data.py 25.3 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, Dict
28
from mongoengine import StringField, DateTimeField, DictField, BooleanField
Markus Scheidgen's avatar
Markus Scheidgen committed
29
import logging
30
from structlog import wrap_logger
31
from contextlib import contextmanager
32
import os.path
Markus Scheidgen's avatar
Markus Scheidgen committed
33

34
from nomad import utils, coe_repo, config, infrastructure, search
Markus Scheidgen's avatar
Markus Scheidgen committed
35
from nomad.files import PathObject, UploadFiles, ExtractError, ArchiveBasedStagingUploadFiles
36
from nomad.processing.base import Proc, process, task, PENDING, SUCCESS, FAILURE
37
from nomad.parsing import parser_dict, match_parser
Markus Scheidgen's avatar
Markus Scheidgen committed
38
from nomad.normalizing import normalizers
Markus Scheidgen's avatar
Markus Scheidgen committed
39
from nomad.datamodel import UploadWithMetadata, CalcWithMetadata
Markus Scheidgen's avatar
Markus Scheidgen committed
40
41


Markus Scheidgen's avatar
Markus Scheidgen committed
42
class Calc(Proc):
Markus Scheidgen's avatar
Markus Scheidgen committed
43
44
45
46
47
48
49
50
51
52
    """
    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:
53
        calc_id: the calc_id of this calc
Markus Scheidgen's avatar
Markus Scheidgen committed
54
55
56
57
        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
    """
58
    calc_id = StringField(primary_key=True)
Markus Scheidgen's avatar
Markus Scheidgen committed
59
60
61
62
    upload_id = StringField()
    mainfile = StringField()
    parser = StringField()

63
64
    queue = 'calcs'

Markus Scheidgen's avatar
Markus Scheidgen committed
65
    meta: Any = {
Markus Scheidgen's avatar
Markus Scheidgen committed
66
67
        'indexes': [
            'upload_id', 'mainfile', 'parser', 'tasks_status'
Markus Scheidgen's avatar
Markus Scheidgen committed
68
69
70
71
72
73
        ]
    }

    def __init__(self, *args, **kwargs):
        super().__init__(*args, **kwargs)
        self._parser_backend = None
74
75
        self._upload: Upload = None
        self._upload_files: ArchiveBasedStagingUploadFiles = None
76
        self._calc_proc_logwriter = None
77
        self._calc_proc_logwriter_ctx: ContextManager = None
Markus Scheidgen's avatar
Markus Scheidgen committed
78
79
80

    @classmethod
    def get(cls, id):
81
        return cls.get_by_id(id, 'calc_id')
Markus Scheidgen's avatar
Markus Scheidgen committed
82

Markus Scheidgen's avatar
Markus Scheidgen committed
83
    @property
84
85
    def mainfile_file(self) -> PathObject:
        return self.upload_files.raw_file_object(self.mainfile)
Markus Scheidgen's avatar
Markus Scheidgen committed
86

87
88
89
90
    @property
    def upload(self) -> 'Upload':
        if not self._upload:
            self._upload = Upload.get(self.upload_id)
91
            self._upload.worker_hostname = self.worker_hostname
92
93
        return self._upload

94
95
96
    @property
    def upload_files(self) -> ArchiveBasedStagingUploadFiles:
        if not self._upload_files:
97
98
            self._upload_files = ArchiveBasedStagingUploadFiles(
                self.upload_id, is_authorized=lambda: True, upload_path=self.upload.upload_path)
99
100
        return self._upload_files

Markus Scheidgen's avatar
Markus Scheidgen committed
101
    def get_logger(self, **kwargs):
102
103
104
105
        """
        Returns a wrapped logger that additionally saves all entries to the calculation
        processing log in the archive.
        """
106
107
108
        logger = super().get_logger()
        logger = logger.bind(
            upload_id=self.upload_id, mainfile=self.mainfile, calc_id=self.calc_id, **kwargs)
109

