repo.py 31.2 KB
Newer Older
Markus Scheidgen's avatar
Markus Scheidgen committed
1
2
3
4
5
6
7
8
9
10
11
12
13
14
# Copyright 2018 Markus Scheidgen
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#   http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an"AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

15
'''
Markus Scheidgen's avatar
Markus Scheidgen committed
16
17
The repository API of the nomad@FAIRDI APIs. Currently allows to resolve repository
meta-data.
18
'''
Markus Scheidgen's avatar
Markus Scheidgen committed
19

20
from typing import List, Dict, Any
Markus Scheidgen's avatar
Markus Scheidgen committed
21
from flask_restplus import Resource, abort, fields
22
from flask import request, g
23
from elasticsearch_dsl import Q
24
from elasticsearch.exceptions import NotFoundError
25
import elasticsearch.helpers
Markus Scheidgen's avatar
Markus Scheidgen committed
26
from datetime import datetime
Markus Scheidgen's avatar
Markus Scheidgen committed
27

28
from nomad import search, utils, datamodel, processing as proc, infrastructure
29
from nomad.datamodel import Dataset, User, EditableUserMetadata
30
from nomad.app import common
31
from nomad.app.common import RFC3339DateTime, DotKeyNested
Markus Scheidgen's avatar
Markus Scheidgen committed
32

Markus Scheidgen's avatar
Markus Scheidgen committed
33
from .api import api
34
from .auth import authenticate
35
36
from .common import search_model, calc_route, add_pagination_parameters,\
    add_scroll_parameters, add_search_parameters, apply_search_parameters,\
37
    query_api_python, query_api_curl, _search_quantities
Markus Scheidgen's avatar
Markus Scheidgen committed
38

39
ns = api.namespace('repo', description='Access repository metadata.')
Markus Scheidgen's avatar
Markus Scheidgen committed
40
41
42
43
44


@calc_route(ns)
class RepoCalcResource(Resource):
    @api.response(404, 'The upload or calculation does not exist')
45
    @api.response(401, 'Not authorized to access the calculation')
46
    @api.response(200, 'Metadata send', fields.Raw)
47
    @api.doc('get_repo_calc')
48
    @authenticate()
49
    def get(self, upload_id, calc_id):
50
        '''
Markus Scheidgen's avatar
Markus Scheidgen committed
51
52
        Get calculation metadata in repository form.

53
        Repository metadata only entails the quantities shown in the repository.
54
        Calcs are references via *upload_id*, *calc_id* pairs.
55
        '''
Markus Scheidgen's avatar
Markus Scheidgen committed
56
        try:
57
58
59
60
61
62
63
64
            calc = search.Entry.get(calc_id)
        except NotFoundError:
            abort(404, message='There is no calculation %s/%s' % (upload_id, calc_id))

        if calc.with_embargo or not calc.published:
            if g.user is None:
                abort(401, message='Not logged in to access %s/%s.' % (upload_id, calc_id))

65
            if not (any(g.user.user_id == user.user_id for user in calc.owners) or g.user.is_admin):
66
67
                abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))

68
        result = calc.to_dict()
69
70
        result['python'] = query_api_python('archive', upload_id, calc_id)
        result['curl'] = query_api_curl('archive', upload_id, calc_id)
Markus Scheidgen's avatar
Markus Scheidgen committed
71

72
        return result, 200
Markus Scheidgen's avatar
Markus Scheidgen committed
73

