summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/normal.py
blob: 399276511b6ecd2bf73ede233e000e814ae98975 (plain)
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
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557

"""
A bunch of helpers to parse and normalize strings: external identifiers,
free-form input, titles, etc.
"""

import re
import base64
from typing import Optional
import unicodedata

import ftfy
import langdetect
import pycountry

DOI_REGEX = re.compile(r"^10.\d{3,6}/\S+$")

def clean_doi(raw):
    """
    Removes any:
    - padding whitespace
    - 'doi:' prefix
    - URL prefix

    Does not try to un-URL-encode

    Returns None if not a valid DOI
    """
    if not raw:
        return None
    raw = raw.strip()
    if '\u2013' in raw:
        # Do not attempt to normalize "en dash" and since FC does not allow
        # unicode in DOI, treat this as invalid.
        return None
    if len(raw.split()) != 1:
        return None
    if raw.startswith("doi:"):
        raw = raw[4:]
    if raw.startswith("http://"):
        raw = raw[7:]
    if raw.startswith("https://"):
        raw = raw[8:]
    if raw.startswith("doi.org/"):
        raw = raw[8:]
    if raw.startswith("dx.doi.org/"):
        raw = raw[11:]
    if raw[7:9] == "//":
        raw = raw[:8] + raw[9:]

    # fatcatd uses same REGEX, but Rust regex rejects these characters, while
    # python doesn't. DOIs are syntaxtually valid, but very likely to be typos;
    # for now filter them out.
    for c in ('¬', ):
        if c in raw:
            return None

    if not raw.startswith("10."):
        return None
    if not DOI_REGEX.fullmatch(raw):
        return None
    # will likely want to expand DOI_REGEX to exclude non-ASCII characters, but
    # for now block specific characters so we can get PubMed importer running
    # again.
    if 'ä' in raw or '\u200e' in raw:
        return None
    return raw

def test_clean_doi():
    assert clean_doi("10.1234/asdf ") == "10.1234/asdf"
    assert clean_doi("10.1037//0002-9432.72.1.50") == "10.1037/0002-9432.72.1.50"
    assert clean_doi("10.1037/0002-9432.72.1.50") == "10.1037/0002-9432.72.1.50"
    assert clean_doi("10.23750/abm.v88i2 -s.6506") == None
    assert clean_doi("10.17167/mksz.2017.2.129–155") == None
    assert clean_doi("http://doi.org/10.1234/asdf ") == "10.1234/asdf"
    assert clean_doi("https://dx.doi.org/10.1234/asdf ") == "10.1234/asdf"
    assert clean_doi("doi:10.1234/asdf ") == "10.1234/asdf"
    assert clean_doi("doi:10.1234/ asdf ") == None
    assert clean_doi("10.4149/gpb¬_2017042") == None  # "logical negation" character
    assert clean_doi("10.6002/ect.2020.häyry") == None  # this example via pubmed (pmid:32519616)
    assert clean_doi("10.30466/vrf.2019.98547.2350\u200e") == None


ARXIV_ID_REGEX = re.compile(r"^(\d{4}.\d{4,5}|[a-z\-]+(\.[A-Z]{2})?/\d{7})(v\d+)?$")

def clean_arxiv_id(raw):
    """
    Removes any:
    - 'arxiv:' prefix

    Works with versioned or un-versioned arxiv identifiers.
    """
    if not raw:
        return None
    raw = raw.strip()
    if raw.lower().startswith("arxiv:"):
        raw = raw[6:]
    if raw.lower().startswith("https://arxiv.org/abs/"):
        raw = raw[22:]
    if not ARXIV_ID_REGEX.fullmatch(raw):
        return None
    return raw

