aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/importers/crossref.py
blob: d0a69cd62094f13f0256d1e8db0f157d2ce87e6c (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

import sys
import json
import sqlite3
import datetime
import itertools
import fatcat_client
from .common import FatcatImporter


# The docs/guide should be the cannonical home for these mappings; update there
# first
CROSSREF_TYPE_MAP = {
    'book': 'book',
    'book-chapter': 'chapter',
    'book-part': 'chapter',
    'book-section': 'chapter',
    'component': None,
    'dataset': 'dataset',
    'dissertation': 'thesis',
    'edited-book': 'book',
    'journal-article': 'article-journal',
    'monograph': 'monograph',
    'other': None,
    'peer-review': 'peer_review',
    'posted-content': 'post',
    'proceedings-article': 'paper-conference',
    'reference-book': 'book',
    'reference-entry': 'entry',
    'report': 'report',
    'standard': 'standard',
}

class CrossrefImporter(FatcatImporter):
    """
    Importer for Crossref metadata.

    Can use a local sqlite3 file for faster "external identifier" lookups

    See https://github.com/CrossRef/rest-api-doc for JSON schema notes
    """

    def __init__(self, host_url, issn_map_file, extid_map_file=None, create_containers=True):
        super().__init__(host_url, issn_map_file)
        self.extid_map_db = None
        if extid_map_file:
            db_uri = "file:{}?mode=ro".format(extid_map_file)
            print("Using external ID map: {}".format(db_uri))
            self.extid_map_db = sqlite3.connect(db_uri, uri=True)
        else:
            print("Not using external ID map")
        self.create_containers = create_containers

    def lookup_ext_ids(self, doi):
        if self.extid_map_db is None:
            return dict(core_id=None, pmid=None, pmcid=None, wikidata_qid=None)
        row = self.extid_map_db.execute("SELECT core, pmid, pmcid, wikidata FROM ids WHERE doi=? LIMIT 1",
            [doi.lower()]).fetchone()
        if row is None:
            return dict(core_id=None, pmid=None, pmcid=None, wikidata_qid=None)
        row = [str(cell or '') or None for cell in row]
        return dict(
            core_id=row[0],
            pmid=row[1],
            pmcid=row[2],
            wikidata_qid=row[3])

    def map_release_type(self, crossref_type):
        return CROSSREF_TYPE_MAP.get(crossref_type)

    def parse_crossref_dict(self, obj):
        """
        obj is a python dict (parsed from json).
        returns a ReleaseEntity
        """

        # This work is out of scope if it doesn't have authors and a title
        if (not 'author' in obj) or (not 'title' in obj):
            return None

        # Other ways to be out of scope (provisionally)
        # journal-issue and journal-volume map to None, but allowed for now
        if obj.get('type') in (None, 'journal', 'proceedings',
                'standard-series', 'report-series', 'book-series', 'book-set',
                'book-track', 'proceedings-series'):
            return None

        # contribs
        def do_contribs(obj_list, ctype):
            contribs = []
            for i, am in enumerate(obj_list):
                creator_id = None
                if 'ORCID' in am.keys():
                    creator_id = self.lookup_orcid(am['ORCID'].split('/')[-1])
                # Sorry humans :(
                if am.get('given') and am.get('family'):
                    raw_name = "{} {}".format(am['given'], am['family'])
                elif am.get('family'):
                    raw_name = am['family']
                else:
                    # TODO: defaults back to a pseudo-null value
                    raw_name = am.get('given', '<blank>')
                extra = dict()
                if ctype == "author":
                    index = i
                else:
                    index = None
                if am.get('affiliation'):
                    # note: affiliation => affiliations
                    extra['affiliations'] = am.get('affiliation')
                if am.get('sequence') and am.get('sequence') != "additional":
                    extra['sequence'] = am.get('sequence')
                if not extra:
                    extra = None
                assert ctype in ("author", "editor", "translator")
                contribs.append(fatcat_client.ReleaseContrib(
                    creator_id=creator_id,
                    index=index,
                    raw_name=raw_name,
                    role=ctype,
                    extra=extra))
            return contribs
        contribs = do_contribs(obj['author'], "author")
        contribs.extend(do_contribs(obj.get('editor', []), "editor"))
        contribs.extend(do_contribs(obj.get('translator', []), "translator"))

        # container
        issn = obj.get('ISSN', [None])[0]
        issnl = self.issn2issnl(issn)
        container_id = None
        if issnl:
            container_id = self.lookup_issnl(issnl)
        publisher = obj.get('publisher')

        ce = None
        if (container_id is None and self.create_containers and (issnl is not None)
            and obj.get('container-title') and len(obj['container-title']) > 0):
            ce = fatcat_client.ContainerEntity(
                issnl=issnl,
                publisher=publisher,
                name=obj['container-title'][0])

