archive.py 15.4 KB
Newer Older
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
'''
16
17
The archive API of the nomad@FAIRDI APIs. This API is about serving processed
(parsed and normalized) calculation data in nomad's *meta-info* format.
18
'''
19

20
from typing import Dict, Any, List
21
from io import BytesIO
22
import os.path
23
24
from flask import send_file, request, g
from flask_restplus import abort, Resource, fields
25
import json
26
import importlib
27
import urllib.parse
28

29
import metainfo
30

31
from nomad.files import UploadFiles, Restricted
32
from nomad import search, config, archive
33
from nomad.app import common
34

35
from .auth import authenticate, create_authorization_predicate
Markus Scheidgen's avatar
Markus Scheidgen committed
36
from .api import api
37
38
from .common import calc_route, streamed_zipfile, search_model, add_search_parameters, apply_search_parameters, query_model

39
40

ns = api.namespace(
41
42
    'archive',
    description='Access archive data and archive processing logs.')
43
44
45
46


@calc_route(ns, '/logs')
class ArchiveCalcLogResource(Resource):
47
    @api.doc('get_archive_logs')
48
    @api.response(404, 'The upload or calculation does not exist')
49
    @api.response(401, 'Not authorized to access the data.')
50
    @api.response(200, 'Archive data send', headers={'Content-Type': 'application/plain'})
51
    @authenticate(signature_token=True)
52
    def get(self, upload_id, calc_id):
53
        '''
54
55
        Get calculation processing log.

56
        Calcs are references via *upload_id*, *calc_id* pairs.
57
        '''
58
        archive_id = '%s/%s' % (upload_id, calc_id)
59

60
        upload_files = UploadFiles.get(
61
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
62

63
        if upload_files is None:
64
            abort(404, message='Upload %s does not exist.' % upload_id)
65
66
67

        try:
            return send_file(
68
                upload_files.archive_log_file(calc_id, 'rb'),
69
70
                mimetype='text/plain',
                as_attachment=True,
71
                cache_timeout=0,
72
73
                attachment_filename='%s.log' % archive_id)
        except Restricted:
74
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
75
76
        except KeyError:
            abort(404, message='Calculation %s does not exist.' % archive_id)
77
78
79
80


@calc_route(ns)
class ArchiveCalcResource(Resource):
81
    @api.doc('get_archive_calc')
82
    @api.response(404, 'The upload or calculation does not exist')
83
    @api.response(401, 'Not authorized to access the data.')
84
    @api.response(200, 'Archive data send')
85
    @authenticate(signature_token=True)
86
    def get(self, upload_id, calc_id):
87
        '''
88
89
        Get calculation data in archive form.

90
        Calcs are references via *upload_id*, *calc_id* pairs.
91
        '''
92
        archive_id = '%s/%s' % (upload_id, calc_id)
93

94
        upload_file = UploadFiles.get(
95
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
96

97
        if upload_file is None:
98
            abort(404, message='Archive %s does not exist.' % upload_id)
99
100
101

        try:
            return send_file(
102
                upload_file.archive_file(calc_id, 'rb'),
103
104
                mimetype='application/json',
                as_attachment=True,
105
                cache_timeout=0,
106
107
                attachment_filename='%s.json' % archive_id)
        except Restricted:
108
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
109
        except KeyError:
110
            abort(404, message='Calculation %s does not exist.' % archive_id)
111
112


113
114
115
_archive_download_parser = api.parser()
add_search_parameters(_archive_download_parser)
_archive_download_parser.add_argument(
116
117
118
    name='compress', type=bool, help='Use compression on .zip files, default is not.',
    location='args')

119
120
121

@ns.route('/download')
class ArchiveDownloadResource(Resource):
122
123
    manifest_quantities = ['upload_id', 'calc_id', 'external_id', 'raw_id', 'pid', 'calc_hash']

124
    @api.doc('archive_download')
125
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
126
    @api.expect(_archive_download_parser, validate=True)
127
128
129
    @api.response(200, 'File(s) send', headers={'Content-Type': 'application/zip'})
    @authenticate(signature_token=True)
    def get(self):
130
        '''
131
132
133
134
135
136
137
138
139
        Get calculation data in archive form from all query results.

        See ``/repo`` endpoint for documentation on the search
        parameters.

        Zip files are streamed; instead of 401 errors, the zip file will just not contain
        any files that the user is not authorized to access.

