data.py 20.9 KB
Newer Older
Markus Scheidgen's avatar
Markus Scheidgen committed
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
# 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.

"""
This module comprises a set of persistent document classes that hold all user related
data. These are information about users, their uploads and datasets, the associated
calculations, and files


.. autoclass:: Calc
    :members:
.. autoclass:: Upload
    :members:
"""

27
from typing import List, Any, ContextManager, Tuple, Generator
28
from elasticsearch.exceptions import NotFoundError
Markus Scheidgen's avatar
Markus Scheidgen committed
29
from mongoengine import StringField, BooleanField, DateTimeField, DictField, IntField
Markus Scheidgen's avatar
Markus Scheidgen committed
30
import logging
Markus Scheidgen's avatar
Markus Scheidgen committed
31
import time
32
from structlog import wrap_logger
33
from contextlib import contextmanager
Markus Scheidgen's avatar
Markus Scheidgen committed
34

35
from nomad import utils, coe_repo
Markus Scheidgen's avatar
Markus Scheidgen committed
36
from nomad.files import UploadFile, ArchiveFile, ArchiveLogFile, File
Markus Scheidgen's avatar
Markus Scheidgen committed
37
from nomad.repo import RepoCalc
38
from nomad.processing.base import Proc, Chord, process, task, PENDING, SUCCESS, FAILURE, RUNNING
Markus Scheidgen's avatar
Markus Scheidgen committed
39
from nomad.parsing import parsers, parser_dict
Markus Scheidgen's avatar
Markus Scheidgen committed
40
from nomad.normalizing import normalizers
Markus Scheidgen's avatar
Markus Scheidgen committed
41
from nomad.utils import lnr
Markus Scheidgen's avatar
Markus Scheidgen committed
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


class NotAllowedDuringProcessing(Exception): pass


class Calc(Proc):
    """
    Instances of this class represent calculations. This class manages the elastic
    search index entry, files, and archive for the respective calculation.

    It also contains the calculations processing and its state.

    The attribute list, does not include the various repository properties generated
    while parsing, including ``program_name``, ``program_version``, etc.

    Attributes:
        archive_id: the hash based archive id of the calc
        parser: the name of the parser used to process this calc
        upload_id: the id of the upload used to create this calculation
        mainfile: the mainfile (including path in upload) that was used to create this calc
        mainfile_tmp_path: path to the mainfile extracted for processing
    """
    archive_id = StringField(primary_key=True)
    upload_id = StringField()
    mainfile = StringField()
    parser = StringField()
    mainfile_tmp_path = StringField()

    meta: Any = {
        'indices': [
72
            'upload_id', 'mainfile', 'code', 'parser', 'status'
Markus Scheidgen's avatar
Markus Scheidgen committed
73
74
75
76
77
78
79
        ]
    }

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self._parser_backend = None
        self._upload = None
80
        self._calc_proc_logwriter = None
Markus Scheidgen's avatar
Markus Scheidgen committed
81
        self._calc_proc_logfile = None
82
        self._calc_proc_logwriter_ctx: ContextManager = None
Markus Scheidgen's avatar
Markus Scheidgen committed
83
84
85
86
87

    @classmethod
    def get(cls, id):
        return cls.get_by_id(id, 'archive_id')

Markus Scheidgen's avatar
Markus Scheidgen committed
88
89
90
91
    @property
    def mainfile_file(self) -> File:
        return File(self.mainfile_tmp_path)

92
93
94
95
96
97
    @property
    def upload(self) -> 'Upload':
        if not self._upload:
            self._upload = Upload.get(self.upload_id)
        return self._upload

Markus Scheidgen's avatar
Markus Scheidgen committed
98
99
100
101
    def delete(self):
        """
        Delete this calculation and all associated data. This includes all files,
        the archive, and this search index entry.
102
        TODO is this needed? Or do we always delete hole uploads in bulk.
Markus Scheidgen's avatar
Markus Scheidgen committed
103
104
105
        """
        # delete the archive
        if self.archive_id is not None:
106
            ArchiveFile(self.archive_id).delete()
Markus Scheidgen's avatar
Markus Scheidgen committed
107
108