        # references
        refs = []
        for i, rm in enumerate(obj.get('reference', [])):
            try:
                year = int(rm.get('year'))
                # NOTE: will need to update/config in the future!
                # NOTE: are there crossref works with year < 100?
                if year > 2025 or year < 100:
                    year = None
            except:
                year = None
            extra = rm.copy()
            if rm.get('DOI'):
                extra['doi'] = rm.get('DOI').lower()
            key = rm.get('key')
            if key and key.startswith(obj['DOI'].upper()):
                key = key.replace(obj['DOI'].upper() + "-", '')
                key = key.replace(obj['DOI'].upper(), '')
            container_name = rm.get('volume-title')
            if not container_name:
                container_name = rm.get('journal-title')
            extra.pop('DOI', None)
            extra.pop('key', None)
            extra.pop('year', None)
            extra.pop('volume-name', None)
            extra.pop('journal-title', None)
            extra.pop('title', None)
            extra.pop('first-page', None)
            extra.pop('doi-asserted-by', None)
            if extra:
                extra = dict(crossref=extra)
            else:
                extra = None
            refs.append(fatcat_client.ReleaseRef(
                index=i,
                # doing lookups would be a second import pass
                target_release_id=None,
                key=key,
                year=year,
                container_name=container_name,
                title=rm.get('title'),
                locator=rm.get('first-page'),
                # TODO: just dump JSON somewhere here?
                extra=extra))

        # abstracts
        abstracts = []
        if obj.get('abstract') != None:
            abstracts.append(fatcat_client.ReleaseEntityAbstracts(
                mimetype="application/xml+jats",
                content=obj.get('abstract')))

        # extra fields
        extra = dict()
        for key in ('subject', 'type', 'license', 'alternative-id',
                'container-title', 'original-title', 'subtitle', 'archive',
                'funder', 'group-title'):
            # TODO: unpack "container-title" array
            val = obj.get(key)
            if val:
                extra[key] = val
        if 'license' in extra and extra['license']:
            for i in range(len(extra['license'])):
                if 'start' in extra['license'][i]:
                    extra['license'][i]['start'] = extra['license'][i]['start']['date-time']
        if len(obj['title']) > 1:
            extra['other-titles'] = obj['title'][1:]
        # TODO: this should be top-level
        extra['is_kept'] = len(obj.get('archive', [])) > 0

        # ISBN
        isbn13 = None
        for raw in obj.get('ISBN', []):
            # TODO: convert if not ISBN-13 format
            if len(raw) == 17:
                isbn13 = raw
                break

        # release status
        if obj['type'] in ('journal-article', 'conference-proceeding', 'book',
                'dissertation', 'book-chapter'):
            release_status = "published"
        else:
            # unknown
            release_status = None

        # external identifiers
        extids = self.lookup_ext_ids(doi=obj['DOI'].lower())

        # TODO: filter out huge releases; we'll get them later (and fix bug in
        # fatcatd)
        if max(len(contribs), len(refs), len(abstracts)) > 750:
            return None

        # release date parsing is amazingly complex
        release_date = obj['issued']['date-parts'][0]
        if not release_date or not release_date[0]:
            # got some NoneType, even though at least year is supposed to be set
            release_date = None
        elif len(release_date) == 3:
            release_date = datetime.date(year=release_date[0], month=release_date[1], day=release_date[2])
        else:
            # only the year is actually required; mangle to first day for date
            # (TODO: something better?)
            release_date = datetime.date(year=release_date[0], month=1, day=1)

        re = fatcat_client.ReleaseEntity(
            work_id=None,
            title=obj['title'][0],
            contribs=contribs,
            refs=refs,
            container_id=container_id,
            publisher=publisher,
            release_type=self.map_release_type(obj['type']),
            release_status=release_status,
            doi=obj['DOI'].lower(),
            isbn13=isbn13,
            core_id=extids['core_id'],
            pmid=extids['pmid'],
            pmcid=extids['pmcid'],
            wikidata_qid=extids['wikidata_qid'],
            release_date=release_date,
            issue=obj.get('issue'),
            volume=obj.get('volume'),
            pages=obj.get('page'),
            abstracts=abstracts,
            extra=dict(crossref=extra))
        return (re, ce)

    def create_row(self, row, editgroup=None):
        if row is None:
            return
        obj = json.loads(row)
        entities = self.parse_crossref_dict(obj)
        if entities is not None:
            (re, ce) = entities
            if ce is not None:
                container = self.api.create_container(ce, editgroup=editgroup)
                re.container_id = container.ident
                self._issnl_id_map[ce.issnl] = container.ident
            self.api.create_release(re, editgroup=editgroup)
            self.counts['insert'] += 1

    def create_batch(self, batch, editgroup=None):
        """Current work/release pairing disallows batch creation of releases.
        Could do batch work creation and then match against releases, but meh."""
        release_batch = []
        for row in batch:
            if row is None:
                continue
            obj = json.loads(row)
            entities = self.parse_crossref_dict(obj)
            if entities is not None:
                (re, ce) = entities
                if ce is not None:
                    ce_eg = self.api.create_editgroup(
                        fatcat_client.Editgroup(editor_id='aaaaaaaaaaaabkvkaaaaaaaaae'))
                    container = self.api.create_container(ce, editgroup=ce_eg.id)
                    self.api.accept_editgroup(ce_eg.id)
                    re.container_id = container.ident
                    self._issnl_id_map[ce.issnl] = container.ident
                release_batch.append(re)
        self.api.create_release_batch(release_batch, autoaccept="true", editgroup=editgroup)
        self.counts['insert'] += len(release_batch)