def test_clean_arxiv_id():
    assert clean_arxiv_id("0806.2878v1") == "0806.2878v1"
    assert clean_arxiv_id("0806.2878") == "0806.2878"
    assert clean_arxiv_id("1501.00001v1") == "1501.00001v1"
    assert clean_arxiv_id("1501.00001") == "1501.00001"
    assert clean_arxiv_id("hep-th/9901001v1") == "hep-th/9901001v1"
    assert clean_arxiv_id("hep-th/9901001") == "hep-th/9901001"
    assert clean_arxiv_id("math.CA/0611800v2") == "math.CA/0611800v2"
    assert clean_arxiv_id("math.CA/0611800") == "math.CA/0611800"
    assert clean_arxiv_id("0806.2878v1 ") == "0806.2878v1"

    assert clean_arxiv_id("https://arxiv.org/abs/0806.2878v1") == "0806.2878v1"
    assert clean_arxiv_id("arxiv:0806.2878v1") == "0806.2878v1"
    assert clean_arxiv_id("arXiv:0806.2878v1") == "0806.2878v1"

    assert clean_arxiv_id("hep-TH/9901001v1") == None
    assert clean_arxiv_id("hßp-th/9901001v1") == None
    assert clean_arxiv_id("math.CA/06l1800v2") == None
    assert clean_arxiv_id("mßth.ca/0611800v2") == None
    assert clean_arxiv_id("MATH.CA/0611800v2") == None
    assert clean_arxiv_id("0806.2878v23") == "0806.2878v23"  # ?
    assert clean_arxiv_id("0806.2878v") == None
    assert clean_arxiv_id("0806.2878") == "0806.2878"
    assert clean_arxiv_id("006.2878v1") == None
    assert clean_arxiv_id("0806.v1") == None
    assert clean_arxiv_id("08062878v1") == None

def clean_pmid(raw):
    if not raw:
        return None
    raw = raw.strip()
    if len(raw.split()) != 1:
        return None
    if raw.isdigit():
        return raw
    return None

def test_clean_pmid():
    assert clean_pmid("1234") == "1234"
    assert clean_pmid("1234 ") == "1234"
    assert clean_pmid("PMC123") == None
    assert clean_sha1("qfba3") == None
    assert clean_sha1("") == None

def clean_pmcid(raw):
    if not raw:
        return None
    raw = raw.strip()
    if len(raw.split()) != 1:
        return None
    if raw.startswith("PMC") and raw[3:] and raw[3:].isdigit():
        return raw
    return None

def clean_sha1(raw):
    if not raw:
        return None
    raw = raw.strip().lower()
    if len(raw.split()) != 1:
        return None
    if len(raw) != 40:
        return None
    for c in raw:
        if c not in "0123456789abcdef":
            return None
    return raw

def test_clean_sha1():
    assert clean_sha1("0fba3fba0e1937aa0297de3836b768b5dfb23d7b") == "0fba3fba0e1937aa0297de3836b768b5dfb23d7b"
    assert clean_sha1("0fba3fba0e1937aa0297de3836b768b5dfb23d7b ") == "0fba3fba0e1937aa0297de3836b768b5dfb23d7b"
    assert clean_sha1("fba3fba0e1937aa0297de3836b768b5dfb23d7b") == None
    assert clean_sha1("qfba3fba0e1937aa0297de3836b768b5dfb23d7b") == None
    assert clean_sha1("0fba3fb a0e1937aa0297de3836b768b5dfb23d7b") == None

def clean_sha256(raw):
    raw = raw.strip().lower()
    if len(raw.split()) != 1:
        return None
    if len(raw) != 64:
        return None
    for c in raw:
        if c not in "0123456789abcdef":
            return None
    return raw

def test_clean_sha256():
    assert clean_sha256("6cc853f2ae75696b2e45f476c76b946b0fc2df7c52bb38287cb074aceb77bc7f") == "6cc853f2ae75696b2e45f476c76b946b0fc2df7c52bb38287cb074aceb77bc7f"
    assert clean_sha256("0fba3fba0e1937aa0297de3836b768b5dfb23d7b") == None

