summaryrefslogtreecommitdiffstats
path: root/python/fatcat_tools/importers/ingest.py
blob: ae3e147a98459f1043db8320a864a939d6c72f75 (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
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
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521

import datetime

import fatcat_openapi_client
from .common import EntityImporter, make_rel_url


class IngestFileResultImporter(EntityImporter):

    def __init__(self, api, require_grobid=True, **kwargs):

        eg_desc = kwargs.pop('editgroup_description', None) or "Files crawled from web using sandcrawler ingest tool"
        eg_extra = kwargs.pop('editgroup_extra', dict())
        eg_extra['agent'] = eg_extra.get('agent', 'fatcat_tools.IngestFileResultImporter')
        kwargs['do_updates'] = kwargs.get("do_updates", False)
        super().__init__(api,
            editgroup_description=eg_desc,
            editgroup_extra=eg_extra,
            **kwargs)
        self.use_glutton_match = False
        self.default_link_rel = kwargs.get("default_link_rel", "web")
        assert self.default_link_rel
        self.require_grobid = require_grobid
        if self.require_grobid:
            print("Requiring GROBID status == 200 (for PDFs)")
        else:
            print("NOT checking GROBID success")
        self.ingest_request_source_allowlist = [
            'fatcat-changelog',
            'fatcat-ingest-container',
            'fatcat-ingest',
            'arabesque',
            #'mag-corpus',
            #'mag',
            'unpaywall-corpus',
            'unpaywall',
            #'s2-corpus',
            #'s2',
            'doaj',
            'dblp',
        ]
        if kwargs.get('skip_source_allowlist', False):
            self.ingest_request_source_allowlist = []

    def want_file(self, row) -> bool:
        """
        File-specific part of want(). Generic across general ingest and save-paper-now.
        """

        if not row.get('file_meta'):
            self.counts['skip-file-meta'] += 1
            return False

        # type-specific filters
        if row['request'].get('ingest_type') == 'pdf':
            if self.require_grobid and row.get('grobid', {}).get('status_code') != 200:
                self.counts['skip-grobid'] += 1
                return False
            if row['file_meta'].get('mimetype') not in ("application/pdf",):
                self.counts['skip-mimetype'] += 1
                return False
        elif row['request'].get('ingest_type') == 'xml':
            if row['file_meta'].get('mimetype') not in ("application/xml",
                    "application/jats+xml", "application/tei+xml", "text/xml"):
                self.counts['skip-mimetype'] += 1
                return False
        else:
            self.counts['skip-ingest-type'] += 1
            return False

        return True

    def want_ingest(self, row) -> bool:
        """
        Sandcrawler ingest-specific part of want(). Generic across file and
        webcapture ingest.
        """
        if row.get('hit') != True:
            self.counts['skip-hit'] += 1
            return False
        source = row['request'].get('ingest_request_source')
        if not source:
            self.counts['skip-ingest_request_source'] += 1
            return False
        if self.ingest_request_source_allowlist and source not in self.ingest_request_source_allowlist:
            self.counts['skip-ingest_request_source'] += 1
            return False

        if row['request'].get('link_source') not in ('arxiv', 'pmc', 'unpaywall', 'doi', 'mag', 's2', 'doaj', 'dblp'):
            self.counts['skip-link-source'] += 1
            return False

        if source.startswith('savepapernow'):
            # never process async savepapernow requests
            self.counts['skip-savepapernow'] += 1
            return False

        return True

    def want(self, row):
        """
        Overall logic here probably needs work (TODO):

        - Direct ingests via DOI from fatcat-changelog should probably go
          through regardless of GROBID status
        - We should filter/block things like single-page PDFs here
        - public/anonymous submissions could require successful biblio-glutton
          match, or some other sanity check on the fatcat side (eg, fuzzy title
          match)
        - handle the case of release_stage not being 'published'; if pre-print,
          potentially create a new release.