110
        if self._calc_proc_logwriter_ctx is None:
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
            try:
                self._calc_proc_logwriter_ctx = self.upload_files.archive_log_file(self.calc_id, 'wt')
                self._calc_proc_logwriter = self._calc_proc_logwriter_ctx.__enter__()  # pylint: disable=E1101
            except KeyError:
                # cannot open log file
                pass

        if self._calc_proc_logwriter_ctx is None:
            return logger
        else:
            def save_to_calc_log(logger, method_name, event_dict):
                if self._calc_proc_logwriter is not None:
                    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

            return wrap_logger(logger, processors=[save_to_calc_log])
136

Markus Scheidgen's avatar
Markus Scheidgen committed
137
    @process
138
    def process_calc(self):
139
        logger = self.get_logger()
140
        if self.upload is None:
141
            logger.error('calculation upload does not exist')
Markus Scheidgen's avatar
Markus Scheidgen committed
142
143
144
145
146
147

        try:
            self.parsing()
            self.normalizing()
            self.archiving()
        finally:
Markus Scheidgen's avatar
Markus Scheidgen committed
148
            # close loghandler that was not closed due to failures
149
            try:
150
151
152
                if self._calc_proc_logwriter is not None:
                    self._calc_proc_logwriter.close()
                    self._calc_proc_logwriter = None
153
154
155
            except Exception as e:
                logger.error('could not close calculation proc log', exc_info=e)

156
157
158
159
160
161
162
    def on_process_complete(self, process_name):
        # the save might be necessary to correctly read the join condition from the db
        self.save()
        # in case of error, the process_name might be unknown
        if process_name == 'process_calc' or process_name is None:
            self.upload.reload()
            self.upload.check_join()
Markus Scheidgen's avatar
Markus Scheidgen committed
163
164
165

    @task
    def parsing(self):
166
        context = dict(parser=self.parser, step=self.parser)
167
        logger = self.get_logger(**context)
168
        parser = parser_dict[self.parser]
Markus Scheidgen's avatar
Markus Scheidgen committed
169

170
        with utils.timer(logger, 'parser executed', input_size=self.mainfile_file.size):
171
172
173
174
175
176
177
            try:
                self._parser_backend = parser.run(
                    self.upload_files.raw_file_object(self.mainfile).os_path, logger=logger)
            except Exception as e:
                self.fail(
                    'parser failed with exception', level=logging.ERROR,
                    exc_info=e, error=str(e), **context)
178
                return
Markus Scheidgen's avatar
Markus Scheidgen committed
179

180
181
        self._parser_backend.openNonOverlappingSection('section_calculation_info')
        self._parser_backend.addValue('upload_id', self.upload_id)
182
183
        self._parser_backend.addValue('calc_id', self.calc_id)
        self._parser_backend.addValue('calc_hash', self.upload_files.calc_hash(self.mainfile))
184
185
186
        self._parser_backend.addValue('main_file', self.mainfile)
        self._parser_backend.addValue('parser_name', self.parser)

Markus Scheidgen's avatar
Markus Scheidgen committed
187
        if self._parser_backend.status[0] != 'ParseSuccess':
188
            logger.error(self._parser_backend.status[1])
Markus Scheidgen's avatar
Markus Scheidgen committed
189
            error = self._parser_backend.status[1]
190
            self._parser_backend.addValue('parse_status', 'ParseFailure')
191
            self.fail(error, level=logging.INFO, **context)
192
193
194
195
196
        else:
            self._parser_backend.addValue('parse_status', 'ParseSuccess')

        self._parser_backend.closeNonOverlappingSection('section_calculation_info')

197
198
199
200
201
202
        self._parser_backend.openNonOverlappingSection('section_repository_info')
        self._parser_backend.addValue('repository_archive_gid', '%s/%s' % (self.upload_id, self.calc_id))
        self._parser_backend.addValue(
            'repository_filepaths', self.upload_files.calc_files(self.mainfile))
        self._parser_backend.closeNonOverlappingSection('section_repository_info')

203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
        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:
225
            errors = self._parser_backend.status[1]
226
227
228
229
            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
230
231
232
233
234

    @task
    def normalizing(self):
        for normalizer in normalizers:
            normalizer_name = normalizer.__name__
235
            context = dict(normalizer=normalizer_name, step=normalizer_name)
236
            logger = self.get_logger(**context)
Markus Scheidgen's avatar
Markus Scheidgen committed
237
238

            with utils.timer(
239
                    logger, 'normalizer executed', input_size=self.mainfile_file.size):
240
                with self.use_parser_backend(normalizer_name) as backend:
241
242
243
244
245
246
247
                    try:
                        normalizer(backend).normalize(logger=logger)
                    except Exception as e:
                        self.fail(
                            'normalizer failed with exception', level=logging.ERROR,
                            exc_info=e, error=str(e), **context)
                        self._parser_backend.status = ['ParseFailure', str(e)]
Markus Scheidgen's avatar
Markus Scheidgen committed
248

