-
Notifications
You must be signed in to change notification settings - Fork 0
/
Copy pathcldfbench_elcat.py
392 lines (365 loc) · 18.2 KB
/
cldfbench_elcat.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
import re
import json
import hashlib
import pathlib
import subprocess
import collections
import urllib.parse
import urllib.request
import requests
from cldfbench import Dataset as BaseDataset
from cldfbench import CLDFSpec
from clldutils.misc import nfilter
from csvw.metadata import URITemplate
from util import (
iter_langs, SCORED_PARAMETERS, split, bibliography, get_doc, COMPOSITE_PARAMETERS,
norm_text, norm_coords, Bib, norm_country, INVALID_LANGUAGE_IDS,
)
class Dataset(BaseDataset):
dir = pathlib.Path(__file__).parent
id = "elcat"
def cldf_specs(self):
return CLDFSpec(module='StructureDataset', dir=self.cldf_dir)
def get(self, url, force=False):
if url.startswith('/'):
url = 'https://endangeredlanguages.com' + url
parsed = urllib.parse.urlparse(url)
p = self.raw_dir / 'html' / parsed.path[1:].replace('/', '_')
if force or (not p.exists()):
p.write_text(requests.get(url).text, encoding='utf8')
print(p)
return get_doc(p)
def cmd_download(self, args):
urllib.request.urlretrieve(
"https://endangeredlanguages.com/userquery/download/",
self.raw_dir / 'languages.csv')
for row in self.raw_dir.read_csv('languages.csv'):
if row[0] not in INVALID_LANGUAGE_IDS:
self.get('/lang/{}'.format(row[0]))
self.get('/lang/{}/bibliography'.format(row[0]))
def cmd_makecldf(self, args):
self.schema(args)
# Add parameter descriptions:
for pname, levels in SCORED_PARAMETERS.items():
if isinstance(pname, tuple):
pname = pname[1]
pid = pname.lower().replace(' ', '_')
args.writer.objects['ParameterTable'].append(dict(ID=pid, Name=pname))
for score, level in enumerate(levels):
args.writer.objects['CodeTable'].append(dict(
ID='{}-{}'.format(pid, score),
Parameter_ID=pid,
Name=SCORED_PARAMETERS['Score'][score],
Description=level,
))
args.writer.objects['ParameterTable'].append(dict(
ID='LEI',
Name='Language Endangerment Index',
Description='Language Endangerment Index computed per source. '
'The level of endangerment presented for each language is not meant to be '
'the final word on the matter. The scores for individual languages will '
'change as more information becomes available. These scores are provided '
'for practical purposes, to give a quick but rough visual indication of a '
'language’s endangerment status. The level of certainty accompanying each '
'endangerment score shows the degree of confidence in that score; a label '
'of “uncertain” may indicate that the level is not yet known, or that the '
'score has been computed but further evaluation is needed. Detailed '
'information about how a language’s level of endangerment is calculated is '
'given at https://endangeredlanguages.com/about_catalogue/',
))
for pid, desc, schema in COMPOSITE_PARAMETERS:
args.writer.objects['ParameterTable'].append(dict(
ID=pid,
Name=pid,
Description=desc,
ColumnSpec=dict(datatype=dict(base='json', format=json.dumps(schema))),
))
args.writer.objects['ParameterTable'].append(dict(
ID='bib',
Name='Bibliography',
Description='Other sources about a language.',
))
# Add languages:
coords, iso2gc = {}, {}
for l in args.glottolog.api.languoids():
if l.iso:
iso2gc[l.iso] = l.id
if l.latitude is not None:
coords[l.id] = (l.latitude, l.longitude)
# Read hand-curated, explicit mappings between ElCat and Glottolog (dialects):
id2gc = {
r['ID']: r['Glottocode'] for r in self.etc_dir.read_csv('languages.csv', dicts=True)}
cols = {9: 'Comment', 10: 'Countries', 11: 'ELCatMacroareas'}
lmeta = {}
countries = {}
for row in self.raw_dir.read_csv('languages.csv'):
lmeta[row[0]] = {n: split(row[i]) if i in {11, 10} else row[i] for i, n in cols.items()}
lmeta[row[0]]['Countries'] = [norm_country(c, countries) for c in lmeta[row[0]]['Countries']]
for c in sorted(countries.values(), key=lambda i: i['name']):
args.writer.objects['countries.csv'].append(c)
sources = []
for obj in iter_langs(self.raw_dir / 'html'):
if obj is None:
continue
if isinstance(obj, Bib):
sources.append(obj)
continue
lang = obj
# Determine a language's ISO and Glottocode prefering
# 1. our hand-crated mappings over
# 2. explicit ElCat mappings, via "code_authority"
# 3. derived mappings from ElCat mappings to ISO 639-3.
iso_codes, glottocodes = [], []
for code in lang.metadata.language_code:
code = code.replace('\ufeff', '')
if len(code) == 3 and (
('ISO 639-3' in lang.metadata.code_authority) or
(not lang.metadata.code_authority)):
iso_codes.append(code)
if len(code) == 8 and ('Glottolog' in lang.metadata.code_authority):
glottocodes.append(code)
if iso_codes and not glottocodes: # Map ISO to Glottocode.
