admin.py 20.3 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
from typing import Callable, List
16
import click
17
import datetime
18
from elasticsearch_dsl import Q
19
import elasticsearch.helpers
20
import sys
21
22
23
24
25
import io
import re
import uuid
import json
from io import StringIO
26
import threading
27
28
29
30
31
32
33

import numpy as np
import requests
from ase import Atoms
import ase.io
from bs4 import BeautifulSoup
from matid import SymmetryAnalyzer
34

35
from nomad import processing as proc, search, datamodel, infrastructure, utils, config
Markus Scheidgen's avatar
Markus Scheidgen committed
36
from nomad.normalizing.aflow_prototypes import get_normalized_wyckoff
37
from nomad.cli.cli import cli
38
from nomad import config
Markus Scheidgen's avatar
Markus Scheidgen committed
39
from nomad.normalizing.springer import update_springer_data
40
41


42
def __run_processing(
43
44
        uploads, parallel: int, process: Callable[[proc.Upload], None], label: str,
        reprocess_running: bool = False):
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
    if isinstance(uploads, (tuple, list)):
        uploads_count = len(uploads)

    else:
        uploads_count = uploads.count()
        uploads = list(uploads)  # copy the whole mongo query set to avoid cursor timeouts

    cv = threading.Condition()
    threads: List[threading.Thread] = []

    state = dict(
        completed_count=0,
        skipped_count=0,
        available_threads_count=parallel)

    logger = utils.get_logger(__name__)

    print('%d uploads selected, %s ...' % (uploads_count, label))

    def process_upload(upload: proc.Upload):
        logger.info('%s started' % label, upload_id=upload.upload_id)

        completed = False
68
        if upload.process_running and not reprocess_running:
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
            logger.warn(
                'cannot trigger %s, since the upload is already/still processing' % label,
                current_process=upload.current_process,
                current_task=upload.current_task, upload_id=upload.upload_id)

        else:
            upload.reset()
            process(upload)
            upload.block_until_complete(interval=.5)

            if upload.tasks_status == proc.FAILURE:
                logger.info('%s with failure' % label, upload_id=upload.upload_id)

            completed = True

            logger.info('%s complete' % label, upload_id=upload.upload_id)

        with cv:
            state['completed_count'] += 1 if completed else 0
            state['skipped_count'] += 1 if not completed else 0
            state['available_threads_count'] += 1

            print(
                '   %s %s and skipped %s of %s uploads' %
                (label, state['completed_count'], state['skipped_count'], uploads_count))

            cv.notify()

    for upload in uploads:
        with cv:
            cv.wait_for(lambda: state['available_threads_count'] > 0)
            state['available_threads_count'] -= 1
            thread = threading.Thread(target=lambda: process_upload(upload))
            threads.append(thread)
            thread.start()

    for thread in threads:
        thread.join()


109
110
111
112
113
@cli.group(help='''The nomad admin commands to do nasty stuff directly on the databases.
                     Remember: With great power comes great responsibility!''')
@click.pass_context
def admin(ctx):
    pass
114
115


116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
@admin.command(help='Reset/remove all databases.')
@click.option('--remove', is_flag=True, help='Do not just reset all dbs, but also remove them.')
@click.option('--i-am-really-sure', is_flag=True, help='Must be set for the command to to anything.')
def reset(remove, i_am_really_sure):
    if not i_am_really_sure:
        print('You do not seem to be really sure about what you are doing.')
        sys.exit(1)

    infrastructure.setup_logging()
    infrastructure.setup_mongo()
    infrastructure.setup_elastic()

    infrastructure.reset(remove)