74

75
76
77
78
79
_search_request_parser = api.parser()
add_pagination_parameters(_search_request_parser)
add_scroll_parameters(_search_request_parser)
add_search_parameters(_search_request_parser)
_search_request_parser.add_argument(
80
    'date_histogram', type=bool, help='Add an additional aggregation over the upload time')
81
_search_request_parser.add_argument(
82
    'metrics', type=str, action='append', help=(
83
        'Metrics to aggregate over all quantities and their values as comma separated list. '
84
        'Possible values are %s.' % ', '.join(search.metrics_names)))
85
_search_request_parser.add_argument(
Markus Scheidgen's avatar
Markus Scheidgen committed
86
    'statistics', type=bool, help=('Return statistics.'))
87
88
_search_request_parser.add_argument(
    'exclude', type=str, action='split', help='Excludes the given keys in the returned data.')
Markus Scheidgen's avatar
Markus Scheidgen committed
89
for group_name in search.groups:
90
    _search_request_parser.add_argument(
91
        'group_%s' % group_name, type=bool, help=('Return %s group data.' % group_name))
92
    _search_request_parser.add_argument(
Markus Scheidgen's avatar
Markus Scheidgen committed
93
94
95
        '%s_after' % group_name, type=str,
        help='The last %s id of the last scroll window for the %s group' % (group_name, group_name))

96
97
98
99
100
_repo_calcs_model_fields = {
    'statistics': fields.Raw(description=(
        'A dict with all statistics. Each statistic is dictionary with a metrics dict as '
        'value and quantity value as key. The possible metrics are code runs(calcs), %s. '
        'There is a pseudo quantity "total" with a single value "all" that contains the '
101
102
        ' metrics over all results. ' % ', '.join(search.metrics_names)))}

103
for group_name in search.groups:
104
    _repo_calcs_model_fields[group_name] = (DotKeyNested if '.' in group_name else fields.Nested)(api.model('RepoGroup', {
105
        'after': fields.String(description='The after value that can be used to retrieve the next %s.' % group_name),
106
        'values': fields.Raw(description='A dict with %s as key. The values are dicts with "total" and "examples" keys.' % group_name)
107
    }), skip_none=True)
108

109
110
for qualified_name, quantity in search.search_quantities.items():
    _repo_calcs_model_fields[qualified_name] = fields.Raw(
111
112
        description=quantity.description, allow_null=True, skip_none=True)

113
_repo_calcs_model = api.inherit('RepoCalculations', search_model, _repo_calcs_model_fields)
114
115


Markus Scheidgen's avatar
Markus Scheidgen committed
116
117
@ns.route('/')
class RepoCalcsResource(Resource):
118
    @api.doc('search')
119
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
120
121
    @api.expect(_search_request_parser, validate=True)
    @api.marshal_with(_repo_calcs_model, skip_none=True, code=200, description='Search results send')
122
    @authenticate()
Markus Scheidgen's avatar
Markus Scheidgen committed
123
    def get(self):
124
        '''
125
        Search for calculations in the repository form, paginated.
126
127

        The ``owner`` parameter determines the overall entries to search through.
128
129
130
131
        Possible values are: ``all`` (show all entries visible to the current user), ``public``
        (show all publically visible entries), ``user`` (show all user entries, requires login),
        ``staging`` (show all user entries in staging area, requires login).

132
133
134
135
136
        You can use the various quantities to search/filter for. For some of the
        indexed quantities this endpoint returns aggregation information. This means
        you will be given a list of all possible values and the number of entries
        that have the certain value. You can also use these aggregations on an empty
        search to determine the possible values.
137
138
139

        The pagination parameters allows determine which page to return via the
        ``page`` and ``per_page`` parameters. Pagination however, is limited to the first
Markus Scheidgen's avatar
Markus Scheidgen committed
140
141
142
143
144
145
146
147
        100k (depending on ES configuration) hits.

        An alternative to pagination is to use ``scroll`` and ``scroll_id``. With ``scroll``
        you will get a ``scroll_id`` on the first request. Each call with ``scroll`` and
        the respective ``scroll_id`` will return the next ``per_page`` (here the default is 1000)
        results. Scroll however, ignores ordering and does not return aggregations.
        The scroll view used in the background will stay alive for 1 minute between requests.
        If the given ``scroll_id`` is not available anymore, a HTTP 400 is raised.
148
149
150
151

        The search will return aggregations on a predefined set of quantities. Aggregations
        will tell you what quantity values exist and how many entries match those values.

