forked from basak/glacier-cli
-
Notifications
You must be signed in to change notification settings - Fork 0
/
glacier.py
736 lines (633 loc) · 28.8 KB
/
glacier.py
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
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
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
109
110
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
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
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
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
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
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
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
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
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
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
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
#!/usr/bin/env python
# Copyright (c) 2012 Robie Basak
#
# Permission is hereby granted, free of charge, to any person obtaining a
# copy of this software and associated documentation files (the
# "Software"), to deal in the Software without restriction, including
# without limitation the rights to use, copy, modify, merge, publish, dis-
# tribute, sublicense, and/or sell copies of the Software, and to permit
# persons to whom the Software is furnished to do so, subject to the fol-
# lowing conditions:
#
# The above copyright notice and this permission notice shall be included
# in all copies or substantial portions of the Software.
#
# THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS
# OR IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABIL-
# ITY, FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT
# SHALL THE AUTHOR BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER LIABILITY,
# WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
# OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS
# IN THE SOFTWARE.
from __future__ import print_function
from __future__ import unicode_literals
import argparse
import calendar
import errno
import itertools
import os
import os.path
import sys
import time
import boto.glacier
import iso8601
import sqlalchemy
import sqlalchemy.ext.declarative
import sqlalchemy.orm
__version__ = '0.1.0'
# There is a lag between an archive being created and the archive
# appearing on an inventory. Even if the inventory has an InventoryDate
# of after the archive was created, it still doesn't necessarily appear.
# So only warn of a missing archive if the archive still hasn't appeared
# on an inventory created INVENTORY_LAG seconds after the archive was
# uploaded successfully.
INVENTORY_LAG = 24 * 60 * 60 * 3
PROGRAM_NAME = 'glacier'
class ConsoleError(RuntimeError):
def __init__(self, m):
self.message = m
class RetryConsoleError(ConsoleError): pass
def info(message):
print(insert_prefix_to_lines('%s: info: ' % PROGRAM_NAME, message),
file=sys.stderr)
def warn(message):
print(insert_prefix_to_lines('%s: warning: ' % PROGRAM_NAME, message),
file=sys.stderr)
def mkdir_p(path):
"""Create path if it doesn't exist already"""
try:
os.makedirs(path)
except OSError as e:
if e.errno != errno.EEXIST:
raise
def insert_prefix_to_lines(prefix, lines):
return "\n".join([prefix + line for line in lines.split("\n")])
def iso8601_to_unix_timestamp(iso8601_date_str):
return calendar.timegm(iso8601.parse_date(iso8601_date_str).utctimetuple())
def get_user_cache_dir():
xdg_cache_home = os.getenv('XDG_CACHE_HOME')
if xdg_cache_home is not None:
return xdg_cache_home
home = os.getenv('HOME')
if home is None:
raise RuntimeError('Cannot find user home directory')
return os.path.join(home, '.cache')
class Cache(object):
