OLD | NEW |
| (Empty) |
1 # Copyright 2011 Google Inc. All Rights Reserved. | |
2 # | |
3 # Licensed under the Apache License, Version 2.0 (the "License"); | |
4 # you may not use this file except in compliance with the License. | |
5 # You may obtain a copy of the License at | |
6 # | |
7 # http://www.apache.org/licenses/LICENSE-2.0 | |
8 # | |
9 # Unless required by applicable law or agreed to in writing, software | |
10 # distributed under the License is distributed on an "AS IS" BASIS, | |
11 # WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. | |
12 # See the License for the specific language governing permissions and | |
13 # limitations under the License. | |
14 | |
15 from boto.s3.deletemarker import DeleteMarker | |
16 from gslib.bucket_listing_ref import BucketListingRef | |
17 from gslib.command import Command | |
18 from gslib.command import COMMAND_NAME | |
19 from gslib.command import COMMAND_NAME_ALIASES | |
20 from gslib.command import CONFIG_REQUIRED | |
21 from gslib.command import FILE_URIS_OK | |
22 from gslib.command import MAX_ARGS | |
23 from gslib.command import MIN_ARGS | |
24 from gslib.command import PROVIDER_URIS_OK | |
25 from gslib.command import SUPPORTED_SUB_ARGS | |
26 from gslib.command import URIS_START_ARG | |
27 from gslib.exception import CommandException | |
28 from gslib.help_provider import HELP_NAME | |
29 from gslib.help_provider import HELP_NAME_ALIASES | |
30 from gslib.help_provider import HELP_ONE_LINE_SUMMARY | |
31 from gslib.help_provider import HELP_TEXT | |
32 from gslib.help_provider import HelpType | |
33 from gslib.help_provider import HELP_TYPE | |
34 from gslib.plurality_checkable_iterator import PluralityCheckableIterator | |
35 from gslib.util import ListingStyle | |
36 from gslib.util import MakeHumanReadable | |
37 from gslib.util import NO_MAX | |
38 from gslib.wildcard_iterator import ContainsWildcard | |
39 import boto | |
40 | |
41 _detailed_help_text = (""" | |
42 <B>SYNOPSIS</B> | |
43 gsutil ls [-a] [-b] [-l] [-L] [-R] [-p proj_id] uri... | |
44 | |
45 | |
46 <B>LISTING PROVIDERS, BUCKETS, SUBDIRECTORIES, AND OBJECTS</B> | |
47 If you run gsutil ls without URIs, it lists all of the Google Cloud Storage | |
48 buckets under your default project ID: | |
49 | |
50 gsutil ls | |
51 | |
52 (For details about projects, see "gsutil help projects" and also the -p | |
53 option in the OPTIONS section below.) | |
54 | |
55 If you specify one or more provider URIs, gsutil ls will list buckets at | |
56 each listed provider: | |
57 | |
58 gsutil ls gs:// | |
59 | |
60 If you specify bucket URIs, gsutil ls will list objects at the top level of | |
61 each bucket, along with the names of each subdirectory. For example: | |
62 | |
63 gsutil ls gs://bucket | |
64 | |
65 might produce output like: | |
66 | |
67 gs://bucket/obj1.htm | |
68 gs://bucket/obj2.htm | |
69 gs://bucket/images1/ | |
70 gs://bucket/images2/ | |
71 | |
72 The "/" at the end of the last 2 URIs tells you they are subdirectories, | |
73 which you can list using: | |
74 | |
75 gsutil ls gs://bucket/images* | |
76 | |
77 If you specify object URIs, gsutil ls will list the specified objects. For | |
78 example: | |
79 | |
80 gsutil ls gs://bucket/*.txt | |
81 | |
82 will list all files whose name matches the above wildcard at the top level | |
83 of the bucket. | |
84 | |
85 See "gsutil help wildcards" for more details on working with wildcards. | |
86 | |
87 | |
88 <B>DIRECTORY BY DIRECTORY, FLAT, and RECURSIVE LISTINGS</B> | |
89 Listing a bucket or subdirectory (as illustrated near the end of the previous | |
90 section) only shows the objects and names of subdirectories it contains. You | |
91 can list all objects in a bucket by using the -R option. For example: | |
92 | |
93 gsutil ls -R gs://bucket | |
94 | |