        # delete the search index entry
109
110
111
112
113
114
        try:
            elastic_entry = RepoCalc.get(self.archive_id)
            if elastic_entry is not None:
                elastic_entry.delete()
        except NotFoundError:
            pass
Markus Scheidgen's avatar
Markus Scheidgen committed
115
116
117
118
119
120
121
122
123

        # delete this mongo document
        super().delete()

    def get_logger(self, **kwargs):
        upload_hash, calc_hash = self.archive_id.split('/')
        logger = super().get_logger()
        logger = logger.bind(
            upload_id=self.upload_id, mainfile=self.mainfile,
124
            upload_hash=upload_hash, calc_hash=calc_hash,
Markus Scheidgen's avatar
Markus Scheidgen committed
125
            archive_id=self.archive_id, **kwargs)
126

Markus Scheidgen's avatar
Markus Scheidgen committed
127
128
        return logger

129
130
131
132
133
134
135
136
    def get_calc_logger(self, **kwargs):
        """
        Returns a wrapped logger that additionally saves all entries to the calculation
        processing log in the archive.
        """
        logger = self.get_logger(**kwargs)

        if self._calc_proc_logwriter is None:
Markus Scheidgen's avatar
Markus Scheidgen committed
137
            self._calc_proc_logfile = ArchiveLogFile(self.archive_id)
138
139
            self._calc_proc_logwriter_ctx = self._calc_proc_logfile.open('wt')
            self._calc_proc_logwriter = self._calc_proc_logwriter_ctx.__enter__()  # pylint: disable=E1101
140

141
        def save_to_calc_log(logger, method_name, event_dict):
142
143
144
145
146
147
148
149
150
151
152
            program = event_dict.get('normalizer', 'parser')
            event = event_dict.get('event', '')
            entry = '[%s] %s: %s' % (method_name, program, event)
            if len(entry) > 120:
                self._calc_proc_logwriter.write(entry[:120])
                self._calc_proc_logwriter.write('...')
            else:
                self._calc_proc_logwriter.write(entry)
            self._calc_proc_logwriter.write('\n')
            return event_dict

153
        return wrap_logger(logger, processors=[save_to_calc_log])
154

Markus Scheidgen's avatar
Markus Scheidgen committed
155
156
    @process
    def process(self):
157
        logger = self.get_logger()
158
        if self.upload is None:
159
            logger.error('calculation upload does not exist')
Markus Scheidgen's avatar
Markus Scheidgen committed
160
161
162
163
164
165

        try:
            self.parsing()
            self.normalizing()
            self.archiving()
        finally:
Markus Scheidgen's avatar
Markus Scheidgen committed
166
            # close loghandler that was not closed due to failures
167
            try:
168
169
170
                if self._calc_proc_logwriter is not None:
                    self._calc_proc_logwriter.close()
                    self._calc_proc_logwriter = None
171
172
173
174
            except Exception as e:
                logger.error('could not close calculation proc log', exc_info=e)

            # inform parent proc about completion
175
            self.upload.completed_child()
Markus Scheidgen's avatar
Markus Scheidgen committed
176
177
178

    @task
    def parsing(self):
179
180
        context = dict(parser=self.parser, step=self.parser)
        logger = self.get_calc_logger(**context)
181
        parser = parser_dict[self.parser]
Markus Scheidgen's avatar
Markus Scheidgen committed
182

183
        with utils.timer(logger, 'parser executed', input_size=self.mainfile_file.size):
184
            self._parser_backend = parser.run(self.mainfile_tmp_path, logger=logger)
Markus Scheidgen's avatar
Markus Scheidgen committed
185

186
187
188
189
190
191
        self._parser_backend.openNonOverlappingSection('section_calculation_info')
        self._parser_backend.addValue('upload_id', self.upload_id)
        self._parser_backend.addValue('archive_id', self.archive_id)
        self._parser_backend.addValue('main_file', self.mainfile)
        self._parser_backend.addValue('parser_name', self.parser)

Markus Scheidgen's avatar
Markus Scheidgen committed
192
        if self._parser_backend.status[0] != 'ParseSuccess':
193
            logger.error(self._parser_backend.status[1])
Markus Scheidgen's avatar
Markus Scheidgen committed
194
            error = self._parser_backend.status[1]
195
            self._parser_backend.addValue('parse_status', 'ParseFailure')
196
            self.fail(error, level=logging.DEBUG, **context)
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
        else:
            self._parser_backend.addValue('parse_status', 'ParseSuccess')

        self._parser_backend.closeNonOverlappingSection('section_calculation_info')

        self.add_processor_info(self.parser)