Base = sqlalchemy.ext.declarative.declarative_base()
class Archive(Base):
__tablename__ = 'archive'
id = sqlalchemy.Column(sqlalchemy.String, primary_key=True)
name = sqlalchemy.Column(sqlalchemy.String)
vault = sqlalchemy.Column(sqlalchemy.String, nullable=False)
key = sqlalchemy.Column(sqlalchemy.String, nullable=False)
last_seen_upstream = sqlalchemy.Column(sqlalchemy.Integer)
created_here = sqlalchemy.Column(sqlalchemy.Integer)
deleted_here = sqlalchemy.Column(sqlalchemy.Integer)
def __init__(self, *args, **kwargs):
self.created_here = time.time()
super(Cache.Archive, self).__init__(*args, **kwargs)
Session = sqlalchemy.orm.sessionmaker()
def __init__(self, key, db_path=None):
self.key = key
if db_path is None:
db_path = os.path.join(get_user_cache_dir(), 'glacier-cli', 'db')
if db_path != ':memory:':
mkdir_p(os.path.dirname(db_path))
self.engine = sqlalchemy.create_engine('sqlite:///%s' % db_path)
self.Base.metadata.create_all(self.engine)
self.Session.configure(bind=self.engine)
self.session = self.Session()
def add_archive(self, vault, name, id):
self.session.add(self.Archive(key=self.key,
vault=vault, name=name, id=id))
self.session.commit()
def _get_archive_query_by_ref(self, vault, ref):
if ref.startswith('id:'):
filter = {'id': ref[3:]}
elif ref.startswith('name:'):
filter = {'name': ref[5:]}
else:
filter = {'name': ref}
return self.session.query(self.Archive).filter_by(
key=self.key, vault=vault, deleted_here=None, **filter)
def get_archive_id(self, vault, ref):
try:
result = self._get_archive_query_by_ref(vault, ref).one()
except sqlalchemy.orm.exc.NoResultFound:
raise KeyError(ref)
return result.id
def get_archive_name(self, vault, ref):
try:
result = self._get_archive_query_by_ref(vault, ref).one()
except sqlalchemy.orm.exc.NoResultFound:
raise KeyError(ref)
return result.name
def get_archive_last_seen(self, vault, ref):
try:
result = self._get_archive_query_by_ref(vault, ref).one()
except sqlalchemy.orm.exc.NoResultFound:
raise KeyError(ref)
return result.last_seen_upstream or result.created_here
def delete_archive(self, vault, ref):
try:
result = self._get_archive_query_by_ref(vault, ref).one()
except sqlalchemy.orm.exc.NoResultFound:
raise KeyError(name)
result.deleted_here = time.time()
self.session.commit()
@staticmethod
def _archive_ref(archive, force_id=False):
if archive.name and not force_id:
if (archive.name.startswith('name:') or
archive.name.startswith('id:')):
return "name:%s" % archive.name
else:
return archive.name
else:
return 'id:' + archive.id
def _get_archive_list_objects(self, vault):
for archive in (
self.session.query(self.Archive).
filter_by(key=self.key,
vault=vault,
deleted_here=None).
order_by(self.Archive.name)):
yield archive
def get_archive_list(self, vault):
def force_id(archive):
return "\t".join([
self._archive_ref(archive, force_id=True),
"%s" % archive.name
])
for archive_name, archive_iterator in (
itertools.groupby(
self._get_archive_list_objects(vault),
lambda archive: archive.name)):
# Yield self._archive_ref(..., force_id=True) if there is more than
# one archive with the same name; otherwise use force_id=False.
first_archive = next(archive_iterator)
try:
second_archive = next(archive_iterator)
except StopIteration:
yield self._archive_ref(first_archive, force_id=False)
else:
yield force_id(first_archive)
yield force_id(second_archive)
for subsequent_archive in archive_iterator:
yield force_id(subsequent_archive)
def get_archive_list_with_ids(self, vault):
for archive in self._get_archive_list_objects(vault):
yield "\t".join([
self._archive_ref(archive, force_id=True),
"%s" % archive.name,
])
def mark_seen_upstream(
self, vault, id, name, upstream_creation_date,
upstream_inventory_date, upstream_inventory_job_creation_date,
fix=False):