249
250
            failed = self._parser_backend.status[0] != 'ParseSuccess'
            if failed:
251
                logger.error(self._parser_backend.status[1])
Markus Scheidgen's avatar
Markus Scheidgen committed
252
                error = self._parser_backend.status[1]
253
                self.fail(error, level=logging.WARNING, error=error, **context)
254
255
256
257
                break
            else:
                logger.debug(
                    'completed normalizer successfully', normalizer=normalizer_name)
Markus Scheidgen's avatar
Markus Scheidgen committed
258
259
260

    @task
    def archiving(self):
261
262
        logger = self.get_logger()

263
264
        calc_with_metadata = self._parser_backend.to_calc_with_metadata()

265
        # persist the repository metadata
266
        with utils.timer(logger, 'saved repo metadata', step='metadata'):
267
268
269
            self.upload_files.metadata.insert(calc_with_metadata.to_dict())

        # index in search
270
        with utils.timer(logger, 'indexed', step='index'):
271
272
            calc_with_metadata.update(published=False, uploader=self.upload.uploader.to_popo())
            search.Entry.from_calc_with_metadata(calc_with_metadata).save()
273

274
        # persist the archive
Markus Scheidgen's avatar
Markus Scheidgen committed
275
        with utils.timer(
276
                logger, 'archived', step='archive',
Markus Scheidgen's avatar
Markus Scheidgen committed
277
                input_size=self.mainfile_file.size) as log_data:
278
            with self.upload_files.archive_file(self.calc_id, 'wt') as out:
279
280
                self._parser_backend.write_json(out, pretty=True)

281
            log_data.update(archive_size=self.upload_files.archive_file_object(self.calc_id).size)
Markus Scheidgen's avatar
Markus Scheidgen committed
282
283
284
285

        # close loghandler
        if self._calc_proc_logwriter is not None:
            with utils.timer(
286
                    logger, 'archived log', step='logs',
Markus Scheidgen's avatar
Markus Scheidgen committed
287
                    input_size=self.mainfile_file.size) as log_data:
288
                self._calc_proc_logwriter_ctx.__exit__(None, None, None)  # pylint: disable=E1101
289
                self._calc_proc_logwriter = None
Markus Scheidgen's avatar
Markus Scheidgen committed
290

291
                log_data.update(log_size=self.upload_files.archive_log_file_object(self.calc_id).size)
Markus Scheidgen's avatar
Markus Scheidgen committed
292

293
    def __str__(self):
294
        return 'calc %s calc_id=%s upload_id%s' % (super().__str__(), self.calc_id, self.upload_id)
295

296

297
class Upload(Proc):
Markus Scheidgen's avatar
Markus Scheidgen committed
298
299
300
301
302
303
    """
    Represents uploads in the databases. Provides persistence access to the files storage,
    and processing state.

    Attributes:
        name: optional user provided upload name
304
305
        upload_path: the path were the uploaded files was stored
        temporary: True if the uploaded file should be removed after extraction
306
        metadata: optional user provided additional meta data
Markus Scheidgen's avatar
Markus Scheidgen committed
307
308
        upload_id: the upload id generated by the database
        upload_time: the timestamp when the system realised the upload
309
        user_id: the id of the user that created this upload
Markus Scheidgen's avatar
Markus Scheidgen committed
310
311
312
313
    """
    id_field = 'upload_id'

    upload_id = StringField(primary_key=True)
314
315
    upload_path = StringField(default=None)
    temporary = BooleanField(default=False)
Markus Scheidgen's avatar
Markus Scheidgen committed
316
317

    name = StringField(default=None)
318
    metadata = DictField(default=None)
Markus Scheidgen's avatar
Markus Scheidgen committed
319
    upload_time = DateTimeField()
320
    user_id = StringField(required=True)
Markus Scheidgen's avatar
Markus Scheidgen committed
321

322
323
    queue = 'uploads'

Markus Scheidgen's avatar
Markus Scheidgen committed
324
325
    meta: Any = {
        'indexes': [
326
            'user_id', 'tasks_status'
Markus Scheidgen's avatar
Markus Scheidgen committed
327
328
329
330
331
        ]
    }

    def __init__(self, **kwargs):
        super().__init__(**kwargs)
332
        self._upload_files: ArchiveBasedStagingUploadFiles = None
Markus Scheidgen's avatar
Markus Scheidgen committed
333
334
335
336
337
338