ISSN_REGEX = re.compile(r"^\d{4}-\d{3}[0-9X]$")

def clean_issn(raw):
    if not raw:
        return None
    raw = raw.strip().upper()
    if len(raw) != 9:
        return None
    if not ISSN_REGEX.fullmatch(raw):
        return None
    return raw

def test_clean_issn():
    assert clean_issn("1234-4567") == "1234-4567"
    assert clean_issn("1234-456X") == "1234-456X"
    assert clean_issn("134-4567") == None
    assert clean_issn("123X-4567") == None

ISBN13_REGEX = re.compile(r"^97(?:8|9)-\d{1,5}-\d{1,7}-\d{1,6}-\d$")

def clean_isbn13(raw):
    if not raw:
        return None
    raw = raw.strip()
    if not ISBN13_REGEX.fullmatch(raw):
        return None
    return raw

def test_clean_isbn13():
    assert clean_isbn13("978-1-56619-909-4") == "978-1-56619-909-4"
    assert clean_isbn13("978-1-4028-9462-6") == "978-1-4028-9462-6"
    assert clean_isbn13("978-1-56619-909-4 ") == "978-1-56619-909-4"
    assert clean_isbn13("9781566199094") == None

ORCID_REGEX = re.compile(r"^\d{4}-\d{4}-\d{4}-\d{3}[\dX]$")

def clean_orcid(raw):
    if not raw:
        return None
    raw = raw.strip()
    if not ORCID_REGEX.fullmatch(raw):
        return None
    return raw

def test_clean_orcid():
    assert clean_orcid("0123-4567-3456-6789") == "0123-4567-3456-6789"
    assert clean_orcid("0123-4567-3456-678X") == "0123-4567-3456-678X"
    assert clean_orcid("0123-4567-3456-6789 ") == "0123-4567-3456-6789"
    assert clean_orcid("01234567-3456-6780") == None
    assert clean_orcid("0x23-4567-3456-6780") == None


def clean_str(thing: Optional[str], force_xml: bool = False) -> Optional[str]:
    """
    This function is appropriate to be called on any random, non-markup string,
    such as author names, titles, etc.

    It will try to clean up common unicode mangles, HTML characters, etc.

    This will detect XML/HTML and "do the right thing" (aka, not remove
    entities like '&amp' if there are tags in the string), unless you pass the
    'force_xml' parameter, which might be appropriate for, eg, names and
    titles, which generally should be projected down to plain text.

    Also strips extra whitespace.
    """
    if not thing:
        return None
    fix_entities = 'auto'
    if force_xml:
        fix_entities = True
    fixed = ftfy.fix_text(thing, fix_entities=fix_entities).strip()
    if not fixed or len(fixed) <= 1:
        # wasn't zero-length before, but is now; return None
        return None
    return fixed

def test_clean_str():

    assert clean_str(None) == None
    assert clean_str('') == None
    assert clean_str('1') == None
    assert clean_str('123') == '123'
    assert clean_str('a&amp;b') == 'a&b'
    assert clean_str('<b>a&amp;b</b>') == '<b>a&amp;b</b>'
    assert clean_str('<b>a&amp;b</b>', force_xml=True) == '<b>a&b</b>'

def b32_hex(s):
    s = s.strip().split()[0].lower()
    if s.startswith("sha1:"):
        s = s[5:]
    if len(s) != 32:
        return s
    return base64.b16encode(base64.b32decode(s.upper())).lower().decode('utf-8')

def is_cjk(s):
    if not s:
        return False
    for c in s:
        if c.isalpha():
            lang_prefix = unicodedata.name(c).split()[0]
            return lang_prefix in ('CJK', 'HIRAGANA', 'KATAKANA', 'HANGUL')
    return False

def test_is_cjk():
    assert is_cjk(None) is False
    assert is_cjk('') is False
    assert is_cjk('blah') is False
    assert is_cjk('岡, 鹿, 梨, 阜, 埼') is True
    assert is_cjk('[岡, 鹿, 梨, 阜, 埼]') is True
    assert is_cjk('菊') is True
    assert is_cjk('岡, 鹿, 梨, 阜, 埼 with eng after') is True
    assert is_cjk('水道') is True
    assert is_cjk('オウ, イク') is True # kanji
    assert is_cjk('ひヒ') is True
    assert is_cjk('き゚ゅ') is True
    assert is_cjk('ㄴ, ㄹ, ㅁ, ㅂ, ㅅ') is True