# Inventories don't get recreated unless the vault has changed.
# See: https://forums.aws.amazon.com/thread.jspa?threadID=106541
#
# The cache's last_seen_upstream is supposed to contain a point in time
# at which we know for sure that an archive existed, but this can fall
# too far behind if a vault doesn't change. So assume that an archive
# that appears in an inventory that hasn't been updated recently
# nevertheless existed at around the time the inventory _could_ have
# been regenerated, ie. at some point prior to the date that we
# requested the inventory retrieval job.
#
# This is preferred over using the job completion date as an archive
# could in theory be deleted while an inventory job is in progress and
# would still appear in that inventory.
#
# Making up a date prior to the inventory job's creation could mean
# that last_seen_upstream ends up claiming that an archive existed even
# before it was created, but this will not cause a problem. Better that
# it's too far back in time than too far ahead.
#
# With thanks to Wolfgang Nagele.
last_seen_upstream = max(
upstream_inventory_date,
upstream_inventory_job_creation_date - INVENTORY_LAG
)
try:
archive = self.session.query(self.Archive).filter_by(
key=self.key, vault=vault, id=id).one()
except sqlalchemy.orm.exc.NoResultFound:
self.session.add(
self.Archive(
key=self.key, vault=vault, name=name, id=id,
last_seen_upstream=last_seen_upstream
)
)
else:
if not archive.name:
archive.name = name
elif archive.name != name:
if fix:
warn('archive %r appears to have changed name from %r ' %
(archive.id, archive.name) + 'to %r (fixed)' % (name))
archive.name = name
else:
warn('archive %r appears to have changed name from %r ' %
(archive.id, archive.name) + 'to %r' % (name))
if archive.deleted_here:
archive_ref = self._archive_ref(archive)
if archive.deleted_here < upstream_inventory_date:
warn('archive %r marked deleted but still present' %
archive_ref)
else:
warn('archive %r deletion not yet in inventory' %
archive_ref)
archive.last_seen_upstream = last_seen_upstream
def mark_only_seen(self, vault, inventory_date, ids, fix=False):
upstream_ids = set(ids)
our_ids = set([r[0] for r in
self.session.query(self.Archive.id)
.filter_by(key=self.key, vault=vault).all()])
missing_ids = our_ids - upstream_ids
for id in missing_ids:
archive = (self.session.query(self.Archive)
.filter_by(key=self.key,
vault=vault, id=id)
.one())
archive_ref = self._archive_ref(archive)
if archive.deleted_here and archive.deleted_here < inventory_date:
self.session.delete(archive)
info('deleted archive %r has left inventory; ' % archive_ref +
'removed from cache')
elif not archive.deleted_here and (
archive.last_seen_upstream or
(archive.created_here and
archive.created_here < inventory_date - INVENTORY_LAG)):
if fix:
self.session.delete(archive)
warn('archive disappeared: %r (removed from cache)' %
archive_ref)
else:
warn('archive disappeared: %r' % archive_ref)
else:
warn('new archive not yet in inventory: %r' % archive_ref)
def mark_commit(self):
self.session.commit()
def get_connection_account(connection):
"""Return some account key associated with the connection.
This is used to key a cache, so that the same cache can serve multiple
accounts. The only requirement is that multiple namespaces of vaults and/or
archives can never collide for connections that return the same key with
this function. The cache will more more efficient if the same Glacier
namespace sets always result in the same key.
"""
return connection.layer1.aws_access_key_id
def find_retrieval_jobs(vault, archive_id):
return [job for job in vault.list_jobs() if job.archive_id == archive_id]
def find_inventory_jobs(vault, max_age_hours=0):
if max_age_hours:
def recent_enough(job):
if not job.completed:
return True