    @contextmanager
    def use_parser_backend(self, processor_name):
        self._parser_backend.reset_status()
        yield self._parser_backend
        self.add_processor_info(processor_name)

    def add_processor_info(self, processor_name: str) -> None:
        self._parser_backend.openContext('/section_calculation_info/0')
        self._parser_backend.openNonOverlappingSection('section_archive_processing_info')
        self._parser_backend.addValue('archive_processor_name', processor_name)

        if self._parser_backend.status[0] == 'ParseSuccess':
            warnings = getattr(self._parser_backend, '_warnings', [])
            if len(warnings) > 0:
                self._parser_backend.addValue('archive_processor_status', 'WithWarnings')
                self._parser_backend.addValue('archive_processor_warning_number', len(warnings))
                self._parser_backend.addArrayValues('archive_processor_warnings', [str(warning) for warning in warnings])
            else:
                self._parser_backend.addValue('archive_processor_status', 'Success')
        else:
            errors = self._parser_backend.status[1]
            self._parser_backend.addValue('archive_processor_error', str(errors))

        self._parser_backend.closeNonOverlappingSection('section_archive_processing_info')
        self._parser_backend.closeContext('/section_calculation_info/0')
Markus Scheidgen's avatar
Markus Scheidgen committed
229
230
231
232
233

    @task
    def normalizing(self):
        for normalizer in normalizers:
            normalizer_name = normalizer.__name__
234
235
            context = dict(normalizer=normalizer_name, step=normalizer_name)
            logger = self.get_calc_logger(**context)
Markus Scheidgen's avatar
Markus Scheidgen committed
236
237

            with utils.timer(
238
                    logger, 'normalizer executed', input_size=self.mainfile_file.size):
239
240
                with self.use_parser_backend(normalizer_name) as backend:
                    normalizer(backend).normalize(logger=logger)
Markus Scheidgen's avatar
Markus Scheidgen committed
241

242
243
            failed = self._parser_backend.status[0] != 'ParseSuccess'
            if failed:
244
                logger.error(self._parser_backend.status[1])
Markus Scheidgen's avatar
Markus Scheidgen committed
245
                error = self._parser_backend.status[1]
246
                self.fail(error, level=logging.WARNING, **context)
247
248
249
250
                break
            else:
                logger.debug(
                    'completed normalizer successfully', normalizer=normalizer_name)
Markus Scheidgen's avatar
Markus Scheidgen committed
251
252
253

    @task
    def archiving(self):
254
255
256
257
258
        logger = self.get_logger()

        upload_hash, calc_hash = self.archive_id.split('/')
        additional = dict(
            mainfile=self.mainfile,
259
            upload_time=self.upload.upload_time,
260
261
            staging=True,
            restricted=False,
262
263
            user_id=self.upload.user_id,
            aux_files=list(self.upload.upload_file.get_siblings(self.mainfile)))
264
265

        with utils.timer(logger, 'indexed', step='index'):
266
267
268
269
270
271
            # persist to elastic search
            RepoCalc.create_from_backend(
                self._parser_backend,
                additional=additional,
                upload_hash=upload_hash,
                calc_hash=calc_hash,
272
                upload_id=self.upload_id).persist()
273

Markus Scheidgen's avatar
Markus Scheidgen committed
274
275
276
277
        with utils.timer(
                logger, 'archived', step='archive',
                input_size=self.mainfile_file.size) as log_data:

278
            # persist the archive
Markus Scheidgen's avatar
Markus Scheidgen committed
279
280
            archive_file = ArchiveFile(self.archive_id)
            with archive_file.write_archive_json() as out:
281
282
                self._parser_backend.write_json(out, pretty=True)

Markus Scheidgen's avatar
Markus Scheidgen committed
283
284
285
286
287
288
289
            log_data.update(archive_size=archive_file.size)

        # close loghandler
        if self._calc_proc_logwriter is not None:
            with utils.timer(
                    logger, 'archived log', step='archive_log',
                    input_size=self.mainfile_file.size) as log_data:
290
                self._calc_proc_logwriter_ctx.__exit__(None, None, None)  # pylint: disable=E1101
291
                self._calc_proc_logwriter = None
Markus Scheidgen's avatar
Markus Scheidgen committed
292

Markus Scheidgen's avatar
Markus Scheidgen committed
293
294
                log_data.update(log_size=self._calc_proc_logfile.size)

Markus Scheidgen's avatar
Markus Scheidgen committed
295

296
class Upload(Chord):
Markus Scheidgen's avatar
Markus Scheidgen committed
297
298
299
300
301
302
    """
    Represents uploads in the databases. Provides persistence access to the files storage,
    and processing state.