152
153
        Ordering is determined by ``order_by`` and ``order`` parameters. Default is
        ``upload_time`` in decending order.
154
        '''
155
156

        try:
157
            parsed_args = _search_request_parser.parse_args()
158
            args = {
159
                key: value for key, value in parsed_args.items()
160
161
162
163
164
165
166
                if value is not None}

            scroll = args.get('scroll', False)
            scroll_id = args.get('scroll_id', None)
            page = args.get('page', 1)
            per_page = args.get('per_page', 10 if not scroll else 1000)
            order = args.get('order', -1)
167
            order_by = args.get('order_by', 'upload_time')
168
169

            date_histogram = args.get('date_histogram', False)
170
            metrics: List[str] = request.args.getlist('metrics')
Markus Scheidgen's avatar
Markus Scheidgen committed
171

Markus Scheidgen's avatar
Markus Scheidgen committed
172
            with_statistics = args.get('statistics', False) or \
173
                any(args.get('group_%s' % group_name, False) for group_name in search.groups)
174
175
176
177
        except Exception as e:
            abort(400, message='bad parameters: %s' % str(e))

        search_request = search.SearchRequest()
178
        apply_search_parameters(search_request, args)
179
180
        if date_histogram:
            search_request.date_histogram()
181

182
        try:
183
            assert page >= 1
184
            assert per_page >= 0
185
186
187
        except AssertionError:
            abort(400, message='invalid pagination')

188
189
190
        if order not in [-1, 1]:
            abort(400, message='invalid pagination')

191
192
        for metric in metrics:
            if metric not in search.metrics_names:
193
194
                abort(400, message='there is no metric %s' % metric)

Markus Scheidgen's avatar
Markus Scheidgen committed
195
196
        if with_statistics:
            search_request.default_statistics(metrics_to_use=metrics)
197

Markus Scheidgen's avatar
Markus Scheidgen committed
198
            additional_metrics = [
199
200
201
                group_quantity.metric_name
                for group_name, group_quantity in search.groups.items()
                if args.get('group_%s' % group_name, False)]
202
203
204

            total_metrics = metrics + additional_metrics

Markus Scheidgen's avatar
Markus Scheidgen committed
205
206
            search_request.totals(metrics_to_use=total_metrics)
            search_request.statistic('authors', 1000)
207
208
        elif len(metrics) > 0:
            search_request.totals(metrics_to_use=metrics)
209

210
        if 'exclude' in parsed_args:
211
212
213
            excludes = parsed_args['exclude']
            if excludes is not None:
                search_request.exclude(*excludes)
214

215
        try:
216
            if scroll:
217
                results = search_request.execute_scrolled(scroll_id=scroll_id, size=per_page)
218

219
            else:
220
221
                for group_name, group_quantity in search.groups.items():
                    if args.get('group_%s' % group_name, False):
222
                        kwargs: Dict[str, Any] = {}
223
                        if group_name == 'group_uploads':
224
                            kwargs.update(order_by='upload_time', order='desc')
Markus Scheidgen's avatar
Markus Scheidgen committed
225
                        search_request.quantity(
226
                            group_quantity.qualified_name, size=per_page, examples=1,
227
228
                            after=request.args.get('%s_after' % group_name, None),
                            **kwargs)
229

230
231
                results = search_request.execute_paginated(
                    per_page=per_page, page=page, order=order, order_by=order_by)
232
233

                # TODO just a work around to make things prettier
Markus Scheidgen's avatar
Markus Scheidgen committed
234
235
236
237
238
                if with_statistics:
                    statistics = results['statistics']
                    if 'code_name' in statistics and 'currupted mainfile' in statistics['code_name']:
                        del(statistics['code_name']['currupted mainfile'])

239
240
241
                if 'quantities' in results:
                    quantities = results.pop('quantities')