        The current logic is intentionally conservative as a first step.
        """
        if not self.want_ingest(row):
            return False
        if not self.want_file(row):
            return False

        return True

    def parse_ingest_release_ident(self, row):

        request = row['request']
        fatcat = request.get('fatcat')

        release_ident = None
        if fatcat and fatcat.get('release_ident'):
            release_ident = fatcat.get('release_ident')
        elif request.get('ext_ids'):
            # if no fatcat ident, try extids
            for extid_type in ('doi', 'pmid', 'pmcid', 'arxiv', 'doaj', 'dblp'):
                extid = request['ext_ids'].get(extid_type)
                if not extid:
                    continue
                if extid_type == 'doi':
                    extid = extid.lower()
                try:
                    release = self.api.lookup_release(**{extid_type: extid})
                except fatcat_openapi_client.rest.ApiException as err:
                    if err.status == 404:
                        continue
                    elif err.status == 400:
                        self.counts['warn-extid-invalid'] += 1
                        continue
                    raise err
                # verify release_stage
                if request.get('release_stage') and release.release_stage:
                    if request['release_stage'] != release.release_stage:
                        self.counts['skip-release-stage'] += 1
                        return None
                release_ident = release.ident
                break

        if self.use_glutton_match and not release_ident and row.get('grobid'):
            # try biblio-glutton extracted hit
            if row['grobid'].get('fatcat_release'):
                release_ident = row['grobid']['fatcat_release'].split('_')[-1]
                self.counts['glutton-match'] += 1

        return release_ident

    def parse_terminal(self, row):
        terminal = row.get('terminal')
        if not terminal:
            # support old cdx-only ingest results
            cdx = row.get('cdx')
            if not cdx:
                return None
            else:
                terminal = {
                    'terminal_url': cdx['url'],
                    'terminal_dt': cdx['datetime'],
                    'terminal_status_code': cdx.get('status_code') or cdx.get('http_status'),
                }

        # work around old schema
        if not 'terminal_url' in terminal:
            terminal['terminal_url'] = terminal['url']
        if not 'terminal_dt' in terminal:
            terminal['terminal_dt'] = terminal['dt']

        # convert CDX-style digits to ISO-style timestamp
        assert len(terminal['terminal_dt']) == 14
        terminal['terminal_timestamp'] = datetime.datetime.strptime(terminal['terminal_dt'], "%Y%m%d%H%M%S").isoformat() + "Z"
        return terminal

    def parse_urls(self, row, terminal):

        request = row['request']

        default_rel = self.default_link_rel
        if request.get('link_source') == 'doi':
            default_rel = 'publisher'
        default_rel = request.get('rel', default_rel)
        url = make_rel_url(terminal['terminal_url'], default_rel)

        if not url:
            self.counts['skip-url'] += 1
            return None
        wayback = "https://web.archive.org/web/{}/{}".format(
            terminal['terminal_dt'],
            terminal['terminal_url'])
        urls = [url, ("webarchive", wayback)]

        urls = [fatcat_openapi_client.FileUrl(rel=rel, url=url) for (rel, url) in urls]
        return urls

    def parse_edit_extra(self, row):

        request = row['request']
        edit_extra = dict()

        if request.get('edit_extra'):
            edit_extra = request['edit_extra']

        if request.get('ingest_request_source'):
            edit_extra['ingest_request_source'] = request['ingest_request_source']
        if request.get('link_source') and request.get('link_source_id'):
            edit_extra['link_source'] = request['link_source']
            edit_extra['link_source_id'] = request['link_source_id']
            if edit_extra['link_source'] == 'doi':
                edit_extra['link_source_id'] = edit_extra['link_source_id'].lower()

        return edit_extra

    def parse_record(self, row):

        request = row['request']
        file_meta = row['file_meta']

        # double check that want() filtered request correctly (eg, old requests)
        if request.get('ingest_type') not in ('pdf', 'xml'):
            self.counts['skip-ingest-type'] += 1
            return None
        assert (request['ingest_type'], file_meta['mimetype']) in [
            ("pdf", "application/pdf"),
            ("xml", "application/xml"),
            ("xml", "application/jats+xml"),
            ("xml", "application/tei+xml"),
            ("xml", "text/xml"),
        ]