    Attributes:
        name: optional user provided upload name
303
        local_path: optional local path, e.g. for files that are already somewhere on the server
Markus Scheidgen's avatar
Markus Scheidgen committed
304
305
306
307
308
309
        additional_metadata: optional user provided additional meta data
        upload_id: the upload id generated by the database
        in_staging: true if the upload is still in staging and can be edited by the uploader
        is_private: true if the upload and its derivitaves are only visible to the uploader
        upload_time: the timestamp when the system realised the upload
        upload_hash: the hash of the uploaded file
310
        user_id: the id of the user that created this upload
Markus Scheidgen's avatar
Markus Scheidgen committed
311
312
313
314
315
316
    """
    id_field = 'upload_id'

    upload_id = StringField(primary_key=True)

    name = StringField(default=None)
317
    local_path = StringField(default=None)
Markus Scheidgen's avatar
Markus Scheidgen committed
318
319
320
321
322
323
324
325
    additional_metadata = DictField(default=None)

    in_staging = BooleanField(default=True)
    is_private = BooleanField(default=False)

    upload_time = DateTimeField()
    upload_hash = StringField(default=None)

326
    user_id = StringField(required=True)
Markus Scheidgen's avatar
Markus Scheidgen committed
327

328
329
    coe_repo_upload_id = IntField(default=None)

330
331
    _initiated_parsers = IntField(default=-1)

Markus Scheidgen's avatar
Markus Scheidgen committed
332
333
    meta: Any = {
        'indexes': [
334
            'upload_hash', 'user_id', 'status'
Markus Scheidgen's avatar
Markus Scheidgen committed
335
336
337
338
339
        ]
    }

    def __init__(self, **kwargs):
        super().__init__(**kwargs)
340
        self._upload_file = None
Markus Scheidgen's avatar
Markus Scheidgen committed
341
342
343
344
345
346

    @classmethod
    def get(cls, id):
        return cls.get_by_id(id, 'upload_id')

    @classmethod
347
    def user_uploads(cls, user: coe_repo.User) -> List['Upload']:
Markus Scheidgen's avatar
Markus Scheidgen committed
348
        """ Returns all uploads for the given user. Currently returns all uploads. """
349
        return cls.objects(user_id=str(user.user_id), in_staging=True)
Markus Scheidgen's avatar
Markus Scheidgen committed
350
351
352
353
354
355
356
357
358

    def get_logger(self, **kwargs):
        logger = super().get_logger()
        logger = logger.bind(upload_id=self.upload_id, **kwargs)
        return logger

    def delete(self):
        logger = self.get_logger(task='delete')

359
        if not (self.completed or self.current_task == 'uploading'):
Markus Scheidgen's avatar
Markus Scheidgen committed
360
361
362
363
            raise NotAllowedDuringProcessing()

        with lnr(logger, 'delete upload file'):
            try:
364
                UploadFile(self.upload_id, local_path=self.local_path).delete()
Markus Scheidgen's avatar
Markus Scheidgen committed
365
366
367
368
369
370
371
372
373
            except KeyError:
                if self.current_task == 'uploading':
                    logger.debug(
                        'Upload exist, but file does not exist. '
                        'It was probably aborted and deleted.')
                else:
                    logger.debug('Upload exist, but uploaded file does not exist.')

        with lnr(logger, 'deleting calcs'):
374
            # delete archive files
375
            ArchiveFile.delete_archives(upload_hash=self.upload_hash)
376
377

            # delete repo entries
Markus Scheidgen's avatar
Markus Scheidgen committed
378
            RepoCalc.delete_upload(upload_id=self.upload_id)
379
380
381