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

    @classmethod
339
    def user_uploads(cls, user: coe_repo.User) -> List['Upload']:
Markus Scheidgen's avatar
Markus Scheidgen committed
340
        """ Returns all uploads for the given user. Currently returns all uploads. """
341
        return cls.objects(user_id=str(user.user_id))
Markus Scheidgen's avatar
Markus Scheidgen committed
342

343
344
345
346
    @property
    def uploader(self):
        return coe_repo.User.from_user_id(self.user_id)

Markus Scheidgen's avatar
Markus Scheidgen committed
347
348
    def get_logger(self, **kwargs):
        logger = super().get_logger()
349
        logger = logger.bind(upload_id=self.upload_id, upload_name=self.name, **kwargs)
Markus Scheidgen's avatar
Markus Scheidgen committed
350
351
352
353
354
355
356
357
        return logger

    @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.
358
359

        Arguments:
360
            user (coe_repo.User): The user that created the upload.
Markus Scheidgen's avatar
Markus Scheidgen committed
361
        """
362
        user: coe_repo.User = kwargs['user']
363
364
365
        del(kwargs['user'])
        if 'upload_id' not in kwargs:
            kwargs.update(upload_id=utils.create_uuid())
366
        kwargs.update(user_id=str(user.user_id))
Markus Scheidgen's avatar
Markus Scheidgen committed
367
        self = super().create(**kwargs)
368

Markus Scheidgen's avatar
Markus Scheidgen committed
369
        self._continue_with('uploading')
370

Markus Scheidgen's avatar
Markus Scheidgen committed
371
372
        return self

373
374
375
376
377
378
379
380
381
382
383
384
385
386
    def delete(self):
        """ Deletes this upload process state entry and its calcs. """
        Calc.objects(upload_id=self.upload_id).delete()
        super().delete()

    @process
    def delete_upload(self):
        """
        Deletes of the upload, including its processing state and
        staging files.
        """
        logger = self.get_logger()

        with utils.lnr(logger, 'staged upload delete failed'):
387
388
389
390
391
392

            with utils.timer(
                    logger, 'upload deleted from repo db', step='repo',
                    upload_size=self.upload_files.size):
                coe_repo.Upload.delete(self.upload_id)

393
            with utils.timer(
394
                    logger, 'upload deleted from index', step='index',
395
                    upload_size=self.upload_files.size):
396
                search.delete_upload(self.upload_id)
397

398
            with utils.timer(
399
                    logger, 'staged upload deleted', step='files',
400
401
402
403
404
                    upload_size=self.upload_files.size):
                self.upload_files.delete()
                self.delete()

        return True  # do not save the process status on the delete upload
405

406
    @process
407
    def publish_upload(self):
408
409
410
411
412
413
414
        """
        Moves the upload out of staging to add it to the coe repository. It will
        pack the staging upload files in to public upload files, add entries to the
        coe repository db and remove this instance and its calculation from the
        processing state db.
        """
        logger = self.get_logger()
Markus Scheidgen's avatar
Markus Scheidgen committed
415
        logger.info('started to publish', step='publish')
416

417
        with utils.lnr(logger, 'publish failed'):
418
419
            upload_with_metadata = self.to_upload_with_metadata()

420
            with utils.timer(
421
                    logger, 'upload added to repository', step='repo',
422
                    upload_size=self.upload_files.size):
423
                coe_repo.Upload.publish(upload_with_metadata)
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
            with utils.timer(
                    logger, 'staged upload files packed', step='pack',
                    upload_size=self.upload_files.size):
                coe_upload = coe_repo.Upload.from_upload_id(upload_with_metadata.upload_id)
                if coe_upload is not None:
                    for coe_calc in coe_upload.calcs:
                        calc_metadata = coe_calc.to_calc_with_metadata()
                        calc_metadata.published = True
                        self.upload_files.metadata.update(
                            calc_id=calc_metadata.calc_id, updates=calc_metadata.to_dict())
                    logger.info('metadata updated after publish to coe repo', step='publish')

                self.upload_files.pack()

            with utils.timer(
                    logger, 'index updated', step='index',
                    upload_size=self.upload_files.size):
                coe_upload = coe_repo.Upload.from_upload_id(upload_with_metadata.upload_id)
                if coe_upload is not None:
                    search.publish(
                        [coe_calc.to_calc_with_metadata() for coe_calc in coe_upload.calcs])

            with utils.timer(
                    logger, 'staged upload deleted', step='delete staged',
                    upload_size=self.upload_files.size):
                self.upload_files.delete()
                self.delete()
Markus Scheidgen's avatar
Markus Scheidgen committed
452