242
243
244
                for group_name, group_quantity in search.groups.items():
                    if args.get('group_%s' % group_name, False):
                        results[group_name] = quantities[group_quantity.qualified_name]
245

246
            # build python code/curl snippet
247
248
249
250
251
            code_args = dict(request.args)
            if 'statistics' in code_args:
                del(code_args['statistics'])
            results['curl'] = query_api_curl('archive', 'query', query_string=code_args)
            results['python'] = query_api_python('archive', 'query', query_string=code_args)
252

253
            return results, 200
Markus Scheidgen's avatar
Markus Scheidgen committed
254
255
        except search.ScrollIdNotFound:
            abort(400, 'The given scroll_id does not exist.')
256
        except KeyError as e:
257
258
            import traceback
            traceback.print_exc()
259
            abort(400, str(e))
260

261

262
_query_model_parameters = {
263
264
265
266
267
    'owner': fields.String(description='Specify which calcs to return: ``all``, ``public``, ``user``, ``staging``, default is ``all``'),
    'from_time': RFC3339DateTime(description='A yyyy-MM-ddTHH:mm:ss (RFC3339) minimum entry time (e.g. upload time)'),
    'until_time': RFC3339DateTime(description='A yyyy-MM-ddTHH:mm:ss (RFC3339) maximum entry time (e.g. upload time)')
}

268
269
for qualified_name, quantity in search.search_quantities.items():
    if quantity.many_and:
270
271
272
273
        def field(**kwargs):
            return fields.List(fields.String(**kwargs))
    else:
        field = fields.String
274
    _query_model_parameters[qualified_name] = field(description=quantity.description)
275

276
_repo_query_model = api.model('RepoQuery', _query_model_parameters, skip_none=True)
277
278
279
280


def repo_edit_action_field(quantity):
    if quantity.is_scalar:
281
        return fields.Nested(_repo_edit_action_model, description=quantity.description, skip_none=True)
282
283
    else:
        return fields.List(
284
            fields.Nested(_repo_edit_action_model, skip_none=True), description=quantity.description)
285
286


287
_repo_edit_action_model = api.model('RepoEditAction', {
288
    'value': fields.String(description='The value/values that is set as a string.'),
289
290
291
292
    'success': fields.Boolean(description='If this can/could be done. Only in API response.'),
    'message': fields.String(descriptin='A message that details the action result. Only in API response.')
})

293
_repo_edit_model = api.model('RepoEdit', {
294
    'verify': fields.Boolean(description='If true, no action is performed.'),
295
    'query': fields.Nested(_repo_query_model, skip_none=True, description='New metadata will be applied to query results.'),
296
297
298
    'actions': fields.Nested(
        api.model('RepoEditActions', {
            quantity.name: repo_edit_action_field(quantity)
299
            for quantity in EditableUserMetadata.m_def.definitions
300
        }), skip_none=True,
301
302
303
        description='Each action specifies a single value (even for multi valued quantities).'),
    'success': fields.Boolean(description='If the overall edit can/could be done. Only in API response.'),
    'message': fields.String(description='A message that details the overall edit result. Only in API response.')
304
305
})

306
307
308
_editable_quantities = {
    quantity.name: quantity for quantity in EditableUserMetadata.m_def.definitions}

309

310
def edit(parsed_query: Dict[str, Any], mongo_update: Dict[str, Any] = None, re_index=True) -> List[str]:
311
    # get all calculations that have to change
312
    with utils.timer(common.logger, 'edit query executed'):
313
        search_request = search.SearchRequest().include('calc_id', 'upload_id')
314
        apply_search_parameters(search_request, parsed_query)
315
316
317
318
319
        upload_ids = set()
        calc_ids = []
        for hit in search_request.execute_scan():
            calc_ids.append(hit['calc_id'])
            upload_ids.add(hit['upload_id'])
320
321