            # delete calc processings
            Calc.objects(upload_id=self.upload_id).delete()
Markus Scheidgen's avatar
Markus Scheidgen committed
382
383
384
385
386
387
388
389
390
391

        with lnr(logger, 'deleting upload'):
            super().delete()

    @classmethod
    def create(cls, **kwargs) -> 'Upload':
        """
        Creates a new upload for the given user, a user given name is optional.
        It will populate the record with a signed url and pending :class:`UploadProc`.
        The upload will be already saved to the database.
392
393

        Arguments:
394
            user (coe_repo.User): The user that created the upload.
Markus Scheidgen's avatar
Markus Scheidgen committed
395
        """
396
        user: coe_repo.User = kwargs['user']
397
398
399
        del(kwargs['user'])
        if 'upload_id' not in kwargs:
            kwargs.update(upload_id=utils.create_uuid())
400
        kwargs.update(user_id=str(user.user_id))
Markus Scheidgen's avatar
Markus Scheidgen committed
401
        self = super().create(**kwargs)
402

Markus Scheidgen's avatar
Markus Scheidgen committed
403
        self._continue_with('uploading')
404

Markus Scheidgen's avatar
Markus Scheidgen committed
405
406
        return self

Markus Scheidgen's avatar
Markus Scheidgen committed
407
    def unstage(self):
408
        self.get_logger().info('unstage')
409
410
411
412

        if not (self.completed or self.current_task == 'uploading'):
            raise NotAllowedDuringProcessing()

Markus Scheidgen's avatar
Markus Scheidgen committed
413
        self.in_staging = False
414
        RepoCalc.unstage(upload_id=self.upload_id)
415
        coe_repo.add_upload(self, restricted=False)  # TODO allow users to choose restricted
Markus Scheidgen's avatar
Markus Scheidgen committed
416
417
        self.save()

Markus Scheidgen's avatar
Markus Scheidgen committed
418
419
420
421
422
423
424
425
426
    @process
    def process(self):
        self.extracting()
        self.parse_all()

    @task
    def uploading(self):
        pass

427
428
429
430
431
432
433
    @property
    def upload_file(self):
        """ The :class:`UploadFile` instance that represents the uploaded file of this upload. """
        if not self._upload_file:
            self._upload_file = UploadFile(self.upload_id, local_path=self.local_path)
        return self._upload_file

Markus Scheidgen's avatar
Markus Scheidgen committed
434
435
    @task
    def extracting(self):
436
437
438
439
440
441
        """
        Task performed before the actual parsing/normalizing. Extracting and bagging
        the uploaded files, computing all keys, create an *upload* entry in the NOMAD-coe
        repository db, etc.
        """
        # extract the uploaded file, this will also create a bagit bag.
Markus Scheidgen's avatar
Markus Scheidgen committed
442
443
        logger = self.get_logger()
        try:
Markus Scheidgen's avatar
Markus Scheidgen committed
444
445
            with utils.timer(
                    logger, 'upload extracted', step='extracting',
446
447
                    upload_size=self.upload_file.size):
                self.upload_file.extract()
Markus Scheidgen's avatar
Markus Scheidgen committed
448
449
450
451
        except KeyError as e:
            self.fail('process request for non existing upload', level=logging.INFO)
            return

452
        # create and save a hash for the upload
Markus Scheidgen's avatar
Markus Scheidgen committed
453
        try:
454
            self.upload_hash = self.upload_file.upload_hash()
455
        except Exception as e:
Markus Scheidgen's avatar
Markus Scheidgen committed
456
457
458
            self.fail('could not create upload hash', e)
            return

459
        # check if the file was already uploaded and processed before
Markus Scheidgen's avatar
Markus Scheidgen committed
460
        if RepoCalc.upload_exists(self.upload_hash):
Markus Scheidgen's avatar
Markus Scheidgen committed
461
462
463
            self.fail('The same file was already uploaded and processed.', level=logging.INFO)
            return

464
465
466
467
468
469
470
471
472
473
474
475
    def match_mainfiles(self) -> Generator[Tuple[File, str, object], None, None]:
        """
        Generator function that matches all files in the upload to all parsers to
        determine the upload's mainfiles.