MONTH_MAP = {
    "jan":  1, "january":   1,
    "feb":  2, "febuary":   2,
    "mar":  3, "march":     3,
    "apr":  4, "april":     4,
    "may":  5, "may":       5,
    "jun":  6, "june":      6,
    "jul":  7, "july":      7,
    "aug":  8, "august":    8,
    "sep":  9, "september": 9,
    "oct": 10, "october":   10,
    "nov": 11, "nov":       11,
    "dec": 12, "december":  12,
}

def parse_month(raw: Optional[str]) -> Optional[int]:
    """
    Parses a string into a month number (1 to 12)
    """
    if not raw:
        return None
    raw = raw.strip().lower()
    if raw.isdigit():
        raw_int = int(raw)
        if raw_int >= 1 and raw_int <= 12:
            return raw_int
        else:
            return None
    if raw in MONTH_MAP:
        return MONTH_MAP[raw]
    return None

def test_parse_month() -> None:

    assert parse_month(None) == None
    assert parse_month("") == None
    assert parse_month("0") == None
    assert parse_month("10") == 10
    assert parse_month("jan") == 1
    assert parse_month("September") == 9

def detect_text_lang(raw: str) -> Optional[str]:
    """
    Tries to determine language of, eg, an abstract.

    Returns an ISO 631 2-char language code, or None.
    """
    if not raw:
        return None
    try:
        lang = langdetect.detect(raw)
        assert len(lang) == 2
        return lang
    except (langdetect.lang_detect_exception.LangDetectException, TypeError):
        return None
    return None

def test_detect_text_lang() -> None:
    assert detect_text_lang("") == None
    EN_SAMPLE = "this is a string of English text for testing"
    assert detect_text_lang(EN_SAMPLE) == "en"
    JA_SAMPLE = "モーラの種類は、以下に示すように111程度存在する。ただし、研究者により数え方が少しずつ異なる。"
    assert detect_text_lang(JA_SAMPLE) == "ja"

def parse_lang_name(raw: Optional[str]) -> Optional[str]:
    """
    Parses a language name and returns a 2-char ISO 631 language code.
    """
    if not raw:
        return None
    try:
        lang = pycountry.languages.lookup(raw)
        if lang.alpha_3 in ("mul", "mis"):
            return None
        return lang.alpha_2.lower()
    except LookupError:
        #print(f"  unknown language: '{raw}', file=sys.stderr)
        return None
    except AttributeError:
        #print(f"  partial language metadata: '{lang}', file=sys.stderr)
        return None
    return None

def test_parse_lang_name() -> None:

    assert parse_lang_name(None) == None
    assert parse_lang_name("") == None
    assert parse_lang_name("asdf ") == None
    assert parse_lang_name("english") == "en"
    assert parse_lang_name("ENGLISH") == "en"
    assert parse_lang_name("asdf blah") is None
    assert parse_lang_name("en") == "en"
    assert parse_lang_name("EN") == "en"
    assert parse_lang_name("ENG") == "en"
    assert parse_lang_name("English") == "en"
    assert parse_lang_name("Portuguese") == "pt"


def parse_country_name(s: Optional[str]) -> Optional[str]:
    """
    Parses a country name into a ISO country code (2-char).