        # identify release by fatcat ident, or extid lookup, or biblio-glutton match
        release_ident = self.parse_ingest_release_ident(row)

        if not release_ident:
            self.counts['skip-release-not-found'] += 1
            return None

        terminal = self.parse_terminal(row)
        if not terminal:
            # TODO: support archive.org hits?
            self.counts['skip-no-terminal'] += 1
            return None

        urls = self.parse_urls(row, terminal)

        fe = fatcat_openapi_client.FileEntity(
            md5=file_meta['md5hex'],
            sha1=file_meta['sha1hex'],
            sha256=file_meta['sha256hex'],
            size=file_meta['size_bytes'],
            mimetype=file_meta['mimetype'],
            release_ids=[release_ident],
            urls=urls,
        )

        edit_extra = self.parse_edit_extra(row)
        if edit_extra:
            fe.edit_extra = edit_extra
        return fe

    def try_update(self, fe):
        # lookup sha1, or create new entity
        existing = None
        try:
            existing = self.api.lookup_file(sha1=fe.sha1)
        except fatcat_openapi_client.rest.ApiException as err:
            if err.status != 404:
                raise err

        # check for existing edits-in-progress with same file hash
        for other in self._entity_queue:
            if other.sha1 == fe.sha1:
                self.counts['skip-in-queue'] += 1
                return False

        if not existing:
            return True

        # NOTE: the following checks all assume there is an existing item
        if (fe.release_ids[0] in existing.release_ids) and existing.urls:
            # TODO: could still, in theory update with the new URL?
            self.counts['exists'] += 1
            return False

        if not self.do_updates:
            self.counts['skip-update-disabled'] += 1
            return False

        # TODO: for now, never update
        self.counts['skip-update-disabled'] += 1
        return False

    def insert_batch(self, batch):
        self.api.create_file_auto_batch(fatcat_openapi_client.FileAutoBatch(
            editgroup=fatcat_openapi_client.Editgroup(
                description=self.editgroup_description,
                extra=self.editgroup_extra),
            entity_list=batch))


class SavePaperNowFileImporter(IngestFileResultImporter):
    """
    This worker ingests from the same feed as IngestFileResultImporter, but
    only imports files from anonymous save-paper-now requests, and "submits"
    them for further human review (as opposed to accepting by default).
    """

    def __init__(self, api, submit_mode=True, **kwargs):

        eg_desc = kwargs.pop('editgroup_description', None) or "Files crawled after a public 'Save Paper Now' request"
        eg_extra = kwargs.pop('editgroup_extra', dict())
        eg_extra['agent'] = eg_extra.get('agent', 'fatcat_tools.IngestFileSavePaperNow')
        kwargs['submit_mode'] = submit_mode
        kwargs['require_grobid'] = True
        kwargs['do_updates'] = False
        super().__init__(api,
            editgroup_description=eg_desc,
            editgroup_extra=eg_extra,
            **kwargs)

    def want(self, row):

        if not self.want_file(row):
            return False

        source = row['request'].get('ingest_request_source')
        if not source:
            self.counts['skip-ingest_request_source'] += 1
            return False
        if not source.startswith('savepapernow'):
            self.counts['skip-not-savepapernow'] += 1
            return False
        if row.get('hit') != True:
            self.counts['skip-hit'] += 1
            return False

        return True

    def insert_batch(self, batch):
        """
        Usually running in submit_mode, so we can't use auto_batch method
        """
        if self.submit_mode:
            eg = self.api.create_editgroup(fatcat_openapi_client.Editgroup(
                description=self.editgroup_description,
                extra=self.editgroup_extra))
            for fe in batch:
                self.api.create_file(eg.editgroup_id, fe)
            self.api.update_editgroup(eg.editgroup_id, eg, submit=True)
        else:
            self.api.create_file_auto_batch(fatcat_openapi_client.FileAutoBatch(
                editgroup=fatcat_openapi_client.Editgroup(
                    description=self.editgroup_description,
                    extra=self.editgroup_extra),
                entity_list=batch))


class IngestWebResultImporter(IngestFileResultImporter):
    """
    Variant of IngestFileResultImporter for processing HTML ingest requests
    into webcapture objects.
    """

    def __init__(self, api, **kwargs):

        eg_desc = kwargs.pop('editgroup_description', None) or "Webcaptures crawled from web using sandcrawler ingest tool"
        eg_extra = kwargs.pop('editgroup_extra', dict())
        eg_extra['agent'] = eg_extra.get('agent', 'fatcat_tools.IngestWebResultImporter')
        kwargs['do_updates'] = False
        super().__init__(api,
            editgroup_description=eg_desc,
            editgroup_extra=eg_extra,
            **kwargs)

    def want(self, row):

        if not self.want_ingest(row):
            return False

        if not row.get('file_meta'):
            self.counts['skip-file-meta'] += 1
            return False

