break

        if layer_target_range is None:
            layer_target_range = character_range

        if layer_target_range not in layers:
            layers[layer_target_range] = character.lower()
            continue

        layers[layer_target_range] += character.lower()

    return list(layers.values())


def merge_coherence_ratios(results: List[CoherenceMatches]) -> CoherenceMatches:
    """
    This function merge results previously given by the function coherence_ratio.
    The return type is the same as coherence_ratio.
    """
    per_language_ratios: Dict[str, List[float]] = {}
    for result in results:
        for sub_result in result:
            language, ratio = sub_result
            if language not in per_language_ratios:
                per_language_ratios[language] = [ratio]
                continue
            per_language_ratios[language].append(ratio)

    merge = [
        (
            language,
            round(
                sum(per_language_ratios[language]) / len(per_language_ratios[language]),
                4,
            ),
        )
        for language in per_language_ratios
    ]

    return sorted(merge, key=lambda x: x[1], reverse=True)


def filter_alt_coherence_matches(results: CoherenceMatches) -> CoherenceMatches:
    """
    We shall NOT return "English—" in CoherenceMatches because it is an alternative
    of "English". This function only keeps the best match and remove the em-dash in it.
    """
    index_results: Dict[str, List[float]] = dict()

    for result in results:
        language, ratio = result
        no_em_name: str = language.replace("—", "")

        if no_em_name not in index_results:
            index_results[no_em_name] = []

        index_results[no_em_name].append(ratio)

    if any(len(index_results[e]) > 1 for e in index_results):
        filtered_results: CoherenceMatches = []

        for language in index_results:
            filtered_results.append((language, max(index_results[language])))

        return filtered_results

    return results


@lru_cache(maxsize=2048)
def coherence_ratio(
    decoded_sequence: str, threshold: float = 0.1, lg_inclusion: Optional[str] = None
) -> CoherenceMatches:
    """
    Detect ANY language that can be identified in given sequence. The sequence will be analysed by layers.
    A layer = Character extraction by alphabets/ranges.
    """

    results: List[Tuple[str, float]] = []
    ignore_non_latin: bool = False

    sufficient_match_count: int = 0

    lg_inclusion_list = lg_inclusion.split(",") if lg_inclusion is not None else []
    if "Latin Based" in lg_inclusion_list:
        ignore_non_latin = True
        lg_inclusion_list.remove("Latin Based")

    for layer in alpha_unicode_split(decoded_sequence):
        sequence_frequencies: TypeCounter[str] = Counter(layer)
        most_common = sequence_frequencies.most_common()

        character_count: int = sum(o for c, o in most_common)

        if character_count <= TOO_SMALL_SEQUENCE:
            continue

        popular_character_ordered: List[str] = [c for c, o in most_common]

        for language in lg_inclusion_list or alphabet_languages(
            popular_character_ordered, ignore_non_latin
        ):
            ratio: float = characters_popularity_compare(
                language, popular_character_ordered
            )

            if ratio < threshold:
                continue
            elif ratio >= 0.8:
                sufficient_match_count += 1

            results.append((language, round(ratio, 4)))

            if sufficient_match_count >= 3:
                break

    return sorted(
        filter_alt_coherence_matches(results), key=lambda x: x[1], reverse=True
    )