archive.py 15.6 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, fields
25
import json
26
import importlib
27
28
import urllib.parse
from elasticsearch.helpers import ScanError
29

30
31
import nomad_meta_info

32
from nomad.files import UploadFiles, Restricted
33
from nomad import utils, search, config
34

35
from .auth import authenticate, create_authorization_predicate
Markus Scheidgen's avatar
Markus Scheidgen committed
36
from .api import api
37
38
39
from .common import calc_route, streamed_zipfile, search_model, add_pagination_parameters,\
    add_scroll_parameters, add_search_parameters, apply_search_parameters,\
    query_api_python, query_api_curl
40
41

ns = api.namespace(
42
43
    'archive',
    description='Access archive data and archive processing logs.')
44
45
46
47


@calc_route(ns, '/logs')
class ArchiveCalcLogResource(Resource):
48
    @api.doc('get_archive_logs')
49
    @api.response(404, 'The upload or calculation does not exist')
50
    @api.response(401, 'Not authorized to access the data.')
51
    @api.response(200, 'Archive data send', headers={'Content-Type': 'application/plain'})
52
    @authenticate(signature_token=True)
53
    def get(self, upload_id, calc_id):
54
55
56
        """
        Get calculation processing log.

57
        Calcs are references via *upload_id*, *calc_id* pairs.
58
        """
59
        archive_id = '%s/%s' % (upload_id, calc_id)
60

61
        upload_files = UploadFiles.get(
62
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
63

64
        if upload_files is None:
65
            abort(404, message='Upload %s does not exist.' % upload_id)
66
67
68

        try:
            return send_file(
69
                upload_files.archive_log_file(calc_id, 'rb'),
70
71
                mimetype='text/plain',
                as_attachment=True,
72
                cache_timeout=0,
73
74
                attachment_filename='%s.log' % archive_id)
        except Restricted:
75
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
76
77
        except KeyError:
            abort(404, message='Calculation %s does not exist.' % archive_id)
78
79
80
81


@calc_route(ns)
class ArchiveCalcResource(Resource):
82
    @api.doc('get_archive_calc')
83
    @api.response(404, 'The upload or calculation does not exist')
84
    @api.response(401, 'Not authorized to access the data.')
85
    @api.response(200, 'Archive data send')
86
    @authenticate(signature_token=True)
87
    def get(self, upload_id, calc_id):
88
89
90
        """
        Get calculation data in archive form.

91
        Calcs are references via *upload_id*, *calc_id* pairs.
92
        """
93
        archive_id = '%s/%s' % (upload_id, calc_id)
94

95
        upload_file = UploadFiles.get(
96
            upload_id, is_authorized=create_authorization_predicate(upload_id, calc_id))
97

98
        if upload_file is None:
99
            abort(404, message='Archive %s does not exist.' % upload_id)
100
101
102

        try:
            return send_file(
103
                upload_file.archive_file(calc_id, 'rb'),
104
105
                mimetype='application/json',
                as_attachment=True,
106
                cache_timeout=0,
107
108
                attachment_filename='%s.json' % archive_id)
        except Restricted:
109
            abort(401, message='Not authorized to access %s/%s.' % (upload_id, calc_id))
110
        except KeyError:
111
            abort(404, message='Calculation %s does not exist.' % archive_id)
112
113


114
115
116
_archive_download_parser = api.parser()
add_search_parameters(_archive_download_parser)
_archive_download_parser.add_argument(
117
118
119
    name='compress', type=bool, help='Use compression on .zip files, default is not.',
    location='args')

120
121
122

@ns.route('/download')
class ArchiveDownloadResource(Resource):
123
124
    manifest_quantities = ['upload_id', 'calc_id', 'external_id', 'raw_id', 'pid', 'calc_hash']

125
    @api.doc('archive_download')
126
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
127
    @api.expect(_archive_download_parser, validate=True)
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
    @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:
143
            args = _archive_download_parser.parse_args()
144
145
146
147
148
            compress = args.get('compress', False)
        except Exception:
            abort(400, message='bad parameter types')

        search_request = search.SearchRequest()
149
        apply_search_parameters(search_request, args)
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
            manifest = {}
Markus Scheidgen's avatar
Markus Scheidgen committed
158
            upload_files = None
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
            try:
                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:
                            utils.get_logger(__name__).error('upload files do not exist', upload_id=upload_id)
                            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
                    }
            except ScanError:
                utils.get_logger(__name__).warning(
                    'scan error while streaming archive data from query',
                    query=urllib.parse.urlencode(request.args, doseq=True))
190

Markus Scheidgen's avatar
Markus Scheidgen committed
191
192
193
            if upload_files is not None:
                upload_files.close_zipfile_cache()