completion_date = iso8601_to_unix_timestamp(job.completion_date)
return completion_date > time.time() - max_age_hours * 60 * 60
else:
def recent_enough(job):
return not job.completed
return [job for job in vault.list_jobs()
if job.action == 'InventoryRetrieval' and recent_enough(job)]
def find_complete_job(jobs):
for job in sorted(filter(lambda job: job.completed, jobs), key=lambda job: iso8601.parse_date(job.completion_date), reverse=True):
return job
def has_pending_job(jobs):
return any(filter(lambda job: not job.completed, jobs))
def update_job_list(jobs):
for i, job in enumerate(jobs):
jobs[i] = job.vault.get_job(job.id)
def job_oneline(conn, cache, vault, job):
action_letter = {'ArchiveRetrieval': 'a',
'InventoryRetrieval': 'i'}[job.action]
status_letter = {'InProgress': 'p',
'Succeeded': 'd',
'Failed': 'e'}[job.status_code]
date = job.completion_date
if not date:
date = job.creation_date
if job.action == 'ArchiveRetrieval':
try:
name = cache.get_archive_name(vault.name, 'id:' + job.archive_id)
except KeyError:
name = None
if name is None:
name = 'id:' + job.archive_id
elif job.action == 'InventoryRetrieval':
name = ''
return '{action_letter}/{status_letter} {date} {vault.name:10} {name}'.format(
**locals())
def wait_until_job_completed(jobs, sleep=600, tries=144):
update_job_list(jobs)
job = find_complete_job(jobs)
while not job:
tries -= 1
if tries < 0:
raise RuntimeError('Timed out waiting for job completion')
time.sleep(sleep)
update_job_list(jobs)
job = find_complete_job(jobs)
return job
class App(object):
def job_list(self):
for vault in self.connection.list_vaults():
job_list = [job_oneline(self.connection,
self.cache,
vault,
job)
for job in vault.list_jobs()]
if job_list:
print(*job_list, sep="\n")
def vault_list(self):
print(*[vault.name for vault in self.connection.list_vaults()],
sep="\n")
def vault_create(self):
self.connection.create_vault(self.args.name)
def _vault_sync_reconcile(self, vault, job, fix=False):
response = job.get_output()
inventory_date = iso8601_to_unix_timestamp(response['InventoryDate'])
job_creation_date = iso8601_to_unix_timestamp(job.creation_date)
seen_ids = []
for archive in response['ArchiveList']:
id = archive['ArchiveId']
name = archive['ArchiveDescription']
creation_date = iso8601_to_unix_timestamp(archive['CreationDate'])
self.cache.mark_seen_upstream(
vault=vault.name,
id=id,
name=name,
upstream_creation_date=creation_date,
upstream_inventory_date=inventory_date,
upstream_inventory_job_creation_date=job_creation_date,
fix=fix)
seen_ids.append(id)
self.cache.mark_only_seen(vault.name, inventory_date, seen_ids,
fix=fix)
self.cache.mark_commit()
def _vault_sync(self, vault_name, max_age_hours, fix, wait):
vault = self.connection.get_vault(vault_name)
inventory_jobs = find_inventory_jobs(vault,
max_age_hours=max_age_hours)
complete_job = find_complete_job(inventory_jobs)
if complete_job:
self._vault_sync_reconcile(vault, complete_job, fix=fix)
elif has_pending_job(inventory_jobs):
if wait:
complete_job = wait_until_job_completed(inventory_jobs)
else:
raise RetryConsoleError('job still pending for inventory on %r' %
vault.name)
else:
job_id = vault.retrieve_inventory()
job = vault.get_job(job_id)
if wait:
wait_until_job_completed([job])
self._vault_sync_reconcile(vault, job, fix=fix)
else:
raise RetryConsoleError('queued inventory job for %r' %
vault.name)
def vault_sync(self):
return self._vault_sync(vault_name=self.args.name,
max_age_hours=self.args.max_age_hours,
fix=self.args.fix,
wait=self.args.wait)
def archive_list(self):
if self.args.force_ids:
archive_list = list(self.cache.get_archive_list_with_ids(
self.args.vault))
else:
archive_list = list(self.cache.get_archive_list(self.args.vault))
if archive_list:
print(*archive_list, sep="\n")
def archive_upload(self):
