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
|
#!/usr/bin/env python3
"""
Takes *enriched* JSON objects which include fatcat metadata and fulltext
content, and outputs JSON lines in fatcat_fulltext schema.
"""
import sys
import json
import argparse
import datetime
from fatcat_covid19.common import *
def fulltext_to_elasticsearch(row, force_bool=True):
"""
Converts from fulltext content and release model/schema to elasticsearch
oriented schema.
Returns: dict
Raises exception on error (never returns None)
"""
if not 'fatcat_release' in row:
# skip papers that don't match to a fatcat release
return None
release = row['fatcat_release']
abstracts = []
abstract_langs = []
# first, easy fatcat metadata
t = {
'fatcat_ident': release['ident'],
'fatcat_revision': release['revision'],
'fulltext': dict(),
}
BIBLIO_KEYS = [
'work_id',
'title',
'subtitle',
'original_title',
'release_type',
'release_stage',
'release_year',
'release_date',
'withdrawn_status',
'language',
'volume',
'issue',
'pages',
'number',
'license',
]
EXT_IDS = [
'doi',
'pmid',
'pmcid',
'isbn13',
'wikidata_qid',
'arxiv_id',
'jstor_id',
'mag_id',
]
for key in BIBLIO_KEYS:
t[key] = release.get(key) or None
for key in EXT_IDS:
t[key] = release['ext_ids'].get(key) or None
t['contrib_count'] = len(release['contribs'] or [])
if release.get('abstracts'):
for a in release['abstracts']:
abstracts.append(a['content'])
abstract_langs.append(a['lang'])
contrib_names = []
contrib_affiliations = []
creator_ids = []
for c in (release['contribs'] or []):
if c.get('raw_name'):
contrib_names.append(c['raw_name'])
elif c.get('surname'):
contrib_names.append(c['surname'])
if c.get('creator_id'):
creator_ids.append(c['creator_id'])
if c.get('raw_affiliation'):
contrib_affiliations.append(c['raw_affiliation'])
t['contrib_names'] = contrib_names
t['creator_ids'] = creator_ids
t['affiliations'] = contrib_affiliations
container = release.get('container')
if container:
t['publisher'] = container.get('publisher')
t['container_name'] = container.get('name')
t['container_original_name'] = container.get('original_name')
# this is container.ident, not release.container_id, because there may
# be a redirect involved
t['container_id'] = container['ident']
t['container_issnl'] = container.get('issnl')
t['container_type'] = container.get('container_type')
if container.get('extra'):
c_extra = container['extra']
if c_extra.get('country'):
t['country_code'] = c_extra['country']
t['country_code_upper'] = c_extra['country'].upper()
# fall back to release-level container metadata if container not linked or
# missing context
if not t.get('publisher'):
t['publisher'] = release.get('publisher')
if not t.get('container_name') and release.get('extra'):
t['container_name'] = release['extra'].get('container_name')
extra = release['extra'] or dict()
if extra:
if not t.get('container_name'):
t['container_name'] = extra.get('container_name')
# backwards compatible subtitle fetching
if not t['subtitle'] and extra.get('subtitle'):
if type(extra['subtitle']) == list:
t['subtitle'] = extra['subtitle'][0]
else:
t['subtitle'] = extra['subtitle']
t['first_page'] = None
if release.get('pages'):
first = release['pages'].split('-')[0]
first = first.replace('p', '')
if first.isdigit():
t['first_page'] = first
# TODO: non-numerical first pages
t['doi_registrar'] = None
if extra and t['doi']:
for k in ('crossref', 'datacite', 'jalc'):
if k in extra:
t['doi_registrar'] = k
if not 'doi_registrar' in t:
t['doi_registrar'] = 'crossref'
if t['doi']:
t['doi_prefix'] = t['doi'].split('/')[0]
# then the fulltext stuff
t['fulltext']['status'] = row.get('fulltext_status', 'none')
if 'fulltext_file' in row:
full = row['fulltext_file']
t['fulltext']['sha1'] = full['sha1']
t['fulltext']['pdf_url'] = "/" + full['pdf_path']
if full.get('pdftotext_path'):
t['fulltext']['pdftotext_url'] = "/" + full['pdftotext_path']
if full.get('thumbnail_path'):
t['fulltext']['thumbnail_url'] = "/" + full['thumbnail_path']
if full.get('grobid_xml_path'):
t['fulltext']['grobid_xml_url'] = "/" + full['grobid_xml_path']
if 'fulltext_grobid' in row:
grobid = row['fulltext_grobid']
if grobid.get('abstract'):
abstracts.append(grobid['abstract'])
abstract_langs.append(grobid['language_code'])
t['fulltext']['abstract'] = grobid.get('abstract', None)
t['fulltext']['body'] = grobid.get('body', None)
t['fulltext']['acknowledgement'] = grobid.get('acknowledgement', None)
t['fulltext']['annex'] = grobid.get('annex', None)
t['fulltext']['lang'] = grobid.get('language_code', None)
elif 'fulltext_pdftotext' in row:
pdftotext = row['fulltext_pdftotext']
t['fulltext']['body'] = pdftotext.get('body', None)
# then other metadata stuff
if row.get('source_tags'):
# will get set-uniq at the end
t['source_tags'] = row['source_tags']
else:
t['source_tags'] = []
if 'cord19_paper' in row:
t['source_tags'].append('cord19')
paper = row['cord19_paper']
t['cord19_uid'] = paper['cord_uid']
if paper.get('who_covidence_id'):
t['who_covidence_id'] = paper['who_covidence_id']
t['source_tags'].append('who')
if paper.get('abstract') and not abstracts:
abstracts.append(paper['abstract'])
if not t['license']:
t['license'] = paper.get('license') or None
t['abstract'] = abstracts
t['abstract_lang'] = list(set(abstract_langs))
t['source_tags'] = list(set(t['source_tags']))
return t
def run(args):
for l in args.json_file:
l = json.loads(l)
result = fulltext_to_elasticsearch(l, args)
if result:
print(json.dumps(result, sort_keys=True))
def main():
parser = argparse.ArgumentParser(
formatter_class=argparse.ArgumentDefaultsHelpFormatter)
parser.add_argument('json_file',
help="fulltext content input",
type=argparse.FileType('r'))
subparsers = parser.add_subparsers()
args = parser.parse_args()
args.session = requests_retry_session()
run(args)
if __name__ == '__main__':
main()
|