131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
@admin.command(help='Check and lift embargo of data with expired embargo period.')
@click.option('--dry', is_flag=True, help='Do not lift the embargo, just show what needs to be done.')
@click.option('--parallel', default=1, type=int, help='Use the given amount of parallel processes. Default is 1.')
def lift_embargo(dry, parallel):
    infrastructure.setup_logging()
    infrastructure.setup_mongo()
    infrastructure.setup_elastic()

    request = search.SearchRequest()
    request.q = Q('term', with_embargo=True) & Q('term', published=True)
    request.quantity('upload_id', 1000)
    result = request.execute()

    uploads_to_repack = []
    for upload_id in result['quantities']['upload_id']['values']:
        upload = proc.Upload.get(upload_id)
        embargo_length = upload.embargo_length
        if embargo_length is None:
            embargo_length = 36
            upload.embargo_length = 36

        if upload.upload_time + datetime.timedelta(days=int(embargo_length * 365 / 12)) < datetime.datetime.now():
            print('need to lift the embargo of %s (upload_time=%s, embargo=%d)' % (
                upload.upload_id, upload.upload_time, embargo_length))

            if not dry:
                proc.Calc._get_collection().update_many(
                    {'upload_id': upload_id},
                    {'$set': {'metadata.with_embargo': False}})
                uploads_to_repack.append(upload)
                upload.save()

163
164
                with upload.entries_metadata() as entries:
                    search.index_all(entries)
165
166
167
168
169

    if not dry:
        __run_processing(uploads_to_repack, parallel, lambda upload: upload.re_pack(), 're-packing')


170
@admin.command(help='(Re-)index all calcs.')
171
@click.option('--threads', type=int, default=1, help='Number of threads to use.')
172
173
@click.option('--dry', is_flag=True, help='Do not index, just compute entries.')
def index(threads, dry):
174
175
176
177
178
179
180
    infrastructure.setup_logging()
    infrastructure.setup_mongo()
    infrastructure.setup_elastic()

    all_calcs = proc.Calc.objects().count()
    print('indexing %d ...' % all_calcs)

181
    def elastic_updates():
182
183
184
        with utils.ETA(all_calcs, '   index %10d or %10d calcs, ETA %s') as eta:
            for calc in proc.Calc.objects():
                eta.add()
185
186
                entry_metadata = datamodel.EntryMetadata.m_from_dict(calc.metadata)
                entry = entry_metadata.a_elastic.create_index_entry().to_dict(include_meta=True)
187
188
                entry['_op_type'] = 'index'
                yield entry
Markus Scheidgen's avatar
Markus Scheidgen committed
189

190
191
192
    if dry:
        for _ in elastic_updates():
            pass
193
    if threads > 1:
194
        print('  use %d threads' % threads)
195
        for _ in elasticsearch.helpers.parallel_bulk(
196
                infrastructure.elastic_client, elastic_updates(), chunk_size=500,
197
198
                thread_count=threads):
            pass
199
    else:
200
201
202
        elasticsearch.helpers.bulk(
            infrastructure.elastic_client, elastic_updates())
    search.refresh()
203

204
    print('')
205
    print('indexing completed')
206
207


208
@admin.group(help='Generate scripts and commands for nomad operation.')
209
210
211
212
def ops():
    pass


213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
@ops.command(help=('Dump the mongo (calculation metadata) db.'))
@click.option('--restore', is_flag=True, help='Do not dump, but restore.')
def dump(restore: bool):
    date_str = datetime.datetime.utcnow().strftime('%Y_%m_%d')
    print('mongodump --host {} --port {} --db {} -o /backup/fairdi/mongo/{}'.format(
        config.mongo.host, config.mongo.port, config.mongo.db_name, date_str))


@ops.command(help=('Restore the mongo (calculation metadata) db.'))
@click.argument('PATH_TO_DUMP', type=str, nargs=1)
def restore(path_to_dump):
    print('mongorestore --host {} --port {} --db {} {}'.format(
        config.mongo.host, config.mongo.port, config.mongo.db_name, path_to_dump))


228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
@ops.command(help=('Generate an nginx.conf to serve the GUI and proxy pass to API container.'))
@click.option('--prefix', type=str, default='/example_nomad', help='Url path prefix. Default is /example_nomd, can be empty str.')
def nginx_conf(prefix):
    prefix = prefix.rstrip('/')
    prefix = '/%s' % prefix.lstrip('/')

    print('''\
server {{
    listen        80;
    server_name   www.example.com;

    location /{0} {{
        return 301 /example-nomad/gui;
    }}

    location {1}/gui {{
        root      /app/;
        rewrite ^{1}/gui/(.*)$ /nomad/$1 break;
        try_files $uri {1}/gui/index.html;
    }}

    location {1}/gui/service-worker.js {{
        add_header Last-Modified $date_gmt;
        add_header Cache-Control 'no-store, no-cache, must-revalidate, proxy-revalidate, max-age=0';
        if_modified_since off;
        expires off;
        etag off;
        root      /app/;
        rewrite ^{1}/gui/service-worker.js /nomad/service-worker.js break;
    }}