    # perform the update on the mongo db
322
    with utils.timer(common.logger, 'edit mongo update executed', size=len(calc_ids)):
323
324
325
        if mongo_update is not None:
            n_updated = proc.Calc.objects(calc_id__in=calc_ids).update(multi=True, **mongo_update)
            if n_updated != len(calc_ids):
326
                common.logger.error('edit repo did not update all entries', payload=mongo_update)
327
328

    # re-index the affected entries in elastic search
329
    with utils.timer(common.logger, 'edit elastic update executed', size=len(calc_ids)):
330
331
332
        if re_index:
            def elastic_updates():
                for calc in proc.Calc.objects(calc_id__in=calc_ids):
333
334
                    entry = search.create_entry(
                        datamodel.EntryMetadata.m_from_dict(calc['metadata']))
335
336
337
338
339
340
341
342
                    entry = entry.to_dict(include_meta=True)
                    entry['_op_type'] = 'index'
                    yield entry

            _, failed = elasticsearch.helpers.bulk(
                infrastructure.elastic_client, elastic_updates(), stats_only=True)
            search.refresh()
            if failed > 0:
343
                common.logger.error(
344
345
                    'edit repo with failed elastic updates',
                    payload=mongo_update, nfailed=len(failed))
346

347
348
    return list(upload_ids)

349

350
def get_uploader_ids(query):
351
    ''' Get all the uploader from the query, to check coauthers and shared_with for uploaders. '''
352
    search_request = search.SearchRequest()
353
    apply_search_parameters(search_request, query)
354
355
356
357
    search_request.quantity(name='uploader_id')
    return search_request.execute()['quantities']['uploader_id']['values']


358
359
@ns.route('/edit')
class EditRepoCalcsResource(Resource):
360
361
    @api.doc('edit_repo')
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
362
363
    @api.expect(_repo_edit_model)
    @api.marshal_with(_repo_edit_model, skip_none=True, code=200, description='Edit verified/performed')
364
    @authenticate()
365
    def post(self):
366
        ''' Edit repository metadata. '''
367
368

        # basic body parsing and some semantic checks
369
370
371
372
373
374
375
376
377
378
        json_data = request.get_json()
        if json_data is None:
            json_data = {}
        query = json_data.get('query', {})

        owner = query.get('owner', 'user')
        if owner not in ['user', 'staging']:
            abort(400, 'Not a valid owner for edit %s. Edit can only be performed in user or staging' % owner)
        query['owner'] = owner

379
380
381
382
        if 'actions' not in json_data:
            abort(400, 'Missing key actions in edit data')
        actions = json_data['actions']
        verify = json_data.get('verify', False)
383

384
385
        # preparing the query of entries that are edited
        parsed_query = {}
386
387
        for quantity_name, value in query.items():
            if quantity_name in _search_quantities:
388
389
390
391
                quantity = search.search_quantities[quantity_name]
                if quantity.many:
                    if not isinstance(value, list):
                        value = value.split(',')
392
393
                parsed_query[quantity_name] = value
        parsed_query['owner'] = owner
394
        parsed_query['domain'] = query.get('domain')
395

396
        # checking the edit actions and preparing a mongo update on the fly
397
        json_data['success'] = True
398
        mongo_update = {}
399
        uploader_ids = None
400
        lift_embargo = False
401
402
        removed_datasets = None

403
        with utils.timer(common.logger, 'edit verified'):
404
            for action_quantity_name, quantity_actions in actions.items():
405
                quantity = _editable_quantities.get(action_quantity_name)
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
                if quantity is None:
                    abort(400, 'Unknown quantity %s' % action_quantity_name)

                quantity_flask = quantity.m_x('flask', {})
                if quantity_flask.get('admin_only', False):
                    if not g.user.is_admin():
                        abort(404, 'Only the admin user can set %s' % quantity.name)

                if isinstance(quantity_actions, list) == quantity.is_scalar:
                    abort(400, 'Wrong shape for quantity %s' % action_quantity_name)

                if not isinstance(quantity_actions, list):
                    quantity_actions = [quantity_actions]