453
        return True  # do not save the process status on the delete upload
454

Markus Scheidgen's avatar
Markus Scheidgen committed
455
    @process
456
    def process_upload(self):
Markus Scheidgen's avatar
Markus Scheidgen committed
457
458
459
460
461
462
463
        self.extracting()
        self.parse_all()

    @task
    def uploading(self):
        pass

464
    @property
465
466
    def upload_files(self) -> ArchiveBasedStagingUploadFiles:
        if not self._upload_files:
467
468
            self._upload_files = ArchiveBasedStagingUploadFiles(
                self.upload_id, is_authorized=lambda: True, upload_path=self.upload_path)
469
        return self._upload_files
470

Markus Scheidgen's avatar
Markus Scheidgen committed
471
472
    @task
    def extracting(self):
473
474
475
476
477
        """
        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.
        """
478
479
480
481
482
        # extract the uploaded file
        self._upload_files = ArchiveBasedStagingUploadFiles(
            upload_id=self.upload_id, is_authorized=lambda: True, create=True,
            upload_path=self.upload_path)

Markus Scheidgen's avatar
Markus Scheidgen committed
483
484
        logger = self.get_logger()
        try:
Markus Scheidgen's avatar
Markus Scheidgen committed
485
486
            with utils.timer(
                    logger, 'upload extracted', step='extracting',
487
488
                    upload_size=self.upload_files.size):
                self.upload_files.extract()
489
490
491
492

            if self.temporary:
                os.remove(self.upload_path)
                self.upload_path = None
493
        except KeyError:
494
            self.fail('processing requested for non existing upload', log_level=logging.ERROR)
495
496
497
            return
        except ExtractError:
            self.fail('bad .zip/.tar file', log_level=logging.INFO)
Markus Scheidgen's avatar
Markus Scheidgen committed
498
499
            return

500
    def match_mainfiles(self) -> Generator[Tuple[str, object], None, None]:
501
502
503
504
505
506
507
        """
        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
        """
508
        directories_with_match: Dict[str, str] = dict()
509
        for filename in self.upload_files.raw_file_manifest():
510
            try:
511
                parser = match_parser(filename, self.upload_files)
512
                if parser is not None:
513
514
                    directory = os.path.dirname(filename)
                    if directory in directories_with_match:
515
516
517
                        # TODO this might give us the chance to store directory based relationship
                        # between calcs for the future?
                        pass
518
519
                    else:
                        directories_with_match[directory] = filename
520
521

                    yield filename, parser
522
523
524
525
            except Exception as e:
                self.get_logger().error(
                    'exception while matching pot. mainfile',
                    mainfile=filename, exc_info=e)
526

Markus Scheidgen's avatar
Markus Scheidgen committed
527
528
    @task
    def parse_all(self):
529
        """
530
        Identified mainfile/parser combinations among the upload's files, creates
531
532
        respective :class:`Calc` instances, and triggers their processing.
        """
533
534
        logger = self.get_logger()

Markus Scheidgen's avatar
Markus Scheidgen committed
535
536
        with utils.timer(
                logger, 'upload extracted', step='matching',
537
538
                upload_size=self.upload_files.size):
            for filename, parser in self.match_mainfiles():
539
                calc = Calc.create(
540
                    calc_id=self.upload_files.calc_id(filename),
541
                    mainfile=filename, parser=parser.name,
542
                    worker_hostname=self.worker_hostname,
543
544
                    upload_id=self.upload_id)

545
                calc.process_calc()
Markus Scheidgen's avatar
Markus Scheidgen committed
546

547
548
549
550
551
552
553
554
555
556
557
558
    def on_process_complete(self, process_name):
        if process_name == 'process_upload':
            self.check_join()

    def check_join(self):
        total_calcs = self.total_calcs
        processed_calcs = self.processed_calcs

        self.get_logger().debug('check join', processed_calcs=processed_calcs, total_calcs=total_calcs)
        if not self.process_running and processed_calcs >= total_calcs:
            self.get_logger().debug('join')
            self.join()
559
560
561

    def join(self):
        self.cleanup()
Markus Scheidgen's avatar
Markus Scheidgen committed
562
563
564