259
260
261
262
263
264
265
266
267
268
    location {1}/gui/meta.json {{
        add_header Last-Modified $date_gmt;
        add_header Cache-Control 'no-store, no-cache, must-revalidate, proxy-revalidate, max-age=0';
        if_modified_since off;
        expires off;
        etag off;
        root      /app/;
        rewrite ^{1}/gui/meta.json /nomad/meta.json break;
    }}

269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
    location {1}/api {{
        proxy_set_header Host $host;
        proxy_pass_request_headers on;
        proxy_pass http://api:8000;
    }}

    location {1}/api/uploads {{
        client_max_body_size 35g;
        proxy_request_buffering off;
        proxy_set_header Host $host;
        proxy_pass_request_headers on;
        proxy_pass http://api:8000;
    }}

    location {1}/api/raw {{
        proxy_buffering off;
        proxy_set_header Host $host;
        proxy_pass_request_headers on;
        proxy_pass http://api:8000;
    }}
}}
    '''.format(prefix.lstrip('/'), prefix))


@ops.command(help=('Generate a proxy pass config for apache2 reverse proxy servers.'))
294
@click.option('--prefix', type=str, default='app', help='The path prefix under which everything is proxy passed.')
295
296
@click.option('--host', type=str, default='130.183.207.104', help='The host to proxy to.')
@click.option('--port', type=str, default='30001', help='The port to proxy to.')
297
298
def apache_conf(prefix, host, port):
    print('''\
299
300
301
302
ProxyPass "/{0}" "http://{1}:{2}/{0}"
ProxyPassReverse "/{0}" "http://{1}:{2}/{0}"
<Proxy http://{1}:{2}/{0}>
    ProxyPreserveHost On
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
    <IfModule !mod_access_compat.c>
         Require all granted
     </IfModule>
     <IfModule mod_access_compat.c>
         Order allow,deny
         Allow from all
     </IfModule>
</Proxy>

RequestHeader set "X-Forwarded-Proto" expr=%{{REQUEST_SCHEME}}
RequestHeader set "X-Forwarded-SSL" expr=%{{HTTPS}}

ProxyPass /fairdi/keycloak http://{1}:8002/fairdi/keycloak
ProxyPassReverse /fairdi/keycloak http://{1}:8002/fairdi/keycloak
<Proxy http://{1}:8002/app>
     ProxyPreserveHost On
     <IfModule !mod_access_compat.c>
         Require all granted
     </IfModule>
     <IfModule mod_access_compat.c>
         Order allow,deny
         Allow from all
     </IfModule>
</Proxy>

RewriteEngine on
RewriteCond %{QUERY_STRING} ^pid=([^&]+)$
RewriteRule ^/NomadRepository-1.1/views/calculation.zul$ /{0}/gui/entry/pid/%1? [R=301]

AllowEncodedSlashes On
333
'''.format(prefix, host, port))  # type: ignore
334
335
336


def write_prototype_data_file(aflow_prototypes: dict, filepath) -> None:
337
    '''Writes the prototype data file in a compressed format to a python
338
339
340
341
    module.