# XXX: "Leading whitespace in archive descriptions is removed."
# XXX: "The description must be less than or equal to 1024 bytes. The
# allowable characters are 7 bit ASCII without control codes,
# specifically ASCII values 32-126 decimal or 0x20-0x7E
# hexadecimal."
if self.args.name is not None:
name = self.args.name
else:
try:
full_name = self.args.file.name
except:
raise RuntimeError('Archive name not specified. Use --name')
name = os.path.basename(full_name)
vault = self.connection.get_vault(self.args.vault)
archive_id = vault.create_archive_from_file(
file_obj=self.args.file, description=name)
self.cache.add_archive(self.args.vault, name, archive_id)
@staticmethod
def _write_archive_retrieval_job(f, job, multipart_size):
if job.archive_size > multipart_size:
def fetch(start, end):
byte_range = start, end-1
f.write(job.get_output(byte_range).read())
whole_parts = job.archive_size // multipart_size
for first_byte in xrange(0, whole_parts * multipart_size,
multipart_size):
fetch(first_byte, first_byte + multipart_size)
remainder = job.archive_size % multipart_size
if remainder:
fetch(job.archive_size - remainder, job.archive_size)
else:
f.write(job.get_output().read())
# Make sure that the file now exactly matches the downloaded archive,
# even if the file existed before and was longer.
try:
f.truncate(job.archive_size)
except IOError as e:
# Allow ESPIPE, since the "file" couldn't have existed before in
# this case.
if e.errno != errno.ESPIPE:
raise
@classmethod
def _archive_retrieve_completed(cls, args, job, name):
if args.output_filename == '-':
cls._write_archive_retrieval_job(
sys.stdout, job, args.multipart_size)
else:
if args.output_filename:
filename = args.output_filename
else:
filename = os.path.basename(name)
with open(filename, 'wb') as f:
cls._write_archive_retrieval_job(f, job, args.multipart_size)
def archive_retrieve_one(self, name):
try:
archive_id = self.cache.get_archive_id(self.args.vault, name)
except KeyError:
raise ConsoleError('archive %r not found' % name)
vault = self.connection.get_vault(self.args.vault)
retrieval_jobs = find_retrieval_jobs(vault, archive_id)
complete_job = find_complete_job(retrieval_jobs)
if complete_job:
self._archive_retrieve_completed(self.args, complete_job, name)
elif has_pending_job(retrieval_jobs):
if self.args.wait:
complete_job = wait_until_job_completed(retrieval_jobs)
self._archive_retrieve_completed(self.args, complete_job, name)
else:
raise RetryConsoleError('job still pending for archive %r' % name)
else:
# create an archive retrieval job
job = vault.retrieve_archive(archive_id)
if self.args.wait:
wait_until_job_completed([job])
self._archive_retrieve_completed(self.args, job, name)
else:
raise RetryConsoleError('queued retrieval job for archive %r' % name)
def archive_retrieve(self):
if len(self.args.names) > 1 and self.args.output_filename:
raise ConsoleError('cannot specify output filename with multi-archive retrieval')
success_list = []
retry_list = []
for name in self.args.names:
try:
self.archive_retrieve_one(name)
except RetryConsoleError as e:
retry_list.append(e.message)
else:
success_list.append('retrieved archive %r' % name)
if retry_list:
message_list = success_list + retry_list
raise RetryConsoleError("\n".join(message_list))
def archive_delete(self):
try:
archive_id = self.cache.get_archive_id(
self.args.vault, self.args.name)
except KeyError:
raise ConsoleError('archive %r not found' % self.args.name)
vault = self.connection.get_vault(self.args.vault)
vault.delete_archive(archive_id)
self.cache.delete_archive(self.args.vault, self.args.name)
def archive_checkpresent(self):
try:
last_seen = self.cache.get_archive_last_seen(
self.args.vault, self.args.name)
except KeyError:
if self.args.wait:
last_seen = None
else:
if not self.args.quiet:
print(
'archive %r not found' % self.args.name,
file=sys.stderr)
return
def too_old(last_seen):
return (not last_seen or
not self.args.max_age_hours or
(last_seen <
time.time() - self.args.max_age_hours * 60 * 60))
if too_old(last_seen):
# Not recent enough
try:
self._vault_sync(vault_name=self.args.vault,
max_age_hours=self.args.max_age_hours,
fix=False,
