aboutsummaryrefslogtreecommitdiffstats
path: root/python/fatcat_web/search.py
blob: fe0610e5d4f568d2efe9031e86c050b88a08dae9 (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

"""
Helpers for doing elasticsearch queries (used in the web interface; not part of
the formal API)
"""

import sys
import datetime
from dataclasses import dataclass
from typing import List, Optional, Any

import elasticsearch
from elasticsearch_dsl import Search, Q
import elasticsearch_dsl.response

from fatcat_web import app

@dataclass
class ReleaseQuery:
    q: Optional[str] = None
    limit: Optional[int] = None
    offset: Optional[int] = None
    fulltext_only: bool = False
    container_id: Optional[str] = None

    @classmethod
    def from_args(cls, args) -> 'ReleaseQuery':

        query_str = args.get('q') or '*'

        container_id = args.get('container_id')
        # TODO: as filter, not in query string
        if container_id:
            query_str += ' container_id:"{}"'.format(container_id)

        # TODO: where are container_issnl queries actually used?
        issnl = args.get('container_issnl')
        if issnl and query_str:
            query_str += ' container_issnl:"{}"'.format(issnl)

        offset = args.get('offset', '0')
        offset = max(0, int(offset)) if offset.isnumeric() else 0

        return ReleaseQuery(
            q=query_str,
            offset=offset,
            fulltext_only=bool(args.get('fulltext_only')),
            container_id=container_id,
        )

@dataclass
class GenericQuery:
    q: Optional[str] = None
    limit: Optional[int] = None
    offset: Optional[int] = None

    @classmethod
    def from_args(cls, args) -> 'GenericQuery':
        query_str = args.get('q')
        if not query_str:
            query_str = '*'
        offset = args.get('offset', '0')
        offset = max(0, int(offset)) if offset.isnumeric() else 0

        return GenericQuery(
            q=query_str,
            offset=offset,
        )

@dataclass
class SearchHits:
    count_returned: int
    count_found: int
    offset: int
    limit: int
    deep_page_limit: int
    query_time_ms: int
    results: List[Any]


def results_to_dict(response: elasticsearch_dsl.response.Response) -> List[dict]:
    """
    Takes a response returns all the hits as JSON objects.

    Also handles surrogate strings that elasticsearch returns sometimes,
    probably due to mangled data processing in some pipeline. "Crimes against
    Unicode"; production workaround
    """

    results = []
    for h in response:
        r = h._d_
        # print(h.meta._d_)
        results.append(r)

    for h in results:
        for key in h:
            if type(h[key]) is str:
                h[key] = h[key].encode("utf8", "ignore").decode("utf8")
    return results

def wrap_es_execution(search: Search) -> Any:
    """
    Executes a Search object, and converts various ES error types into
    something we can pretty print to the user.
    """
    try:
        resp = search.execute()
    except elasticsearch.exceptions.RequestError as e:
        # this is a "user" error
        print("elasticsearch 400: " + str(e.info), file=sys.stderr)
        if e.info.get("error", {}).get("root_cause", {}):
            raise ValueError(str(e.info["error"]["root_cause"][0].get("reason")))
        else:
            raise ValueError(str(e.info))
    except elasticsearch.exceptions.TransportError as e:
        # all other errors
        print("elasticsearch non-200 status code: {}".format(e.info), file=sys.stderr)
        raise IOError(str(e.info))
    return resp

def do_container_search(
    query: GenericQuery, deep_page_limit: int = 2000
) -> SearchHits:

    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_CONTAINER_INDEX'])

    search = search.query(
        "query_string",
        query=query.q,
        default_operator="AND",
        analyze_wildcard=True,
        allow_leading_wildcard=False,
        lenient=True,
        fields=["biblio"],
    )

    # Sanity checks
    limit = min((int(query.limit or 25), 100))
    offset = max((int(query.offset or 0), 0))
    if offset > deep_page_limit:
        # Avoid deep paging problem.
        offset = deep_page_limit

    search = search[offset : (offset + limit)]

    resp = wrap_es_execution(search)
    results = results_to_dict(resp)

    return SearchHits(
        count_returned=len(results),
        count_found=int(resp.hits.total),
        offset=offset,
        limit=limit,
        deep_page_limit=deep_page_limit,
        query_time_ms=int(resp.took),
        results=results,
    )

def do_release_search(
    query: ReleaseQuery, deep_page_limit: int = 2000
) -> SearchHits:

    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])

    # availability filters
    if query.fulltext_only:
        search = search.filter("term", in_ia=True)

    # Below, we combine several queries to improve scoring.

    # this query use the fancy built-in query string parser
    basic_biblio = Q(
        "query_string",
        query=query.q,
        default_operator="AND",
        analyze_wildcard=True,
        allow_leading_wildcard=False,
        lenient=True,
        fields=[
            "title^2",
            "biblio",
        ],
    )
    has_fulltext = Q("term", in_ia=True)
    poor_metadata = Q(
        "bool",
        should=[
            # if these fields aren't set, metadata is poor. The more that do
            # not exist, the stronger the signal.
            Q("bool", must_not=Q("exists", field="title")),
            Q("bool", must_not=Q("exists", field="release_year")),
            Q("bool", must_not=Q("exists", field="release_type")),
            Q("bool", must_not=Q("exists", field="release_stage")),
            Q("bool", must_not=Q("exists", field="container_id")),
        ],
    )

    search = search.query(
        "boosting",
        positive=Q("bool", must=basic_biblio, should=[has_fulltext],),
        negative=poor_metadata,
        negative_boost=0.5,
    )