    Args:
        aflow_prototypes
342
    '''
343
344
345
346
347
    class NoIndent(object):
        def __init__(self, value):
            self.value = value

    class NoIndentEncoder(json.JSONEncoder):
348
        '''A custom JSON encoder that can pretty-print objects wrapped in the
349
        NoIndent class.
350
        '''
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
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
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
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
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
        def __init__(self, *args, **kwargs):
            super(NoIndentEncoder, self).__init__(*args, **kwargs)
            self.kwargs = dict(kwargs)
            del self.kwargs['indent']
            self._replacement_map = {}

        def default(self, o):  # pylint: disable=E0202
            if isinstance(o, NoIndent):
                key = uuid.uuid4().hex
                self._replacement_map[key] = json.dumps(o.value, **self.kwargs)
                return "@@%s@@" % (key,)
            else:
                return super(NoIndentEncoder, self).default(o)

        def encode(self, o):
            result = super(NoIndentEncoder, self).encode(o)
            for k, v in self._replacement_map.items():
                result = result.replace('"@@%s@@"' % (k,), v)
            return result

    prototype_dict = aflow_prototypes["prototypes_by_spacegroup"]
    for prototypes in prototype_dict.values():
        for prototype in prototypes:
            # Save the information back in a prettified form
            prototype["atom_positions"] = NoIndent(prototype["atom_positions"])
            prototype["atom_labels"] = NoIndent(prototype["atom_labels"])
            prototype["lattice_vectors"] = NoIndent(prototype["lattice_vectors"])
            try:
                prototype["normalized_wyckoff_matid"] = NoIndent(prototype["normalized_wyckoff_matid"])
            except KeyError:
                pass

    # Save the updated data
    with io.open(filepath, "w", encoding="utf8") as f:
        json_dump = json.dumps(aflow_prototypes, ensure_ascii=False, indent=4, sort_keys=True, cls=NoIndentEncoder)
        json_dump = re.sub(r"\"(-?\d+(?:[\.,]\d+)?)\"", r'\1', json_dump)  # Removes quotes around numbers
        f.write("# -*- coding: utf-8 -*-\naflow_prototypes = {}\n".format(json_dump))


@ops.command(help='Updates the AFLOW prototype information using the latest online version and writes the results to a python module in the given FILEPATH.')
@click.argument('FILEPATH', nargs=1, type=str)
@click.option('--matches-only', is_flag=True, help='Only update the match information that depends on the symmetry analysis settings. Will not perform and online update.')
@click.pass_context
def prototypes_update(ctx, filepath, matches_only):

    if matches_only:
        from nomad.normalizing.data.aflow_prototypes import aflow_prototypes
    else:
        # The basic AFLOW prototype data is available in a Javascript file. Here we
        # retrieve it and read only the prototype list from it.
        prototypes_file_url = 'http://aflowlib.org/CrystalDatabase/js/table_sort.js'
        r = requests.get(prototypes_file_url, allow_redirects=True)
        datastring = r.content.decode("utf-8")
        datastring = datastring.split('];')[0]
        datastring = datastring.split('= [')[1]
        data = json.loads('[' + datastring + ']')

        newdictarray = []
        n_prototypes = 0
        n_missing = 0
        for protodict in data:
            n_prototypes += 1
            newdict = {}

            # Make prototype plaintext
            prototype = BeautifulSoup(protodict["Prototype"], "html5lib").getText()

            # Add to new dictionary
            newdict['Notes'] = protodict['Notes']
            newdict['Prototype'] = prototype
            newdict['Space Group Symbol'] = protodict['Space Group Symbol']
            newdict['Space Group Number'] = protodict['Space Group Number']
            newdict['Pearsons Symbol'] = protodict['Pearson Symbol']
            newdict['Strukturbericht Designation'] = protodict['Strukturbericht Designation']
            newdict['aflow_prototype_id'] = protodict['AFLOW Prototype']
            newdict['aflow_prototype_url'] = 'http://www.aflowlib.org/CrystalDatabase/' + protodict['href'][2:]