95 will list the top-level objects and buckets, then the objects and | |
96 buckets under gs://bucket/images1, then those under gs://bucket/images2, etc. | |
97 | |
98 If you want to see all objects in the bucket in one "flat" listing use the | |
99 recursive ("**") wildcard, like: | |
100 | |
101 gsutil ls -R gs://bucket/** | |
102 | |
103 or, for a flat listing of a subdirectory: | |
104 | |
105 gsutil ls -R gs://bucket/dir/** | |
106 | |
107 | |
108 <B>LISTING OBJECT DETAILS</B> | |
109 If you specify the -l option, gsutil will output additional information | |
110 about each matching provider, bucket, subdirectory, or object. For example, | |
111 | |
112 gsutil ls -l gs://bucket/*.txt | |
113 | |
114 will print the object size, creation time stamp, and name of each matching | |
115 object, along with the total count and sum of sizes of all matching objects: | |
116 | |
117 2276224 2012-03-02T19:25:17 gs://bucket/obj1 | |
118 3914624 2012-03-02T19:30:27 gs://bucket/obj2 | |
119 TOTAL: 2 objects, 6190848 bytes (5.9 MB) | |
120 | |
121 Note that the total listed in parentheses above is in mebibytes (or gibibytes, | |
122 tebibytes, etc.), which corresponds to the unit of billing measurement for | |
123 Google Cloud Storage. | |
124 | |
125 You can get a listing of all the objects in the top-level bucket directory | |
126 (along with the total count and sum of sizes) using a command like: | |
127 | |
128 gsutil ls -l gs://bucket | |
129 | |
130 To print additional detail about objects and buckets use the gsutil ls -L | |
131 option. For example: | |
132 | |
133 gsutil ls -L gs://bucket/obj1 | |
134 | |
135 will print something like: | |
136 | |
137 gs://bucket/obj1: | |
138 Creation Time: Fri, 02 Mar 2012 19:25:17 GMT | |
139 Size: 2276224 | |
140 Cache-Control: private, max-age=0 | |
141 Content-Type: application/x-executable | |
142 ETag: 5ca6796417570a586723b7344afffc81 | |
143 ACL: <Owner:00b4903a97163d99003117abe64d292561d2b4074fc90ce5c
0e35ac45f66ad70, <<UserById: 00b4903a97163d99003117abe64d292561d2b4074fc90ce5c0e
35ac45f66ad70>: u'FULL_CONTROL'>> | |
144 TOTAL: 1 objects, 2276224 bytes (2.17 MB) | |
145 | |
146 Note that the -L option is slower and more costly to use than the -l option, | |
147 because it makes a bucket listing request followed by a HEAD request for | |
148 each individual object (rather than just parsing the information it needs | |
149 out of a single bucket listing, the way the -l option does). | |
150 | |
151 See also "gsutil help getacl" for getting a more readable version of the ACL. | |
152 | |
153 | |
154 <B>LISTING BUCKET DETAILS</B> | |
155 If you want to see information about the bucket itself, use the -b | |
156 option. For example: | |
157 | |
158 gsutil ls -L -b gs://bucket | |
159 | |
160 will print something like: | |
161 | |
162 gs://bucket/ : | |
163 24 objects, 29.83 KB | |
164 StorageClass: STANDARD | |
165 LocationConstraint: US | |
166 Versioning enabled: True | |
167 ACL: <Owner:00b4903a9740e42c29800f53bd5a9a62a2f96eb3f64a4313a115df3f
3a776bf7, <<GroupById: 00b4903a9740e42c29800f53bd5a9a62a2f96eb3f64a4313a115df3f3
a776bf7>: u'FULL_CONTROL'>> | |
168 Default ACL: <> | |
169 TOTAL: 24 objects, 30544 bytes (29.83 KB) | |
170 | |
171 | |
172 <B>OPTIONS</B> | |
173 -l Prints long listing (owner, length). | |
174 | |
175 -L Prints even more detail than -l. This is a separate option because | |
176 it makes additional service requests (so, takes longer and adds | |
177 requests costs). | |
178 | |
179 -b Prints info about the bucket when used with a bucket URI. | |
180 | |
181 -p proj_id Specifies the project ID to use for listing buckets. | |
182 | |
183 -R, -r Requests a recursive listing. | |
184 | |
185 -a Includes non-current object versions / generations in the listing | |
186 (only useful with a versioning-enabled bucket). If combined with | |
