archive.py 16.3 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
18
19
"""
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.
"""

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

29
30
import nomad_meta_info

31
from nomad.files import UploadFiles, Restricted
32
from nomad import search, config
33
from nomad.archive_library.utils import get_dbs
Alvin Noe Ladines's avatar
Alvin Noe Ladines committed
34
from nomad import search, config
35
from nomad.app import common
36

37
from .auth import authenticate, create_authorization_predicate
Markus Scheidgen's avatar
Markus Scheidgen committed
38
from .api import api
39
from .common import calc_route, streamed_zipfile, search_model, add_pagination_parameters,\
40
    add_scroll_parameters, add_search_parameters, apply_search_parameters
41
42

ns = api.namespace(
43
44
    'archive',
    description='Access archive data and archive processing logs.')
45
46
47
48


@calc_route(ns, '/logs')
class ArchiveCalcLogResource(Resource):
49
    @api.doc('get_archive_logs')
50
    @api.response(404, 'The upload or calculation does not exist')
51
    @api.response(401, 'Not authorized to access the data.')
52
    @api.response(200, 'Archive data send', headers={'Content-Type': 'application/plain'})
53
    @authenticate(signature_token=True)
54
    def get(self, upload_id, calc_id):
55
56
57
        """
        Get calculation processing log.

58
        Calcs are references via *upload_id*, *calc_id* pairs.
59
        """
60
        archive_id = '%s/%s' % (upload_id, calc_id)
61

62
        upload_files = UploadFiles.get(
63
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
64

65
        if upload_files is None:
66
            abort(404, message='Upload %s does not exist.' % upload_id)
67
68
69

        try:
            return send_file(
70
                upload_files.archive_log_file(calc_id, 'rb'),
71
72
                mimetype='text/plain',
                as_attachment=True,
73
                cache_timeout=0,
74
75
                attachment_filename='%s.log' % archive_id)
        except Restricted:
76
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
77
78
        except KeyError:
            abort(404, message='Calculation %s does not exist.' % archive_id)
79
80
81
82


@calc_route(ns)
class ArchiveCalcResource(Resource):
83
    @api.doc('get_archive_calc')
84
    @api.response(404, 'The upload or calculation does not exist')
85
    @api.response(401, 'Not authorized to access the data.')
86
    @api.response(200, 'Archive data send')
87
    @authenticate(signature_token=True)
88
    def get(self, upload_id, calc_id):
89
90
91
        """
        Get calculation data in archive form.

92
        Calcs are references via *upload_id*, *calc_id* pairs.
93
        """
94
        archive_id = '%s/%s' % (upload_id, calc_id)
95

96
        upload_file = UploadFiles.get(
97
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
98

99
        if upload_file is None:
100
            abort(404, message='Archive %s does not exist.' % upload_id)
101
102
103

        try:
            return send_file(
104
                upload_file.archive_file(calc_id, 'rb'),
105
106
                mimetype='application/json',
                as_attachment=True,
107
                cache_timeout=0,
108
109
                attachment_filename='%s.json' % archive_id)
        except Restricted:
110
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
111
        except KeyError:
112
            abort(404, message='Calculation %s does not exist.' % archive_id)
113
114


115
116
117
_archive_download_parser = api.parser()
add_search_parameters(_archive_download_parser)
_archive_download_parser.add_argument(
118
119
120
    name='compress', type=bool, help='Use compression on .zip files, default is not.',
    location='args')

121
122
123

@ns.route('/download')
class ArchiveDownloadResource(Resource):
124
125
    manifest_quantities = ['upload_id', 'calc_id', 'external_id', 'raw_id', 'pid', 'calc_hash']

126
    @api.doc('archive_download')
127
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
128
    @api.expect(_archive_download_parser, validate=True)
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
    @api.response(200, 'File(s) send', headers={'Content-Type': 'application/zip'})
    @authenticate(signature_token=True)
    def get(self):
        """
        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.
        """
        try:
144
            args = _archive_download_parser.parse_args()
145
146
147
148
149
            compress = args.get('compress', False)
        except Exception:
            abort(400, message='bad parameter types')

        search_request = search.SearchRequest()
150
        apply_search_parameters(search_request, args)
151
        search_request.include('calc_id', 'upload_id', 'mainfile')
152

153
154
155
156
        calcs = search_request.execute_scan(
            order_by='upload_id',
            size=config.services.download_scan_size,
            scroll=config.services.download_scan_timeout)
157
158

        def generator():
159
            try:
160
161
162
                manifest = {}
                upload_files = None

163
164
165
166
167
168
169
170
171
172
173
                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:
174
                            common.logger.error('upload files do not exist', upload_id=upload_id)
175
176
177
178
179
180
181
182
183
184
185
186
187
188
                            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
                    }
189

190
191
192
193
194
195
196
197
                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')
198
                    common.logger.error(
199
200
201
202
203
204
                        '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
205

206
            except Exception as e:
207
                common.logger.warning(
208
209
210
                    'unexpected error while streaming raw data from query',
                    exc_info=e,
                    query=urllib.parse.urlencode(request.args, doseq=True))
211

212
213
214
215
        return streamed_zipfile(
            generator(), zipfile_name='nomad_archive.zip', compress=compress)


216
217
218
_archive_query_parser = api.parser()
add_pagination_parameters(_archive_query_parser)
add_scroll_parameters(_archive_query_parser)
219
add_search_parameters(_archive_query_parser)
220
221
222
223
_archive_query_parser.add_argument(
    'db', type=str, help='Database to use, zip or msg', default='zip', location='args')
_archive_query_parser.add_argument(
    'qschema', type=str, help='Serialized archive dict with null values as placeholder for data.')
224