    # Sanity checks
    limit = min((int(query.limit or 25), 100))
    offset = max((int(query.offset or 0), 0))
    if offset > deep_page_limit:
        # Avoid deep paging problem.
        offset = deep_page_limit

    search = search[offset : (offset + limit)]

    resp = wrap_es_execution(search)
    results = results_to_dict(resp)

    for h in results:
        # Ensure 'contrib_names' is a list, not a single string
        if type(h['contrib_names']) is not list:
            h['contrib_names'] = [h['contrib_names'], ]
        h['contrib_names'] = [name.encode('utf8', 'ignore').decode('utf8') for name in h['contrib_names']]

    return SearchHits(
        count_returned=len(results),
        count_found=int(resp.hits.total),
        offset=offset,
        limit=limit,
        deep_page_limit=deep_page_limit,
        query_time_ms=int(resp.took),
        results=results,
    )

def get_elastic_container_random_releases(ident, limit=5):
    """
    Returns a list of releases from the container.
    """

    assert limit > 0 and limit <= 100

    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])
    search = search.query(
        'bool',
        must=[
            Q('term', container_id=ident),
            Q('range', release_year={ "lte": datetime.datetime.today().year }),
        ]
    )
    search = search.sort('-in_web', '-release_date')
    search = search[:int(limit)]

    search = search.params(request_cache=True)
    resp = wrap_es_execution(search)
    results = results_to_dict(resp)

    return results

def get_elastic_entity_stats():
    """
    TODO: files, filesets, webcaptures (no schema yet)

    Returns dict:
        changelog: {latest: {index, datetime}}
        release: {total, refs_total}
        papers: {total, in_web, in_oa, in_kbart, in_web_not_kbart}
    """

    stats = {}

    # release totals
    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])
    search.aggs.bucket(
        'release_ref_count',
        'sum',
        field='ref_count',
    )
    search = search[:0]  # pylint: disable=unsubscriptable-object

    search = search.params(request_cache=True)
    resp = wrap_es_execution(search)

    stats['release'] = {
        "total": int(resp.hits.total),
        "refs_total": int(resp.aggregations.release_ref_count.value),
    }

    # paper counts
    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])
    search = search.query(
        'terms',
        release_type=[
            "article-journal",
            "paper-conference",
            # "chapter",
            # "thesis",
        ],
    )
    search.aggs.bucket(
        'paper_like',
        'filters',
        filters={
            "in_web": { "term": { "in_web": "true" } },
            "is_oa": { "term": { "is_oa": "true" } },
            "in_kbart": { "term": { "in_kbart": "true" } },
            "in_web_not_kbart": { "bool": { "filter": [
                { "term": { "in_web": "true" } },
                { "term": { "in_kbart": "false" } },
            ]}},
        }
    )
    search = search[:0]

    search = search.params(request_cache=True)
    resp = wrap_es_execution(search)
    buckets = resp.aggregations.paper_like.buckets
    stats['papers'] = {
        'total': resp.hits.total,
        'in_web': buckets.in_web.doc_count,
        'is_oa': buckets.is_oa.doc_count,
        'in_kbart': buckets.in_kbart.doc_count,
        'in_web_not_kbart': buckets.in_web_not_kbart.doc_count,
    }

    # container counts
    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_CONTAINER_INDEX'])
    search.aggs.bucket(
        'release_ref_count',
        'sum',
        field='ref_count',
    )
    search = search[:0]  # pylint: disable=unsubscriptable-object

    search = search.params(request_cache=True)
    resp = wrap_es_execution(search)
    stats['container'] = {
        "total": resp.hits.total,
    }

    return stats

def get_elastic_container_stats(ident, issnl=None):
    """
    Returns dict:
        ident
        issnl (optional)
        total
        in_web
        in_kbart
        preserved
    """

    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])
    search = search.query(
        'term',
        container_id=ident,
    )
    search.aggs.bucket(
        'container_stats',
        'filters',
        filters={
            "in_web": {
                "term": { "in_web": True },
            },
            "in_kbart": {
                "term": { "in_kbart": True },
            },
            "is_preserved": {
                "term": { "is_preserved": True },
            },
        },
    )
    search = search[:0]

    search = search.params(request_cache=True)
    resp = wrap_es_execution(search)

    buckets = resp.aggregations.container_stats.buckets
    stats = {
        'ident': ident,
        'issnl': issnl,
        'total': resp.hits.total,
        'in_web': buckets['in_web']['doc_count'],
        'in_kbart': buckets['in_kbart']['doc_count'],
        'is_preserved': buckets['is_preserved']['doc_count'],
    }

    return stats

def get_elastic_container_histogram(ident):
    """
    Fetches a stacked histogram

    Filters to the past 500 years (at most), or about 1000 values.

    Returns a list of tuples:
        (year, in_ia, count)
    """

    search = Search(using=app.es_client, index=app.config['ELASTICSEARCH_RELEASE_INDEX'])
    search = search.query(
        'bool',
        must=[
            Q("range", release_year={
                "gte": datetime.datetime.today().year - 499,
                "lte": datetime.datetime.today().year,
            }),
        ],
        filter=[
            Q("bool", minimum_should_match=1, should=[
                Q("match", container_id=ident),
            ]),
        ],
    )
    search.aggs.bucket(
        'year_in_ia',
        'composite',
        size=1000,
        sources=[
            {"year": {
                "histogram": {
                    "field": "release_year",
                    "interval": 1,
                },
            }},
            {"in_ia": {
                "terms": {
                    "field": "in_ia",
                },
            }},
        ],
    )
    search = search[:0]

    search = search.params(request_cache='true')
    resp = wrap_es_execution(search)

    buckets = resp.aggregations.year_in_ia.buckets
    vals = [(h['key']['year'], h['key']['in_ia'], h['doc_count'])
            for h in buckets]
    vals = sorted(vals)
    return vals