    This version copied from the chocula repository.
    """
    if not s or s in ("Unknown"):
        return None

    s = s.strip()
    if s.lower() in ("usa", "new york (state)", "washington (state)"):
        return "us"
    if s.lower() in ("russia (federation)", "russia"):
        return "ru"
    if s == "Québec (Province)":
        s = "Canada"
    if s == "China (Republic : 1949- )":
        return "tw"
    if s == "Brunei":
        return "bn"
    if s.startswith("Congo "):
        s = "Congo"
    if s.lower() == "iran":
        return "ir"
    if s.lower() == "bermuda islands":
        return "bm"
    if s.lower() == "burma":
        s = "myanmar"
    if s.lower() in ("korea (south)", "south korea"):
        return "kr"
    if s.lower() in ("england", "scotland", "wales"):
        return "uk"
    s = s.replace(" (Republic)", "").replace(" (Federation)", "")

    try:
        country = pycountry.countries.lookup(s)
    except LookupError:
        country = None

    if country:
        return country.alpha_2.lower()
    try:
        sub = pycountry.subdivisions.lookup(s)
    except LookupError:
        sub = None

    s = s.replace(" (State)", "").replace(" (Province)", "")
    if sub:
        return sub.country_code.lower()

    else:
        # print(f"unknown country: {s}", file=sys.stderr)
        return None


def test_parse_country_name():
    assert parse_country_name("") is None
    assert parse_country_name("asdf blah") is None
    assert parse_country_name("us") == "us"
    assert parse_country_name("USA") == "us"
    assert parse_country_name("United States of America") == "us"
    assert parse_country_name("united States") == "us"
    assert parse_country_name("Massachusetts") == "us"
    assert parse_country_name("Russia") == "ru"
    assert parse_country_name("Japan") == "jp"

# These are very close, but maybe not exactly 1-to-1 with 639-2? Some mix of
# 2/T and 2/B?
# PubMed/MEDLINE and JSTOR use these MARC codes
# https://www.loc.gov/marc/languages/language_name.html
LANG_MAP_MARC = {
    'afr': 'af',
    'alb': 'sq',
    'amh': 'am',
    'ara': 'ar',
    'arm': 'hy',
    'aze': 'az',
    'ben': 'bn',
    'bos': 'bs',
    'bul': 'bg',
    'cat': 'ca',
    'chi': 'zh',
    'cze': 'cs',
    'dan': 'da',
    'dut': 'nl',
    'eng': 'en',
    'epo': 'eo',
    'est': 'et',
    'fin': 'fi',
    'fre': 'fr',
    'geo': 'ka',
    'ger': 'de',
    'gla': 'gd',
    'gre': 'el',
    'heb': 'he',
    'hin': 'hi',
    'hrv': 'hr',
    'hun': 'hu',
    'ice': 'is',
    'ind': 'id',
    'ita': 'it',
    'jpn': 'ja',
    'kin': 'rw',
    'kor': 'ko',
    'lat': 'la',
    'lav': 'lv',
    'lit': 'lt',
    'mac': 'mk',
    'mal': 'ml',
    'mao': 'mi',
    'may': 'ms',
    'nor': 'no',
    'per': 'fa',
    'per': 'fa',
    'pol': 'pl',
    'por': 'pt',
    'pus': 'ps',
    'rum': 'ro',
    'rus': 'ru',
    'san': 'sa',
    'slo': 'sk',
    'slv': 'sl',
    'spa': 'es',
    'srp': 'sr',
    'swe': 'sv',
    'tha': 'th',
    'tur': 'tr',
    'ukr': 'uk',
    'urd': 'ur',
    'vie': 'vi',
    'wel': 'cy',

# additions
    'gle': 'ga', # "Irish" (Gaelic)
    'jav': 'jv', # Javanese
    'welsh': 'cy', # Welsh
    'oci': 'oc', # Occitan

# Don't have ISO 639-1 codes
    'grc': 'el', # Ancient Greek; map to modern greek
    'map': None, # Austronesian (collection)
    'syr': None, # Syriac, Modern
    'gem': None, # Old Saxon
    'non': None, # Old Norse
    'emg': None, # Eastern Meohang
    'neg': None, # Negidal
    'mul': None, # Multiple languages
    'und': None, # Undetermined
}