225
_archive_query_model = api.clone('ArchiveCalculations', search_model)
226
227
228
# scroll model should be capitalized to prevent ambiguity with scroll flag
_archive_query_model['Scroll'] = _archive_query_model.pop('scroll')
_archive_query_model['Pagination'] = _archive_query_model.pop('pagination')
229
230


231
232
@ns.route('/query')
class ArchiveQueryResource(Resource):
233
234
235
236
237
    @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')
238
    @api.expect(_archive_query_model)
239
    @api.marshal_with(_archive_query_model, skip_none=True, code=200, description='Search results sent')
240
    @authenticate()
241
242
    def post(self):
        """
243
        Post a query schema and return it filled with archive data.
244

245
246
        See ``/repo`` endpoint for documentation on the search
        parameters.
247

248
249
250
        The actual data are in results and a supplementary python code (curl) to
        execute search is in python (curl).
        """
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
        try:
            data_in = request.get_json()
            scroll = data_in.get('scroll', None)
            scroll_id = data_in.get('scroll_id', None)
            Scroll = data_in.get('Scroll', None)
            if Scroll:
                scroll = Scroll.get('scroll', scroll)
                scroll_id = Scroll.get('scroll_id', scroll_id)
            pagination = data_in.get('Pagination', None)
            page = data_in.get('page', 1)
            per_page = data_in.get('per_page', 10 if not scroll else 1000)
            order = data_in.get('order', -1)
            order_by = data_in.get('order_by', 'upload_id')
            if pagination:
                page = pagination.get('page', page)
                per_page = pagination.get('per_page', per_page)
                order = pagination.get('order', order)
                order_by = pagination.get('order_by', order_by)
            qschema = data_in.get('results', None)
            if qschema is not None:
                qschema = qschema[-1]
        except Exception:
            abort(400, message='bad parameter types')

        try:
            assert page >= 1
            assert per_page > 0
        except AssertionError:
            abort(400, message='invalid pagination')

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

        search_request = search.SearchRequest()
        apply_search_parameters(search_request, data_in)
286
        search_request.include('calc_id', 'upload_id', 'mainfile')
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306

        try:
            if scroll:
                results = search_request.execute_scrolled(scroll_id=scroll_id, size=per_page)
                results['scroll']['scroll'] = True

            else:
                results = search_request.execute_paginated(
                    per_page=per_page, page=page, order=order, order_by=order_by)

        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']
        try:
307
            msgdbs = None
Alvin Noe Ladines's avatar
Alvin Noe Ladines committed
308
            cur_upload_id = None
309
310
311
            for entry in calcs:
                upload_id = entry['upload_id']
                calc_id = entry['calc_id']
Alvin Noe Ladines's avatar
Alvin Noe Ladines committed
312
                if msgdbs is None or cur_upload_id != upload_id:
313
                    msgdbs = get_dbs(upload_id)
Alvin Noe Ladines's avatar
Alvin Noe Ladines committed
314
                    cur_upload_id = upload_id
315

316
317
318
319
                if len(msgdbs) == 0:
                    raise KeyError

                calc_data = None
320
                for msgdb in msgdbs:
321
322
323
324
325
326
327
                    calc_data = msgdb.query({calc_id: qschema})
                    if calc_data:
                        data.append(calc_data)
                        break

                if calc_data is None:
                    raise Restricted
328
329
330
331
332
333
334
335
336
337
338
339

        except Restricted:
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))

        except KeyError:
            abort(404, message='Calculation %s/%s does not exist.' % (upload_id, calc_id))

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

        # for compatibility with archive model
        # TODO should be changed in search
340
341
342
343
        if scroll:
            results['Scroll'] = results.pop('scroll', None)
        if pagination:
            results['Pagination'] = results.pop('pagination', None)
344
345

        return results, 200
346
347


348
349
@ns.route('/metainfo/<string:metainfo_package_name>')
@api.doc(params=dict(metainfo_package_name='The name of the metainfo package.'))
350
351
352
353
class MetainfoResource(Resource):
    @api.doc('get_metainfo')
    @api.response(404, 'The metainfo does not exist')
    @api.response(200, 'Metainfo data send')
354
    def get(self, metainfo_package_name):
355
356
357
358
        """
        Get a metainfo definition file.
        """
        try:
359
            return load_metainfo(metainfo_package_name), 200
360
        except FileNotFoundError:
361
            parser_prefix = metainfo_package_name[:-len('.nomadmetainfo.json')]
362

363
            try:
364
365
366
                return load_metainfo(dict(
                    parser='%sparser' % parser_prefix,
                    path='%s.nomadmetainfo.json' % parser_prefix)), 200
367
368
            except FileNotFoundError:
                abort(404, message='The metainfo %s does not exist.' % metainfo_package_name)
369
370
371
372
373


metainfo_main_path = os.path.dirname(os.path.abspath(nomad_meta_info.__file__))


374
375
376
def load_metainfo(
        package_name_or_dependency: str, dependency_source: str = None,
        loaded_packages: Dict[str, Any] = None) -> Dict[str, Any]:
377
378
379
380
381
    """
    Loads the given metainfo package and all its dependencies. Returns a dict with
    all loaded package_names and respective packages.

    Arguments:
382
383
        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.
384
385
386
387
388
389
        loaded_packages: Give a dict and the function will added freshly loaded packages
            to it and return it.
    """
    if loaded_packages is None:
        loaded_packages = {}

390
391
    if isinstance(package_name_or_dependency, str):
        package_name = package_name_or_dependency
392
        metainfo_path = os.path.join(metainfo_main_path, package_name)
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
    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)
421
422
423
424
425
426
427
428
429
430
431
432

    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', []):
433
        load_metainfo(dependency, dependency_source=metainfo_path, loaded_packages=loaded_packages)
434
435

    return loaded_packages