glottocodes = [iso2gc[iso] for iso in iso_codes if iso in iso2gc]
if lang.id in id2gc: # Hand-curated mapping exists.
if glottocodes:
if [id2gc[lang.id]] != glottocodes:
assert (lang.id in {'2896', '6029'}) or id2gc[lang.id] in glottocodes, str(lang.id) # This is the only known discrepancy.
glottocodes = [id2gc[lang.id]]
else:
glottocodes = [id2gc[lang.id]]
args.writer.objects['LanguageTable'].append(dict(
ID=lang.id,
Name=lang.name.strip(),
Glottocode=glottocodes[0] if len(glottocodes) == 1 else None,
ISO639P3code=iso_codes[0] if len(iso_codes) == 1 else None,
classification=lang.classification,
endangerment=lang.endangerment,
code_authorities=lang.metadata.code_authority,
codes=lang.metadata.language_code,
alt_names=nfilter(lang.metadata.also_known_as),
**lmeta[lang.id],
))
# Add values:
for param, _, _ in COMPOSITE_PARAMETERS:
for i, (sid, d) in enumerate(getattr(lang, param), start=1):
comment = norm_text(d.pop('Public Comment', '')) or None
if 'Endangerment Level' in d:
if sid == lang.preferred_source:
if lang.endangerment and (lang.endangerment != d['Endangerment Level'].split('(')[0].strip().lower()):
# 29 cases. What to do here?
args.log.warning('{}: {} vs {}'.format(
lang.id, lang.endangerment, d['Endangerment Level'].split('(')[0].strip().lower()))
args.writer.objects['ValueTable'].append(dict(
ID='{}-LEI-{}'.format(lang.id, i),
Language_ID=lang.id,
Parameter_ID='LEI',
Value=d['Endangerment Level'],
Source=[sid] if sid else [],
Comment=comment,
preferred=sid == lang.preferred_source,
))
level, _, certainty = d.pop('Endangerment Level').partition('(')
d['Endangerment'] = dict(Level=level.strip().lower())
m = re.match('([0-9]+)', certainty)
if m:
d['Endangerment']['Certainty'] = float(m.groups()[0]) / 100
if "Places" in d:
d["Places"] = split(d["Places"])
for key in [
'Description',
'Speaker Number Text',
'Speaker Attitude',
'Second Language Speakers',
'Number Speaker Other Languages'
]:
if key in d:
d[key] = norm_text(d[key])
if "Domains Other Langs" in d:
d["Domains Other Langs"] = [norm_text(x) for x in split(d["Domains Other Langs"])]
if 'Coordinates' in d:
d['Coordinates'] = norm_coords(d['Coordinates'])
if not d['Coordinates']:
del d['Coordinates']
else: # Pick the first coordinate as representative location of the language.
coords[lang.id] = d['Coordinates'][0]
for k, v in d.items():
levels = SCORED_PARAMETERS.get((param, k))
if levels: # A value for a scored parameter.
if k == 'Speaker Number': # Speaker Numbers are not given as 10,11,...
if v in levels:
score = levels.index(v)
else:
assert v == 'Awakening', 'Unknown Speaker Number value: {}'.format(v)
continue
else: # Scores given as 10, 11, 12, 13, 14, 15.
score = int(v) - 10
args.writer.objects['ValueTable'].append(dict(
ID='{}-{}-{}'.format(lang.id, k.lower().replace(' ', '_'), i),
Language_ID=lang.id,
Parameter_ID=k.lower().replace(' ', '_'),
Value=score,
Code_ID='{}-{}'.format(k.lower().replace(' ', '_'), score),
Source=[sid] if sid else [],
Comment=comment,
preferred=sid == lang.preferred_source,
))
for key in ['Speaker Number Trends', 'Transmission', 'Domains Of Use']:
if key in d:
# Translate 10,11,12,... notation to score descriptions for the
# composite values.
try:
score = int(d[key]) - 10
d[key] = SCORED_PARAMETERS[('vitality', key)][score]
except ValueError:
del d[key]
if d or comment:
args.writer.objects['ValueTable'].append(dict(
ID='{}-{}-{}'.format(lang.id, param, i),
Language_ID=lang.id,
Parameter_ID=param,
Value=json.dumps(d),
Source=[sid] if sid else [],
Comment=comment,
preferred=sid == lang.preferred_source,
))
for src in bibliography(self.raw_dir / 'html'):
if src:
sources.append(src)
lid2sids = collections.defaultdict(set)
local_id2sid, sids = {}, set()
for i, m in enumerate(Bib.iter_merged(sources), start=1):
src = m.as_source(str(i))
sid = hashlib.md5((src.get('local_ids') or src.bibtex()).encode('utf8')).hexdigest()
src.id = sid
assert sid not in sids
sids.add(sid)
args.writer.cldf.sources.add(m.as_source(sid))
assert m.local_ids or m.language_ids, str(src.bibtex())
for local_id in m.local_ids:
local_id2sid[local_id] = sid
for lid in m.language_ids:
lid2sids[lid].add(sid)
for val in args.writer.objects['ValueTable']:
val['Source'] = [local_id2sid[sid] for sid in val['Source']]
for lid, sources in lid2sids.items():
args.writer.objects['ValueTable'].append(dict(
ID='{}-bib'.format(lid),
Language_ID=lid,
Parameter_ID='bib',
Value='See Source',
Source=list(sorted(sources)),
))