                flask_verify = quantity_flask.get('verify', None)
                mongo_key = 'metadata__%s' % quantity.name
                has_error = False
                for action in quantity_actions:
                    action['success'] = True
                    action['message'] = None
                    action_value = action.get('value')
                    action_value = action_value if action_value is None else action_value.strip()

                    if action_value is None:
                        mongo_value = None
431

432
                    elif action_value == '':
433
                        mongo_value = None
434
435
436
437
438

                    elif flask_verify == datamodel.User:
                        try:
                            mongo_value = User.get(user_id=action_value).user_id
                        except KeyError:
439
440
                            action['success'] = False
                            has_error = True
441
                            action['message'] = 'User does not exist'
442
                            continue
443

444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
                        if uploader_ids is None:
                            uploader_ids = get_uploader_ids(parsed_query)
                        if action_value in uploader_ids:
                            action['success'] = False
                            has_error = True
                            action['message'] = 'This user is already an uploader of one entry in the query'
                            continue

                    elif flask_verify == datamodel.Dataset:
                        try:
                            mongo_value = Dataset.m_def.m_x('me').get(
                                user_id=g.user.user_id, name=action_value).dataset_id
                        except KeyError:
                            action['message'] = 'Dataset does not exist and will be created'
                            mongo_value = None
                            if not verify:
                                dataset = Dataset(
                                    dataset_id=utils.create_uuid(), user_id=g.user.user_id,
Markus Scheidgen's avatar
Markus Scheidgen committed
462
                                    name=action_value, created=datetime.utcnow())
463
464
465
466
467
468
469
470
471
472
                                dataset.m_x('me').create()
                                mongo_value = dataset.dataset_id

                    elif action_quantity_name == 'with_embargo':
                        # ignore the actual value ... just lift the embargo
                        mongo_value = False
                        lift_embargo = True

                        # check if necessary
                        search_request = search.SearchRequest()
473
                        apply_search_parameters(search_request, parsed_query)
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
                        search_request.q = search_request.q & Q('term', with_embargo=True)
                        if search_request.execute()['total'] == 0:
                            action['success'] = False
                            has_error = True
                            action['message'] = 'There is no embargo to lift'
                            continue
                    else:
                        mongo_value = action_value

                    if len(quantity.shape) == 0:
                        mongo_update[mongo_key] = mongo_value
                    else:
                        mongo_values = mongo_update.setdefault(mongo_key, [])
                        if mongo_value is not None:
                            if mongo_value in mongo_values:
                                action['success'] = False
                                has_error = True
                                action['message'] = 'Duplicate values are not allowed'
                                continue
                            mongo_values.append(mongo_value)

                if len(quantity_actions) == 0 and len(quantity.shape) > 0:
                    mongo_update[mongo_key] = []

                if action_quantity_name == 'datasets':
                    # check if datasets edit is allowed and if datasets have to be removed
                    search_request = search.SearchRequest()
501
                    apply_search_parameters(search_request, parsed_query)
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
                    search_request.quantity(name='dataset_id')
                    old_datasets = list(
                        search_request.execute()['quantities']['dataset_id']['values'].keys())

                    removed_datasets = []
                    for dataset_id in old_datasets:
                        if dataset_id not in mongo_update.get(mongo_key, []):
                            removed_datasets.append(dataset_id)

                    doi_ds = Dataset.m_def.m_x('me').objects(
                        dataset_id__in=removed_datasets, doi__ne=None).first()
                    if doi_ds is not None:
                        json_data['success'] = False
                        json_data['message'] = json_data.get('message', '') + \
                            'Edit would remove entries from a dataset with DOI (%s) ' % doi_ds.name
                        has_error = True
518

519
520
521
522
523
524
525
526
        # stop here, if client just wants to verify its actions
        if verify:
            return json_data, 200