        The zip file will contain a ``manifest.json`` with the repository meta data.
140
        '''
141
        try:
142
            args = _archive_download_parser.parse_args()
143
144
145
146
147
            compress = args.get('compress', False)
        except Exception:
            abort(400, message='bad parameter types')

        search_request = search.SearchRequest()
148
        apply_search_parameters(search_request, args)
149
        search_request.include('calc_id', 'upload_id', 'mainfile')
150

151
152
153
154
        calcs = search_request.execute_scan(
            order_by='upload_id',
            size=config.services.download_scan_size,
            scroll=config.services.download_scan_timeout)
155
156

        def generator():
157
            try:
158
159
160
                manifest = {}
                upload_files = None

161
162
163
164
165
166
167
168
169
170
171
                for entry in calcs:
                    upload_id = entry['upload_id']
                    calc_id = entry['calc_id']
                    if upload_files is None or upload_files.upload_id != upload_id:
                        if upload_files is not None:
                            upload_files.close_zipfile_cache()

                        upload_files = UploadFiles.get(
                            upload_id, create_authorization_predicate(upload_id))

                        if upload_files is None:
172
                            common.logger.error('upload files do not exist', upload_id=upload_id)
173
174
175
176
177
178
179
180
181
182
183
184
185
186
                            continue

                        upload_files.open_zipfile_cache()

                    yield (
                        '%s.%s' % (calc_id, upload_files._archive_ext), calc_id,
                        lambda calc_id: upload_files.archive_file(calc_id, 'rb'),
                        lambda calc_id: upload_files.archive_file_size(calc_id))

                    manifest[calc_id] = {
                        key: entry[key]
                        for key in ArchiveDownloadResource.manifest_quantities
                        if entry.get(key) is not None
                    }
187

188
189
190
191
192
193
194
195
                if upload_files is not None:
                    upload_files.close_zipfile_cache()

                try:
                    manifest_contents = json.dumps(manifest).encode('utf-8')
                except Exception as e:
                    manifest_contents = json.dumps(
                        dict(error='Could not create the manifest: %s' % (e))).encode('utf-8')
196
                    common.logger.error(
197
198
199
200
201
202
                        'could not create raw query manifest', exc_info=e)

                yield (
                    'manifest.json', 'manifest',
                    lambda *args: BytesIO(manifest_contents),
                    lambda *args: len(manifest_contents))
Markus Scheidgen's avatar
Markus Scheidgen committed
203

204
            except Exception as e:
205
                common.logger.warning(
206
207
208
                    'unexpected error while streaming raw data from query',
                    exc_info=e,
                    query=urllib.parse.urlencode(request.args, doseq=True))
209

210
211
212
213
        return streamed_zipfile(
            generator(), zipfile_name='nomad_archive.zip', compress=compress)


214
_archive_query_model = api.inherit('ArchiveSearch', search_model, {
215
216
217
    'query': fields.Nested(query_model, description='The query used to find the requested entries.'),
    'query_schema': fields.Raw(description='The query schema that defines what archive data to retrive.')
})
218
219


220
221
@ns.route('/query')
class ArchiveQueryResource(Resource):
222
223
224
225
226
    @api.doc('post_archive_query')
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
    @api.response(401, 'Not authorized to access the data.')
    @api.response(404, 'The upload or calculation does not exist')
    @api.response(200, 'Archive data send')
227
    @api.expect(_archive_query_model)
228
    @api.marshal_with(_archive_query_model, skip_none=True, code=200, description='Search results sent')
229
    @authenticate()
230
    def post(self):
231
        '''
232
        Post a query schema and return it filled with archive data.
233

234
235
        See ``/repo`` endpoint for documentation on the search
        parameters.
236

237
238
        The actual data are in results and a supplementary python code (curl) to
        execute search is in python (curl).
239
        '''
240
241
242
        try:
            data_in = request.get_json()
            scroll = data_in.get('scroll', None)
243
244
245
246
247
248
249
250
251
252
253
            if scroll:
                scroll_id = scroll.get('scroll_id')
                scroll = True

            pagination = data_in.get('pagination', {})
            page = pagination.get('page', 1)
            per_page = pagination.get('per_page', 10 if not scroll else 1000)

            query = data_in.get('query', {})
            query_schema = data_in.get('query_schema', '*')

254
255
256
        except Exception:
            abort(400, message='bad parameter types')

257
        if not (page >= 1 and per_page > 0):
258
259
260
            abort(400, message='invalid pagination')

        search_request = search.SearchRequest()
261
262
263
264
265
266
        if g.user is not None:
            search_request.owner('all', user_id=g.user.user_id)
        else:
            search_request.owner('all')

        apply_search_parameters(search_request, query)
267
        search_request.include('calc_id', 'upload_id', 'with_embargo', 'parser_name')
268
269
270

        try:
            if scroll:
271
272
                results = search_request.execute_scrolled(
                    scroll_id=scroll_id, size=per_page, order_by='upload_id')
273
274
275
276
                results['scroll']['scroll'] = True

            else:
                results = search_request.execute_paginated(
277
                    per_page=per_page, page=page, order_by='upload_id')
278
279
280
281
282
283
284
285
286
287

        except search.ScrollIdNotFound:
            abort(400, 'The given scroll_id does not exist.')
        except KeyError as e:
            import traceback
            traceback.print_exc()
            abort(400, str(e))

        data = []
        calcs = results['results']
288
        archive_readers: List[archive.ArchiveReader] = []
289
        current_upload_id = None
290
291
292
        for entry in calcs:
            upload_id = entry['upload_id']
            calc_id = entry['calc_id']
293
            if current_upload_id is None or current_upload_id != upload_id:
294
295
296
297
298
                upload_files = UploadFiles.get(upload_id, create_authorization_predicate(upload_id))

                if upload_files is None:
                    return []