# Now we can assign locations to the languages, preferentially from ElCat data, as a
# fallback from Glottolog data.
for lang in args.writer.objects['LanguageTable']:
if lang['ID'] in coords:
lang['Latitude'], lang['Longitude'] = coords[lang['ID']]
elif lang['Glottocode'] in coords:
lang['Latitude'], lang['Longitude'] = coords[lang['Glottocode']]
def schema(self, args):
args.writer.cldf.add_component(
'LanguageTable',
{
'name': 'Comment',
'propertyUrl': 'http://cldf.clld.org/v1.0/terms.rdf#comment',
},
{
'name': 'Countries',
'separator': ' ',
'dc:description':
'Countries a language is spoken in given by ISO 3166-1 alpha-2 code',
},
{
'name': 'ELCatMacroareas',
'separator': '; ',
},
{
'name': 'classification',
'dc:description': 'Top-level genealogical unit the language belongs to.',
},
{
'name': 'endangerment',
'dc:description':
"ElCat's aggregated endangerment assessment. Note that in a few cases this "
"endangerment assessment does **not** match the assessment in the preferred "
"source as given for parameter LEI in the ValueTable.",
'datatype': {
'base': 'string',
'format': 'at risk|awakening|critically endangered|dormant|endangered|endangerment|severely endangered|threatened|vulnerable'},
},
{
'name': 'code_authorities',
'dc:description': 'Other language catalogs which have assigned codes to the language.',
'separator': '; ',
'datatype': {'base': 'string', 'format': 'ISO 639-3|Glottolog|LINGUIST List'},
},
{
'name': 'codes',
'dc:description': 'Codes assigned to the language by other language catalogs.',
'separator': '; ',
},
{
'name': 'alt_names',
'dc:description': 'Alternative names used for the language.',
'separator': '; '},
)
args.writer.cldf['LanguageTable', 'ID'].valueUrl = URITemplate(
'https://endangeredlanguages.com/lang/{ID}')
args.writer.cldf.add_component('ParameterTable')
args.writer.cldf.add_component('CodeTable')
args.writer.cldf.add_columns(
'ValueTable',
{
'name': 'preferred',
'dc:description':
'If ElCat contains multiple pieces of information pertaining to the same '
'aspects of a language, the one used to assess the overall endangerment index '
'is marked as *preferred* (based on its source).',
'datatype': {'base': 'boolean', 'format': 'yes|no'}}
)
args.writer.cldf.add_table(
'countries.csv',
{
'name': 'alpha_2',
'propertyUrl': 'http://cldf.clld.org/v1.0/terms.rdf#id',
},
'alpha_3',
{
'name': 'name',
'propertyUrl': 'http://cldf.clld.org/v1.0/terms.rdf#name',
},
'official_name',
)
args.writer.cldf.add_foreign_key('LanguageTable', 'Countries', 'countries.csv', 'alpha_2')
def cmd_readme(self, args):
subprocess.check_call([
'cldfbench',
'cldfviz.map',
str(self.cldf_specs().metadata_path),
'--language-properties', 'endangerment',
'--language-properties-colormaps',
'{"at risk":"#FFF7BC","threatened":"#FEE391","vulnerable":"#FEC44F","endangered":"#FB9A29","severely endangered":"#EC7014","critically endangered":"#CC4C02","awakening":"#993404","dormant":"#662506"}',
'--output', str(self.dir / 'map.png'),
'--width', '20',
'--height', '10',
'--format', 'png',
'--with-ocean',
'--pacific-centered'])
desc = [
'\n\n\n### Parameters\n',
'This dataset contains three sets of [parameters](cldf/parameters.csv).',
'- The four categories from which the Language Endangerment Index is derived.',
'- The computed Language Endangerment Index.',
'- Parameters with composite JSON values, aggregating information from a specific '
'source on a particular topic. Below is a list of '
'[JSON schemas](https://json-schema.org/) describing the values of these parameters.\n'
]
for param in self.cldf_reader()['ParameterTable']:
if param['ColumnSpec']:
desc.append('#### {}\n'.format(param['Name']))
desc.append('```json')
schema = json.loads(param['ColumnSpec']['datatype']['format'])
desc.extend(json.dumps(schema, indent=4).split('\n'))
desc.append('```\n')
pre, head, post = super().cmd_readme(args).partition('## CLDF ')
return pre + '\n'.join(desc) + head + post