    @task
    def cleanup(self):
565
566
        search.refresh()

567
568
569
570
571
572
573
574
575
576
        # send email about process finish
        user = self.uploader
        name = '%s %s' % (user.first_name, user.last_name)
        message = '\n'.join([
            'Dear %s,' % name,
            '',
            'your data %suploaded %s has completed processing.' % (
                self.name if self.name else '', self.upload_time.isoformat()),
            'You can review your data on your upload page: %s' % config.services.upload_url
        ])
577
578
579
580
581
582
583
        try:
            infrastructure.send_mail(
                name=name, email=user.email, message=message, subject='Processing completed')
        except Exception as e:
            # probably due to email configuration problems
            # don't fail or present this error to clients
            self.logger.error('could not send after processing email', exc_info=e)
Markus Scheidgen's avatar
Markus Scheidgen committed
584
585

    @property
586
    def processed_calcs(self):
587
        return Calc.objects(upload_id=self.upload_id, tasks_status__in=[SUCCESS, FAILURE]).count()
588
589
590
591
592
593
594

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

    @property
    def failed_calcs(self):
595
        return Calc.objects(upload_id=self.upload_id, tasks_status=FAILURE).count()
596

597
598
    @property
    def pending_calcs(self):
599
        return Calc.objects(upload_id=self.upload_id, tasks_status=PENDING).count()
600

601
602
    def all_calcs(self, start, end, order_by='mainfile'):
        return Calc.objects(upload_id=self.upload_id)[start:end].order_by(order_by)
603
604
605

    @property
    def calcs(self):
Markus Scheidgen's avatar
Markus Scheidgen committed
606
        return Calc.objects(upload_id=self.upload_id, tasks_status=SUCCESS)
Markus Scheidgen's avatar
Markus Scheidgen committed
607
608

    def to_upload_with_metadata(self) -> UploadWithMetadata:
609
610
611
612
613
614
615
616
617
        # TODO remove the very verbose metadata after debugging and optimizing

        logger = self.get_logger(step='publish')

        # prepare user metadata per upload and per calc
        logger.info('prepare user metadata per upload and per calc')
        calc_metadatas: Dict[str, Any] = dict()
        upload_metadata: Dict[str, Any] = dict()

Markus Scheidgen's avatar
Markus Scheidgen committed
618
        if self.metadata is not None:
619
620
621
            upload_metadata.update(self.metadata)
            if 'calculations' in upload_metadata:
                del(upload_metadata['calculations'])
Markus Scheidgen's avatar
Markus Scheidgen committed
622

623
624
            for calc in self.metadata.get('calculations', []):
                calc_metadatas[calc['mainfile']] = calc
Markus Scheidgen's avatar
Markus Scheidgen committed
625

626
        user_upload_time = upload_metadata.get('_upload_time', None)
Markus Scheidgen's avatar
Markus Scheidgen committed
627
628
629
630
631
        result = UploadWithMetadata(
            upload_id=self.upload_id,
            uploader=utils.POPO(id=int(self.user_id)),
            upload_time=self.upload_time if user_upload_time is None else user_upload_time)

632
633
634
635
636
637
638
639
640
        logger.info('read calc data from files and apply user metadata')
        upload_files = UploadFiles.get(self.upload_id, is_authorized=lambda: True)
        if self.upload_files is None:
            raise KeyError

        def apply_metadata(calc):
            calc_data = upload_files.metadata.get(calc.calc_id)
            calc_with_metadata = CalcWithMetadata(**calc_data)

641
642
            calc_metadata = dict(upload_metadata)
            calc_metadata.update(calc_metadatas.get(calc.mainfile, {}))
643
644
            calc_with_metadata.apply_user_metadata(calc_metadata)

Markus Scheidgen's avatar
Markus Scheidgen committed
645
            logger.debug('prepared calc with metadata', calc_id=calc_with_metadata.calc_id)
646
647
648
649
650
            return calc_with_metadata

        result.calcs = [apply_metadata(calc) for calc in self.calcs]

        logger.info('prepared user metadata')
Markus Scheidgen's avatar
Markus Scheidgen committed
651
652

        return result
653
654

    def __str__(self):
655
        return 'upload %s upload_id%s' % (super().__str__(), self.upload_id)