        # stop if the action were not ok
        if has_error:
            return json_data, 400

527
        # perform the change
Markus Scheidgen's avatar
Markus Scheidgen committed
528
        mongo_update['metadata__last_edit'] = datetime.utcnow()
529
        upload_ids = edit(parsed_query, mongo_update, True)
530
531
532
533
534
535

        # lift embargo
        if lift_embargo:
            for upload_id in upload_ids:
                upload = proc.Upload.get(upload_id)
                upload.re_pack()
536

537
        # remove potentially empty old datasets
538
        if removed_datasets is not None:
539
540
541
            for dataset in removed_datasets:
                if proc.Calc.objects(metadata__dataset_id=dataset).first() is None:
                    Dataset.m_def.m_x('me').objects(dataset_id=dataset).delete()
542

543
        return json_data, 200
544

545

546
547
548
549
550
551
552
553
_repo_quantity_search_request_parser = api.parser()
add_search_parameters(_repo_quantity_search_request_parser)
_repo_quantity_search_request_parser.add_argument(
    'after', type=str, help='The after value to use for "scrolling".')
_repo_quantity_search_request_parser.add_argument(
    'size', type=int, help='The max size of the returned values.')

_repo_quantity_model = api.model('RepoQuantity', {
554
555
556
557
    'after': fields.String(description='The after value that can be used to retrieve the next set of values.'),
    'values': fields.Raw(description='A dict with values as key. Values are dicts with "total" and "examples" keys.')
})

558
559
_repo_quantity_values_model = api.model('RepoQuantityValues', {
    'quantity': fields.Nested(_repo_quantity_model, allow_null=True)
560
561
})

562

563
@ns.route('/quantity/<string:quantity>')
564
565
566
class RepoQuantityResource(Resource):
    @api.doc('quantity_search')
    @api.response(400, 'Invalid requests, e.g. wrong owner type, bad quantity, bad search parameters')
567
568
    @api.expect(_repo_quantity_search_request_parser, validate=True)
    @api.marshal_with(_repo_quantity_values_model, skip_none=True, code=200, description='Search results send')
569
    @authenticate()
570
    def get(self, quantity: str):
571
        '''
572
573
574
575
576
577
578
579
580
581
582
583
        Retrieve quantity values from entries matching the search.

        You can use the various quantities to search/filter for. For some of the
        indexed quantities this endpoint returns aggregation information. This means
        you will be given a list of all possible values and the number of entries
        that have the certain value. You can also use these aggregations on an empty
        search to determine the possible values.

        There is no ordering and no pagination. Instead there is an 'after' key based
        scrolling. The result will contain an 'after' value, that can be specified
        for the next request. You can use the 'size' and 'after' parameters accordingly.

584
585
586
        The result will contain a 'quantity' key with quantity values and the "after"
        value. There will be upto 'size' many values. For the rest of the values use the
        "after" parameter in another request.
587
        '''
588

589
        search_request = search.SearchRequest()
590
591
        args = {
            key: value
592
            for key, value in _repo_quantity_search_request_parser.parse_args().items()
593
            if value is not None}
594

595
        apply_search_parameters(search_request, args)
596
597
        after = args.get('after', None)
        size = args.get('size', 100)
598
599
600
601
602
603
604

        try:
            assert size >= 0
        except AssertionError:
            abort(400, message='invalid size')

        try:
605
            search_request.quantity(quantity, size=size, after=after)
606
607
608
            results = search_request.execute()
            quantities = results.pop('quantities')
            results['quantity'] = quantities[quantity]
609
610
611
612
613
614

            return results, 200
        except KeyError as e:
            import traceback
            traceback.print_exc()
            abort(400, 'Given quantity does not exist: %s' % str(e))
615
616


617
618
619
_repo_quantities_search_request_parser = api.parser()
add_search_parameters(_repo_quantities_search_request_parser)
_repo_quantities_search_request_parser.add_argument(
620
621
    'quantities', type=str, action='append',
    help='The quantities to retrieve values from')
622
_repo_quantities_search_request_parser.add_argument(
623
624
    'size', type=int, help='The max size of the returned values.')