194
195
196
197
198
199
200
201
202
203
204
205
            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')
                utils.get_logger(__name__).error(
                    'could not create raw query manifest', exc_info=e)

            yield (
                'manifest.json', 'manifest',
                lambda *args: BytesIO(manifest_contents),
                lambda *args: len(manifest_contents))
206

207
208
209
210
        return streamed_zipfile(
            generator(), zipfile_name='nomad_archive.zip', compress=compress)


211
212
213
_archive_query_parser = api.parser()
add_pagination_parameters(_archive_query_parser)
add_scroll_parameters(_archive_query_parser)
214
add_search_parameters(_archive_query_parser)
215
216
217
218
219
220
221
222
223
224
225
226
227

_archive_query_model_fields = {
    'results': fields.List(fields.Raw, description=(
        'A list of search results. Each result is a dict with quantities names as key and '
        'values as values')),
    'python': fields.String(description=(
        'A string of python code snippet which can be executed to reproduce the api result.')),
    'curl': fields.String(description=(
        'A string of curl command which can be executed to reproduce the api result.')),
}
_archive_query_model = api.inherit('ArchiveCalculations', search_model, _archive_query_model_fields)


228
229
@ns.route('/query')
class ArchiveQueryResource(Resource):
230
    @api.doc('archive_query')
231
    @api.response(400, 'Invalid requests, e.g. wrong owner type or bad search parameters')
232
233
234
235
236
    @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')
    @api.expect(_archive_query_parser, validate=True)
    @api.marshal_with(_archive_query_model, skip_none=True, code=200, description='Search results sent')
237
    @authenticate()
238
239
240
241
242
243
244
    def get(self):
        """
        Get archive data in json format from all query results.

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

245
246
        The actual data are in archive_data and a supplementary python code (curl) to
        execute search is in python (curl).
247
248
        """
        try:
249
250
251
252
            args = {
                key: value for key, value in _archive_query_parser.parse_args().items()
                if value is not None}

253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
            scroll = args.get('scroll', False)
            scroll_id = args.get('scroll_id', None)
            page = args.get('page', 1)
            per_page = args.get('per_page', 10 if not scroll else 1000)
            order = args.get('order', -1)
            order_by = 'upload_id'
        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()
272
        apply_search_parameters(search_request, args)
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288

        try:
            if scroll:
                results = search_request.execute_scrolled(scroll_id=scroll_id, size=per_page)

            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))

289
        # build python code and curl snippet
290
291
        results['python'] = query_api_python('archive', 'query', query_string=request.args)
        results['curl'] = query_api_curl('archive', 'query', query_string=request.args)
292

293
        data = []
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
        calcs = results['results']
        try:
            upload_files = None
            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:
                        raise KeyError

                    upload_files.open_zipfile_cache()

                fo = upload_files.archive_file(calc_id, 'rb')
313
                data.append(json.loads(fo.read()))
314
315
316
317
318
319
320
321
322
323

            if upload_files is not None:
                upload_files.close_zipfile_cache()

        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))

324
        results['results'] = data
325
326

        return results, 200
327
328


329
330
@ns.route('/metainfo/<string:metainfo_package_name>')
@api.doc(params=dict(metainfo_package_name='The name of the metainfo package.'))
331
332
333
334
class MetainfoResource(Resource):
    @api.doc('get_metainfo')
    @api.response(404, 'The metainfo does not exist')
    @api.response(200, 'Metainfo data send')
335
    def get(self, metainfo_package_name):
336
337
338
339
        """
        Get a metainfo definition file.
        """
        try:
340
            return load_metainfo(metainfo_package_name), 200
341
        except FileNotFoundError:
342
            parser_prefix = metainfo_package_name[:-len('.nomadmetainfo.json')]
343

344
            try:
345
346
347
                return load_metainfo(dict(
                    parser='%sparser' % parser_prefix,
                    path='%s.nomadmetainfo.json' % parser_prefix)), 200
348
349
            except FileNotFoundError:
                abort(404, message='The metainfo %s does not exist.' % metainfo_package_name)
350
351
352
353
354


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


355
356
357
def load_metainfo(
        package_name_or_dependency: str, dependency_source: str = None,
        loaded_packages: Dict[str, Any] = None) -> Dict[str, Any]:
358
359
360
361
362
    """
    Loads the given metainfo package and all its dependencies. Returns a dict with
    all loaded package_names and respective packages.

    Arguments:
363
364
        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.
365
366
367
368
369
370
        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 = {}

371
372
    if isinstance(package_name_or_dependency, str):
        package_name = package_name_or_dependency
373
        metainfo_path = os.path.join(metainfo_main_path, package_name)
374
375
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
    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)
402
403
404
405
406
407
408
409
410
411
412
413

    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', []):
414
        load_metainfo(dependency, dependency_source=metainfo_path, loaded_packages=loaded_packages)
415
416

    return loaded_packages