aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/transforms/elasticsearch.py
blob: a8e0227b0d5639fdc3a7844a56b48b662d55cb92 (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


import collections
from fatcat_client import ApiClient


def check_kbart(year, archive):
    if not archive or not archive.get('year_spans'):
        return None
    for span in archive['year_spans']:
        if year >= span[0] and year <= span[1]:
            return True
    return False

def test_check_kbart():

    assert check_kbart(1990, dict()) == None
    assert check_kbart(1990, dict(year_spans=[[2000, 2000]])) == False
    assert check_kbart(2000, dict(year_spans=[[2000, 2000]])) == True
    assert check_kbart(1950, dict(year_spans=[[1900, 1920], [1990, 2000]])) == False
    assert check_kbart(1950, dict(year_spans=[[1900, 1920], [1930, 2000]])) == True

def release_to_elasticsearch(entity, force_bool=True):
    """
    Converts from an entity model/schema to elasticsearch oriented schema.

    Returns: dict
    Raises exception on error (never returns None)
    """

    if entity.state in ('redirect', 'deleted'):
        return dict(
            ident = entity.ident,
            state = entity.state,
        )
    elif entity.state != 'active':
        raise ValueError("Unhandled entity state: {}".format(entity.state))

    # First, the easy ones (direct copy)
    release = entity
    t = dict(
        ident = release.ident,
        state = release.state,
        revision = release.revision,
        title = release.title,
        subtitle = release.subtitle,
        original_title = release.original_title,
        release_type = release.release_type,
        release_stage = release.release_stage,
        withdrawn_status = release.withdrawn_status,
        language = release.language,
        license = release.license_slug,
        doi = release.ext_ids.doi,
        pmid = release.ext_ids.pmid,
        pmcid = release.ext_ids.pmcid,
        isbn13 = release.ext_ids.isbn13,
        wikidata_qid = release.ext_ids.wikidata_qid,
        core_id = release.ext_ids.core,
        arxiv_id = release.ext_ids.arxiv,
        jstor_id = release.ext_ids.jstor,
        ark_id = release.ext_ids.ark,
        mag_id = release.ext_ids.mag,
    )

    is_oa = None
    is_preserved = None
    is_longtail_oa = None
    in_kbart = None
    in_jstor = False
    in_web = False
    in_dweb = False
    in_ia = False
    in_ia_sim = False
    in_shadow = False

    release_year = release.release_year
    if release.release_date:
        # .isoformat() results in, eg, '2010-10-22' (YYYY-MM-DD)
        t['release_date'] = release.release_date.isoformat()
        if not release_year:
            release_year = release.release_date.year
    if release_year:
        t['release_year'] = release_year

    t['any_abstract'] = len(release.abstracts or []) > 0
    t['ref_count'] = len(release.refs or [])
    t['contrib_count'] = len(release.contribs or [])
    contrib_names = []
    for c in (release.contribs or []):
        if c.raw_name:
            contrib_names.append(c.raw_name)
    t['contrib_names'] = contrib_names

    container = release.container
    if container:
        t['publisher'] = container.publisher
        t['container_name'] = container.name
        t['container_id'] = container.ident
        t['container_issnl'] = container.issnl
        t['container_type'] = container.container_type
        if container.extra:
            c_extra = container.extra
            if c_extra.get('kbart') and release_year:
                in_jstor = check_kbart(release_year, c_extra['kbart'].get('jstor'))
                in_kbart = in_jstor
                for archive in ('portico', 'lockss', 'clockss'):
                    in_kbart = in_kbart or check_kbart(release_year, c_extra['kbart'].get(archive))

            if c_extra.get('ia'):
                if c_extra['ia'].get('sim') and release_year:
                    in_ia_sim = check_kbart(release_year, c_extra['ia']['sim'])
                if c_extra['ia'].get('longtail_oa'):
                    is_longtail_oa = True
            if c_extra.get('sherpa_romeo'):
                if c_extra['sherpa_romeo'].get('color') == 'white':
                    is_oa = False
            if c_extra.get('default_license') and c_extra.get('default_license').startswith('CC-'):
                is_oa = True
            if c_extra.get('doaj'):
                if c_extra['doaj'].get('as_of'):
                    is_oa = True
            if c_extra.get('road'):
                if c_extra['road'].get('as_of'):
                    is_oa = True
    else:
        t['publisher'] = release.publisher

    if release.ext_ids.jstor or (release.ext_ids.doi and release.ext_ids.doi.startswith('10.2307/')):
        in_jstor = True

    files = release.files or []
    t['file_count'] = len(files)
    t['fileset_count'] = len(release.filesets or [])
    t['webcapture_count'] = len(release.webcaptures or [])
    any_pdf_url = None
    good_pdf_url = None
    best_pdf_url = None
    ia_pdf_url = None
    for f in files:
        if f.extra and f.extra.get('shadows'):
            # TODO: shadow check goes here
            in_shadows = True
        is_pdf = 'pdf' in (f.mimetype or '')
        for url in (f.urls or []):
            if not f.mimetype and 'pdf' in url.lower():
                is_pdf = True
            if url.url.lower().startswith('http'):
                in_web = True
            if url.rel in ('dweb', 'p2p', 'ipfs', 'dat', 'torrent'):
                # not sure what rel will be for this stuff
                in_dweb = True
            if is_pdf:
                any_pdf_url = url.url
            if is_pdf and url.rel in ('webarchive', 'repository') and is_pdf:
                is_preserved = True
                good_pdf_url = url.url
            if '//www.jstor.org/' in url.url:
                in_jstor = True
            if '//web.archive.org/' in url.url or '//archive.org/' in url.url:
                in_ia = True
                if is_pdf:
                    best_pdf_url = url.url
                    ia_pdf_url = url.url
    # here is where we bake-in priority; IA-specific
    t['best_pdf_url'] = best_pdf_url or good_pdf_url or any_pdf_url
    t['ia_pdf_url'] = ia_pdf_url