            # Download cif or poscar if possible make ASE Atoms object if possible
            # to obtain labels, positions, cell
            cifurl = 'http://www.aflowlib.org/CrystalDatabase/CIF/' + protodict['href'][2:-5] + '.cif'
            r = requests.get(cifurl, allow_redirects=True)
            cif_str = r.content.decode("utf-8")
            cif_file = StringIO()
            cif_file.write(cif_str)
            cif_file.seek(0)
            try:
                atoms = ase.io.read(cif_file, format='cif')
            except Exception:
                print("Error in getting prototype structure from CIF: {}", format(cifurl))
                # Then try to get structure from POSCAR
                try:
                    poscarurl = 'http://www.aflowlib.org/CrystalDatabase/POSCAR/' + protodict['href'][2:-5] + '.poscar'
                    r = requests.get(poscarurl, allow_redirects=True)
                    poscar_str = r.content.decode("utf-8")
                    poscar_file = StringIO()
                    poscar_file.write(poscar_str)
                    poscar_file.seek(0)
                    atoms = ase.io.read(poscar_file, format='vasp')
                except Exception:
                    print("Error in getting prototype structure from POSCAR: {}".format(poscarurl))
                    print("Could not read prototype structure from CIF or POSCAR file for prototype: {}, {}, ".format(prototype, newdict['aflow_prototype_url']))
                    n_missing += 1
                    continue

            atom_positions = atoms.get_positions()
            atom_labels = atoms.get_chemical_symbols()
            cell = atoms.get_cell()

            newdict['lattice_vectors'] = cell.tolist()
            newdict['atom_positions'] = atom_positions.tolist()
            newdict['atom_labels'] = atom_labels
            newdictarray.append(newdict)

            print("Processed: {}".format(len(newdictarray)))

        # Sort prototype dictionaries by spacegroup and make dictionary
        structure_types_by_spacegroup = {}
        for i_sg in range(1, 231):
            protos_sg = []
            for newdict in newdictarray:
                if newdict['Space Group Number'] == i_sg:
                    protos_sg.append(newdict)
            structure_types_by_spacegroup[i_sg] = protos_sg

        # Wrap in a dictionary that can hold other data, e.g. the symmemtry tolerance parameter.
        aflow_prototypes = {
            "prototypes_by_spacegroup": structure_types_by_spacegroup
        }
        print(
            "Extracted latest AFLOW prototypes online. Total number of "
            "successfully fetched prototypes: {}, missing: {}"
            .format(n_prototypes, n_missing)
        )

    # Update matches
    n_prototypes = 0
    n_failed = 0
    n_unmatched = 0
    prototype_dict = aflow_prototypes["prototypes_by_spacegroup"]

    for aflow_spg_number, prototypes in prototype_dict.items():
        n_prototypes += len(prototypes)
        for prototype in prototypes:

            # Read prototype structure
            pos = np.array(prototype["atom_positions"])
            labels = prototype["atom_labels"]
            cell = np.array(prototype["lattice_vectors"])
            atoms = Atoms(
                symbols=labels,
                positions=pos,
                cell=cell,
                pbc=True
            )

            # Try to first see if the space group can be matched with the one in AFLOW
            tolerance = config.normalize.symmetry_tolerance
            try:
                symm = SymmetryAnalyzer(atoms, tolerance)
                spg_number = symm.get_space_group_number()
                wyckoff_matid = symm.get_wyckoff_letters_conventional()
                norm_system = symm.get_conventional_system()
            except Exception:
                n_failed += 1
            else:
                # If the space group is matched, add the MatID normalized Wyckoff
                # letters to the data.
                if spg_number == aflow_spg_number:
                    atomic_numbers = norm_system.get_atomic_numbers()
                    normalized_wyckoff_matid = get_normalized_wyckoff(atomic_numbers, wyckoff_matid)
                    prototype["normalized_wyckoff_matid"] = normalized_wyckoff_matid
                else:
                    n_unmatched += 1
    print(
        "Updated matches in AFLOW prototype library. Total number of "
        "prototypes: {}, unmatched: {}, failed: {}"
        .format(n_prototypes, n_unmatched, n_failed)
    )

    aflow_prototypes["matid_symmetry_tolerance"] = tolerance

    # Write data file to the specified path
    write_prototype_data_file(aflow_prototypes, filepath)
534
535
536
537
538
539
540


@admin.command(help='Updates the springer database in nomad.config.springer_msg_db_path.')
@click.option('--max-n-query', default=10, type=int, help='Number of unsuccessful springer request before returning an error. Default is 10.')
@click.option('--retry-time', default=120, type=int, help='Time in seconds to retry after unsuccessful request. Default is 120.')
def springer_update(max_n_query, retry_time):
    update_springer_data(max_n_query, retry_time)