wait=self.args.wait)
except RetryConsoleError:
pass
else:
try:
last_seen = self.cache.get_archive_last_seen(
self.args.vault, self.args.name)
except KeyError:
if not self.args.quiet:
print(('archive %r not found, but it may ' +
'not be in the inventory yet')
% self.args.name, file=sys.stderr)
return
if too_old(last_seen):
if not self.args.quiet:
print(('archive %r found, but has not been seen ' +
'recently enough to consider it present') %
self.args.name, file=sys.stderr)
return
print(self.args.name)
def parse_args(self, args=None):
parser = argparse.ArgumentParser()
parser.add_argument('--region', default='us-east-1')
subparsers = parser.add_subparsers()
vault_subparser = subparsers.add_parser('vault').add_subparsers()
vault_subparser.add_parser('list').set_defaults(func=self.vault_list)
vault_create_subparser = vault_subparser.add_parser('create')
vault_create_subparser.set_defaults(func=self.vault_create)
vault_create_subparser.add_argument('name')
vault_sync_subparser = vault_subparser.add_parser('sync')
vault_sync_subparser.set_defaults(func=self.vault_sync)
vault_sync_subparser.add_argument('name', metavar='vault_name')
vault_sync_subparser.add_argument('--wait', action='store_true')
vault_sync_subparser.add_argument('--fix', action='store_true')
vault_sync_subparser.add_argument('--max-age', type=int, default=24,
dest='max_age_hours')
archive_subparser = subparsers.add_parser('archive').add_subparsers()
archive_list_subparser = archive_subparser.add_parser('list')
archive_list_subparser.set_defaults(func=self.archive_list)
archive_list_subparser.add_argument('--force-ids', action='store_true')
archive_list_subparser.add_argument('vault')
archive_upload_subparser = archive_subparser.add_parser('upload')
archive_upload_subparser.set_defaults(func=self.archive_upload)
archive_upload_subparser.add_argument('vault')
archive_upload_subparser.add_argument('file',
type=argparse.FileType('rb'))
archive_upload_subparser.add_argument('--name')
archive_retrieve_subparser = archive_subparser.add_parser('retrieve')
archive_retrieve_subparser.set_defaults(func=self.archive_retrieve)
archive_retrieve_subparser.add_argument('vault')
archive_retrieve_subparser.add_argument('names', nargs='+',
metavar='name')
archive_retrieve_subparser.add_argument('--multipart-size', type=int,
default=(8*1024*1024))
archive_retrieve_subparser.add_argument('-o', dest='output_filename',
metavar='OUTPUT_FILENAME')
archive_retrieve_subparser.add_argument('--wait', action='store_true')
archive_delete_subparser = archive_subparser.add_parser('delete')
archive_delete_subparser.set_defaults(func=self.archive_delete)
archive_delete_subparser.add_argument('vault')
archive_delete_subparser.add_argument('name')
archive_checkpresent_subparser = archive_subparser.add_parser(
'checkpresent')
archive_checkpresent_subparser.set_defaults(
func=self.archive_checkpresent)
archive_checkpresent_subparser.add_argument('vault')
archive_checkpresent_subparser.add_argument('name')
archive_checkpresent_subparser.add_argument('--wait',
action='store_true')
archive_checkpresent_subparser.add_argument('--quiet',
action='store_true')
archive_checkpresent_subparser.add_argument(
'--max-age', type=int, default=80, dest='max_age_hours')
job_subparser = subparsers.add_parser('job').add_subparsers()
job_subparser.add_parser('list').set_defaults(func=self.job_list)
return parser.parse_args(args)
def __init__(self, args=None, connection=None, cache=None):
args = self.parse_args(args)
if connection is None:
connection = boto.glacier.connect_to_region(args.region)
if cache is None:
cache = Cache(get_connection_account(connection))
self.connection = connection
self.cache = cache
self.args = args
def main(self):
try:
self.args.func()
except RetryConsoleError as e:
message = insert_prefix_to_lines(PROGRAM_NAME + ': ', e.message)
print(message, file=sys.stderr)
# From sysexits.h:
# "temp failure; user is invited to retry"
sys.exit(75) # EX_TEMPFAIL
except ConsoleError as e:
message = insert_prefix_to_lines(PROGRAM_NAME + ': ', e.message)
print(message, file=sys.stderr)
sys.exit(1)
def main():
App().main()
if __name__ == '__main__':
main()