        Returns:
            Tuples of mainfile, filename, and parsers
        """
        for filename in self.upload_file.filelist:
            potential_mainfile = self.upload_file.get_file(filename)
            for parser in parsers:
                try:
476
                    with potential_mainfile.open('r') as mainfile_f:
477
478
479
                        if parser.is_mainfile(filename, lambda fn: mainfile_f):
                            yield potential_mainfile, filename, parser
                except Exception as e:
480
                    self.get_logger().error(
481
482
483
                        'exception while matching pot. mainfile',
                        mainfile=filename, exc_info=e)

Markus Scheidgen's avatar
Markus Scheidgen committed
484
485
    @task
    def parse_all(self):
486
        """
487
        Identified mainfile/parser combinations among the upload's files, creates
488
489
        respective :class:`Calc` instances, and triggers their processing.
        """
490
491
        logger = self.get_logger()

Markus Scheidgen's avatar
Markus Scheidgen committed
492
        # TODO: deal with multiple possible parser specs
Markus Scheidgen's avatar
Markus Scheidgen committed
493
494
        with utils.timer(
                logger, 'upload extracted', step='matching',
495
496
                upload_size=self.upload_file.size,
                upload_filecount=len(self.upload_file.filelist)):
497
            total_calcs = 0
498
499
500
501
502
503
504
505
506
            for mainfile, filename, parser in self.match_mainfiles():
                calc = Calc.create(
                    archive_id='%s/%s' % (self.upload_hash, utils.hash(filename)),
                    mainfile=filename, parser=parser.name,
                    mainfile_tmp_path=mainfile.os_path,
                    upload_id=self.upload_id)

                calc.process()
                total_calcs += 1
Markus Scheidgen's avatar
Markus Scheidgen committed
507

508
509
510
511
512
        # have to save the total_calcs information for chord management
        self.spwaned_childred(total_calcs)

    def join(self):
        self.cleanup()
Markus Scheidgen's avatar
Markus Scheidgen committed
513
514
515
516

    @task
    def cleanup(self):
        try:
Markus Scheidgen's avatar
Markus Scheidgen committed
517
            upload = UploadFile(self.upload_id, local_path=self.local_path)
518
519
520
521
522
            with utils.timer(
                    self.get_logger(), 'upload persisted', step='cleaning',
                    upload_size=upload.size):
                upload.persist()

Markus Scheidgen's avatar
Markus Scheidgen committed
523
524
525
526
            with utils.timer(
                    self.get_logger(), 'processing cleaned up', step='cleaning',
                    upload_size=upload.size):
                upload.remove_extract()
Markus Scheidgen's avatar
Markus Scheidgen committed
527
        except KeyError as e:
Markus Scheidgen's avatar
Markus Scheidgen committed
528
            self.fail('Upload does not exist', exc_info=e)
Markus Scheidgen's avatar
Markus Scheidgen committed
529
530
531
532
533
            return

        self.get_logger().debug('closed upload')

    @property
534
535
536
537
538
539
540
541
542
543
544
    def processed_calcs(self):
        return Calc.objects(upload_id=self.upload_id, status__in=[SUCCESS, FAILURE]).count()

    @property
    def total_calcs(self):
        return Calc.objects(upload_id=self.upload_id).count()

    @property
    def failed_calcs(self):
        return Calc.objects(upload_id=self.upload_id, status=FAILURE).count()

545
546
547
548
    @property
    def pending_calcs(self):
        return Calc.objects(upload_id=self.upload_id, status=PENDING).count()

549
550
    def all_calcs(self, start, end, order_by='mainfile'):
        return Calc.objects(upload_id=self.upload_id)[start:end].order_by(order_by)
551

Markus Scheidgen's avatar
Markus Scheidgen committed
552
    @staticmethod
553
554
555
556
557
    def repair_all():
        """
        Utitlity function that will look for suspiciously looking conditions in
        all uncompleted downloads. It ain't a perfect world.
        """
558
559
560
        # TODO this was added as a quick fix to #37.
        # Even though it might be strictly necessary, there should be a tested backup
        # solution for it Chords to not work properly due to failed to fail processings
561
562
563
564
565
566
567
568
569
570
        uploads = Upload.objects(status__in=[PENDING, RUNNING])
        for upload in uploads:
            completed = upload.processed_calcs
            total = upload.total
            pending = upload.pending_calcs

            if completed + pending == total:
                time.sleep(2)
                if pending == upload.pending_calcs:
                    Calc.objects(upload_id=upload.upload_id, status=PENDING).delete()