Markus Scheidgen's avatar
Markus Scheidgen committed
625
_repo_quantities_model = api.model('RepoQuantities', {
626
627
628
    'quantities': fields.List(fields.Nested(_repo_quantity_model))
})

629
630
631
632
633

@ns.route('/quantities')
class RepoQuantitiesResource(Resource):
    @api.doc('quantities_search')
    @api.response(400, 'Invalid requests, e.g. wrong owner type, bad quantity, bad search parameters')
634
    @api.expect(_repo_quantities_search_request_parser, validate=True)
Markus Scheidgen's avatar
Markus Scheidgen committed
635
    @api.marshal_with(_repo_quantities_model, skip_none=True, code=200, description='Search results send')
636
637
    @authenticate()
    def get(self):
638
        '''
639
640
641
642
643
644
645
646
647
648
649
650
651
        Retrieve quantity values for multiple quantities at once.

        You can use the various quantities to search/filter for. For some of the
        indexed quantities this endpoint returns aggregation information. This means
        you will be given a list of all possible values and the number of entries
        that have the certain value. You can also use these aggregations on an empty
        search to determine the possible values.

        There is no ordering and no pagination and not after key based scrolling. Instead
        there is an 'after' key based scrolling.

        The result will contain a 'quantities' key with a dict of quantity names and the
        retrieved values as values.
652
        '''
653
654
655
656

        search_request = search.SearchRequest()
        args = {
            key: value
657
            for key, value in _repo_quantities_search_request_parser.parse_args().items()
658
659
            if value is not None}

660
        apply_search_parameters(search_request, args)
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
        quantities = args.get('quantities', [])
        size = args.get('size', 5)

        try:
            assert size >= 0
        except AssertionError:
            abort(400, message='invalid size')

        for quantity in quantities:
            try:
                search_request.quantity(quantity, size=size)
            except KeyError as e:
                import traceback
                traceback.print_exc()
                abort(400, 'Given quantity does not exist: %s' % str(e))

        return search_request.execute(), 200


Markus Scheidgen's avatar
Markus Scheidgen committed
680
_repo_calc_id_model = api.model('RepoCalculationId', {
681
682
683
684
    'upload_id': fields.String(), 'calc_id': fields.String()
})


685
@ns.route('/pid/<path:pid>')
686
687
688
class RepoPidResource(Resource):
    @api.doc('resolve_pid')
    @api.response(404, 'Entry with PID does not exist')
Markus Scheidgen's avatar
Markus Scheidgen committed
689
    @api.marshal_with(_repo_calc_id_model, skip_none=True, code=200, description='Entry resolved')
Markus Scheidgen's avatar
Markus Scheidgen committed
690
    @authenticate()
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
    def get(self, pid: str):
        if '/' in pid:
            prefix, pid = pid.split('/')
            if prefix != '21.11132':
                abort(400, 'Wrong PID format')
            try:
                pid_int = utils.decode_handle_id(pid)
            except ValueError:
                abort(400, 'Wrong PID format')
        else:
            try:
                pid_int = int(pid)
            except ValueError:
                abort(400, 'Wrong PID format')

706
        search_request = search.SearchRequest().include('upload_id', 'calc_id')
707
708
709

        if g.user is not None:
            search_request.owner('all', user_id=g.user.user_id)
710
        else:
711
712
            search_request.owner('all')

713
        search_request.search_parameter('pid', pid_int)
714
715
716
717
718

        results = list(search_request.execute_scan())
        total = len(results)

        if total == 0:
719
            abort(404, 'Entry with PID %s does not exist' % pid)
720
721

        if total > 1:
722
            common.logger.error('Two entries for the same pid', pid=pid_int)
723
724
725
726
727

        result = results[0]
        return dict(
            upload_id=result['upload_id'],
            calc_id=result['calc_id'])