    if release.license_slug:
        # TODO: more/better checks here, particularly strict *not* OA licenses
        if release.license_slug.startswith("CC-"):
            is_oa = True

    extra = release.extra or dict()
    if extra:
        if extra.get('is_oa'):
            # NOTE: not actually setting this anywhere... but could
            is_oa = True
        if extra.get('longtail_oa'):
            # sometimes set by GROBID/matcher
            is_oa = True
            is_longtail_oa = True
        if not t.get('container_name'):
            t['container_name'] = extra.get('container_name')
        if extra.get('crossref'):
            if extra['crossref'].get('archive'):
                # all crossref archives are KBART, I believe
                in_kbart = True

    if is_longtail_oa:
        is_oa = True

    if force_bool:
        t['is_oa'] = bool(is_oa)
        t['is_longtail_oa'] = bool(is_longtail_oa)
        t['in_kbart'] = bool(in_kbart)
        t['in_ia_sim'] = bool(in_ia_sim)
        t['in_jstor'] = bool(in_jstor)
        t['in_web'] = bool(in_web)
        t['in_dweb'] = bool(in_dweb)
    else:
        t['is_oa'] = is_oa
        t['is_longtail_oa'] = is_longtail_oa
        t['in_kbart'] = in_kbart
        t['in_ia_sim'] = in_ia_sim
        t['in_jstor'] = in_jstor
        t['in_web'] = in_web
        t['in_dweb'] = in_dweb

    t['in_ia'] = bool(in_ia)
    t['is_preserved'] = bool(is_preserved or in_ia or in_kbart or in_jstor)
    return t

def container_to_elasticsearch(entity, force_bool=True):
    """
    Converts from an entity model/schema to elasticsearch oriented schema.

    Returns: dict
    Raises exception on error (never returns None)
    """

    if entity.state in ('redirect', 'deleted'):
        return dict(
            ident = entity.ident,
            state = entity.state,
        )
    elif entity.state != 'active':
        raise ValueError("Unhandled entity state: {}".format(entity.state))

    # First, the easy ones (direct copy)
    t = dict(
        ident = entity.ident,
        state = entity.state,
        revision = entity.revision,

        name = entity.name,
        publisher = entity.publisher,
        container_type = entity.container_type,
        issnl = entity.issnl,
        wikidata_qid = entity.wikidata_qid,
    )

    # TODO: region, discipline
    # TODO: single primary language?
    if not entity.extra:
        entity.extra = dict()
    for key in ('country', 'languages', 'mimetypes', 'first_year', 'last_year'):
        if entity.extra.get(key):
            t[key] = entity.extra[key]

    in_doaj = None
    in_road = None
    # TODO: not currently implemented
    in_doi = None
    # TODO: would be nice to have 'in_doaj_works', or maybe just "any_pid"
    #in_doaj_works = None
    in_sherpa_romeo = None
    is_oa = None
    # TODO: not actually set/stored anywhere?
    is_longtail_oa = None
    any_kbart = None
    any_jstor = None
    any_ia_sim = None

    extra = entity.extra
    if extra.get('doaj'):
        if extra['doaj'].get('as_of'):
            in_doaj = True
    if extra.get('road'):
        if extra['road'].get('as_of'):
            in_road = True
    if extra.get('default_license'):
        if extra['default_license'].startswith('CC-'):
            is_oa = True
    if extra.get('sherpa_romeo'):
        in_sherpa_romeo = True
        if extra['sherpa_romeo'].get('color') == 'white':
            is_oa = False
    if extra.get('kbart'):
        any_kbart = True
        if extra['kbart'].get('jstor'):
            any_jstor = True
    if extra.get('ia'):
        if extra['ia'].get('sim'):
            any_ia_sim = True

    t['in_doaj'] = bool(in_doaj)
    t['in_road'] = bool(in_road)
    t['in_sherpa_romeo'] = bool(in_sherpa_romeo)
    t['any_kbart'] = bool(any_kbart)
    t['is_longtail_oa'] = bool(is_longtail_oa)
    if force_bool:
        t['in_doi'] = bool(in_doi)
        t['is_oa'] = bool(in_doaj or in_road or is_longtail_oa or is_oa)
        t['any_jstor'] = bool(any_jstor)
        t['any_ia_sim'] = bool(any_ia_sim)
    else:
        t['in_doi'] = in_doi
        t['is_oa'] = in_doaj or in_road or is_longtail_oa or is_oa
        t['any_jstor'] = any_jstor
        t['any_ia_sim'] = any_ia_sim
    return t


def changelog_to_elasticsearch(entity):

    editgroup = entity.editgroup
    t = dict(
        index=entity.index,
        editgroup_id=entity.editgroup_id,
        timestamp=entity.timestamp,
        editor_id=editgroup.editor_id,
    )

    extra = editgroup.extra or dict()
    if extra.get('agent'):
        t['agent'] = extra['agent']

    t['containers'] = len(editgroup.edits.containers)
    t['creators'] = len(editgroup.edits.containers)
    t['files'] = len(editgroup.edits.containers)
    t['filesets'] = len(editgroup.edits.containers)
    t['webcaptures'] = len(editgroup.edits.containers)
    t['releases'] = len(editgroup.edits.containers)
    t['works'] = len(editgroup.edits.containers)

    # TODO: parse and pull out counts
    #created = 0
    #updated = 0
    #deleted = 0
    #t['created'] = created
    #t['updated'] = updated
    #t['deleted'] = deleted
    #t['total'] = created + updated + deleted
    return t