aboutsummaryrefslogtreecommitdiffstats
path: root/fatcat_covid19/transform.py
blob: 1002c105bf23761ebdf2d35d52eaf6d139c496b0 (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

import sys
import json
import argparse
import datetime

from fatcat_covid19.common import *


UNWANTED_ABSTRACT_PREFIXES = [
    # roughly sort this long to short
    'Abstract No Abstract ',
    'Publisher Summary ',
    'Abstract ',
    'ABSTRACT ',
    'Summary ',
    'Background: ',
    'Background ',
]

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']:

            # hack to (partially) clean up common JATS abstract display case
            if a.get('mimetype') == 'application/xml+jats':
                for tag in ('p', 'jats', 'jats:p', 'jats:title'):
                    a['content'] = a['content'].replace('<{}>'.format(tag), '')
                    a['content'] = a['content'].replace('</{}>'.format(tag), '')
                    # ugh, double encoding happens
                    a['content'] = a['content'].replace('&lt;/{}&gt;'.format(tag), '')
                    a['content'] = a['content'].replace('&lt;{}&gt;'.format(tag), '')

            # hack to remove abstract prefixes
            for prefix in UNWANTED_ABSTRACT_PREFIXES:
                if a['content'].startswith(prefix):
                    a['content'] = a['content'][len(prefix):]
                    break
            a['content'] = a['content'].strip()
            if a['content']:
                abstracts.append(a['content'].strip())
                if a.get('lang'):
                    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]

    # special-case medrxiv/biorxiv content
    if not t.get('release_stage') and not t.get('container_name') and t.get('doi', '').startswith('10.1101/20'):
        t['container_name'] = 'biorXiv/medrXiv'
        t['release_stage'] = 'draft'
        if t.get('release_type') in ['post', None]:
            t['release_type'] = 'article-journal'

    # special-case arxiv
    if not t.get('container_name') and t.get('arxiv_id'):
        t['container_name'] = 'arXiv'

    # 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']
        for url in full.get('urls', []):
            if url.get('rel') in ('webarchive', 'archive') and 'archive.org/' in url['url']:
                t['fulltext']['ia_pdf_url'] = url['url']
                break

    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 transform_es_file(json_input, json_output):
    """
    Takes *enriched* JSON objects which include fatcat metadata and fulltext
    content, and outputs JSON lines in fatcat_fulltext schema.
    """
    for l in json_input:
        l = json.loads(l)
        result = fulltext_to_elasticsearch(l)
        if result:
            print(json.dumps(result, sort_keys=True), file=json_output)