        # webcapture-specific filters
        if row['request'].get('ingest_type') != 'html':
            self.counts['skip-ingest-type'] += 1
            return False
        if row['file_meta'].get('mimetype') not in ("text/html", "application/xhtml+xml"):
            self.counts['skip-mimetype'] += 1
            return False

        return True

    def parse_record(self, row):

        request = row['request']
        file_meta = row['file_meta']

        # double check that want() filtered request correctly (eg, old requests)
        if request.get('ingest_type') != "html":
            self.counts['skip-ingest-type'] += 1
            return None
        if file_meta['mimetype'] not in ("text/html", "application/xhtml+xml"):
            self.counts['skip-mimetype'] += 1
            return None

        # identify release by fatcat ident, or extid lookup
        release_ident = self.parse_ingest_release_ident(row)

        if not release_ident:
            self.counts['skip-release-not-found'] += 1
            return None

        terminal = self.parse_terminal(row)
        if not terminal:
            # TODO: support archive.org hits?
            self.counts['skip-no-terminal'] += 1
            return None

        urls = self.parse_urls(row, terminal)
        archive_urls = [u for u in urls if u.rel == 'webarchive']

        if terminal['terminal_status_code'] != 200:
            self.counts['skip-terminal-status-code'] += 1
            return None

        terminal_cdx = row['cdx']
        if 'revisit_cdx' in row:
            terminal_cdx = row['revisit_cdx']
        assert terminal_cdx['surt']
        if terminal_cdx['url'] != terminal['terminal_url']:
            self.counts['skip-terminal-url-mismatch'] += 1
            return None

        wc_cdx = []
        # primary resource first
        wc_cdx.append(fatcat_openapi_client.WebcaptureCdxLine(
            surt=terminal_cdx['surt'],
            timestamp=terminal['terminal_timestamp'],
            url=terminal['terminal_url'],
            mimetype=file_meta['mimetype'],
            status_code=terminal['terminal_status_code'],
            sha1=file_meta['sha1hex'],
            sha256=file_meta['sha256hex'],
            size=file_meta['size_bytes'],
        ))

        for resource in row.get('html_resources', []):
            timestamp = resource['timestamp']
            if not "+" in timestamp and not "Z" in timestamp:
                timestamp += "Z"
            wc_cdx.append(fatcat_openapi_client.WebcaptureCdxLine(
                surt=resource['surt'],
                timestamp=timestamp,
                url=resource['url'],
                mimetype=resource.get('mimetype'),
                size=resource.get('size'),
                sha1=resource.get('sha1hex'),
                sha256=resource.get('sha256hex'),
            ))

        wc = fatcat_openapi_client.WebcaptureEntity(
            cdx=wc_cdx,
            archive_urls=archive_urls,
            original_url=terminal['terminal_url'],
            timestamp=terminal['terminal_timestamp'],
            release_ids=[release_ident],
        )

        edit_extra = self.parse_edit_extra(row)

        if edit_extra:
            wc.edit_extra = edit_extra
        return wc

    def try_update(self, wc):

        # check for existing edits-in-progress with same URL
        for other in self._entity_queue:
            if other.original_url == wc.original_url:
                self.counts['skip-in-queue'] += 1
                return False

        # lookup sha1, or create new entity (TODO: API doesn't support this yet)
        #existing = None

        # TODO: currently only allow one release per webcapture
        release = self.api.get_release(wc.release_ids[0], expand="webcaptures")
        if release.webcaptures:
            # check if this is an existing match, or just a similar hit
            for other in release.webcaptures:
                if wc.original_url == other.original_url:
                    # TODO: compare very similar timestamps of same time (different formats)
                    self.counts['exists'] += 1
                    return False
            self.counts['skip-release-has-webcapture'] += 1
            return False

        # Ok, if we got here then no existing web capture for (first) release,
        # so go ahead and insert!
        return True

    def insert_batch(self, batch):
        self.api.create_webcapture_auto_batch(fatcat_openapi_client.WebcaptureAutoBatch(
            editgroup=fatcat_openapi_client.Editgroup(
                description=self.editgroup_description,
                extra=self.editgroup_extra),
            entity_list=batch))