299
300
301
302
303
304
305
306
                for archive_reader in archive_readers:
                    if archive_reader is not None:
                        archive_reader.close()

                archive_readers = [
                    archive.ArchiveReader(f) if f is not None else None
                    for f in upload_files.archive_file_msgs()]

307
                current_upload_id = upload_id
308
309

            if entry['with_embargo']:
310
                archive_reader = archive_readers[1]
311
            else:
312
                archive_reader = archive_readers[0]
313

314
            if archive_reader is None:
315
                continue
316

317
318
319
320
321
322
323
            data.append(
                {
                    'calc_id': calc_id,
                    'parser_name': entry['parser_name'],
                    'archive': archive.query_archive(
                        archive_reader, {calc_id: query_schema})[calc_id]
                })
324
325
326
327

        # assign archive data to results
        results['results'] = data

328
        return results, 200
329
330


331
332
@ns.route('/metainfo/<string:metainfo_package_name>')
@api.doc(params=dict(metainfo_package_name='The name of the metainfo package.'))
333
334
335
336
class MetainfoResource(Resource):
    @api.doc('get_metainfo')
    @api.response(404, 'The metainfo does not exist')
    @api.response(200, 'Metainfo data send')
337
    def get(self, metainfo_package_name):
338
        '''
339
        Get a metainfo definition file.
340
        '''
341
        try:
342
            return load_metainfo(metainfo_package_name), 200
343
        except FileNotFoundError:
344
            parser_prefix = metainfo_package_name[:-len('.nomadmetainfo.json')]
345

346
            try:
347
348
349
                return load_metainfo(dict(
                    parser='%sparser' % parser_prefix,
                    path='%s.nomadmetainfo.json' % parser_prefix)), 200
350
351
            except FileNotFoundError:
                abort(404, message='The metainfo %s does not exist.' % metainfo_package_name)
352
353


354
metainfo_main_path = os.path.dirname(os.path.abspath(metainfo.__file__))
355
356


357
358
359
def load_metainfo(
        package_name_or_dependency: str, dependency_source: str = None,
        loaded_packages: Dict[str, Any] = None) -> Dict[str, Any]:
360
    '''
361
362
363
364
    Loads the given metainfo package and all its dependencies. Returns a dict with
    all loaded package_names and respective packages.

    Arguments:
365
366
        package_name_or_dependency: The name of the package, or a nomadmetainfo dependency object.
        dependency_source: The path of the metainfo that uses this function to load a relative dependency.
367
368
        loaded_packages: Give a dict and the function will added freshly loaded packages
            to it and return it.
369
    '''
370
371
372
    if loaded_packages is None:
        loaded_packages = {}

373
374
    if isinstance(package_name_or_dependency, str):
        package_name = package_name_or_dependency
375
        metainfo_path = os.path.join(metainfo_main_path, package_name)
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
    else:
        dependency = package_name_or_dependency
        if 'relativePath' in dependency:
            if dependency_source is None:
                raise Exception(
                    'Can only load relative dependency from within another metainfo package')

            metainfo_path = os.path.join(
                os.path.dirname(dependency_source), dependency['relativePath'])

        elif 'metainfoPath' in dependency:
            metainfo_path = os.path.join(metainfo_main_path, dependency['metainfoPath'])

        elif 'parser' in dependency:
            parser = dependency['parser']
            path = dependency['path']
            try:
                parser_module = importlib.import_module(parser).__file__
            except Exception:
                raise Exception('Parser not installed %s for metainfo path %s' % (parser, metainfo_path))

            parser_directory = os.path.dirname(parser_module)
            metainfo_path = os.path.join(parser_directory, path)

        else:
            raise Exception('Invalid dependency type in metainfo package %s' % metainfo_path)

        package_name = os.path.basename(metainfo_path)
404
405
406
407
408
409
410
411
412
413
414
415

    package_name = os.path.basename(package_name)

    if package_name in loaded_packages:
        return loaded_packages

    with open(metainfo_path, 'rt') as f:
        metainfo_json = json.load(f)

    loaded_packages[package_name] = metainfo_json

    for dependency in metainfo_json.get('dependencies', []):
416
        load_metainfo(dependency, dependency_source=metainfo_path, loaded_packages=loaded_packages)
417
418

    return loaded_packages