187 -l option also prints meta-generation for each listed object. | |
188 """) | |
189 | |
190 | |
191 class LsCommand(Command): | |
192 """Implementation of gsutil ls command.""" | |
193 | |
194 # Command specification (processed by parent class). | |
195 command_spec = { | |
196 # Name of command. | |
197 COMMAND_NAME : 'ls', | |
198 # List of command name aliases. | |
199 COMMAND_NAME_ALIASES : ['dir', 'list'], | |
200 # Min number of args required by this command. | |
201 MIN_ARGS : 0, | |
202 # Max number of args required by this command, or NO_MAX. | |
203 MAX_ARGS : NO_MAX, | |
204 # Getopt-style string specifying acceptable sub args. | |
205 SUPPORTED_SUB_ARGS : 'ablLp:rR', | |
206 # True if file URIs acceptable for this command. | |
207 FILE_URIS_OK : False, | |
208 # True if provider-only URIs acceptable for this command. | |
209 PROVIDER_URIS_OK : True, | |
210 # Index in args of first URI arg. | |
211 URIS_START_ARG : 0, | |
212 # True if must configure gsutil before running command. | |
213 CONFIG_REQUIRED : True, | |
214 } | |
215 help_spec = { | |
216 # Name of command or auxiliary help info for which this help applies. | |
217 HELP_NAME : 'ls', | |
218 # List of help name aliases. | |
219 HELP_NAME_ALIASES : ['dir', 'list'], | |
220 # Type of help: | |
221 HELP_TYPE : HelpType.COMMAND_HELP, | |
222 # One line summary of this help. | |
223 HELP_ONE_LINE_SUMMARY : 'List providers, buckets, or objects', | |
224 # The full help text. | |
225 HELP_TEXT : _detailed_help_text, | |
226 } | |
227 | |
228 def _PrintBucketInfo(self, bucket_uri, listing_style): | |
229 """Print listing info for given bucket. | |
230 | |
231 Args: | |
232 bucket_uri: StorageUri being listed. | |
233 listing_style: ListingStyle enum describing type of output desired. | |
234 | |
235 Returns: | |
236 Tuple (total objects, total bytes) in the bucket. | |
237 """ | |
238 bucket_objs = 0 | |
239 bucket_bytes = 0 | |
240 if listing_style == ListingStyle.SHORT: | |
241 print bucket_uri | |
242 else: | |
243 for obj in self.WildcardIterator( | |
244 bucket_uri.clone_replace_name('**')).IterKeys(): | |
245 bucket_objs += 1 | |
246 bucket_bytes += obj.size | |
247 if listing_style == ListingStyle.LONG: | |
248 print '%s : %s objects, %s' % ( | |
249 bucket_uri, bucket_objs, MakeHumanReadable(bucket_bytes)) | |
250 else: # listing_style == ListingStyle.LONG_LONG: | |
251 location_constraint = bucket_uri.get_location(validate=False, | |
252 headers=self.headers) | |
253 location_output = '' | |
254 if location_constraint: | |
255 location_output = '\n\tLocationConstraint: %s' % location_constraint | |
256 storage_class = bucket_uri.get_storage_class(validate=False, | |
257 headers=self.headers) | |
258 self.proj_id_handler.FillInProjectHeaderIfNeeded( | |
259 'get_acl', bucket_uri, self.headers) | |
260 print('%s :\n\t%d objects, %s\n\tStorageClass: %s%s\n' | |
261 '\tVersioning enabled: %s\n\tACL: %s\n' | |
262 '\tDefault ACL: %s' % ( | |
263 bucket_uri, bucket_objs, MakeHumanReadable(bucket_bytes), | |
264 storage_class, location_output, | |
265 bucket_uri.get_versioning_config(), | |
266 bucket_uri.get_acl(False, self.headers), | |
267 bucket_uri.get_def_acl(False, self.headers))) | |
268 return (bucket_objs, bucket_bytes) | |
269 | |
270 def _UriStrForObj(self, uri, obj): | |
271 """Constructs a URI string for the given object. | |
272 | |
273 For example if we were iterating gs://*, obj could be an object in one | |
274 of the user's buckets enumerated by the ls command. | |
275 | |
276 Args: | |
277 uri: base StorageUri being iterated. | |
278 obj: object (Key) being listed. | |
279 | |
280 Returns: | |
281 URI string. | |
282 """ | |
283 version_info = '' | |
284 if self.all_versions: | |
285 if uri.get_provider().name == 'google' and obj.generation: | |
286 version_info = '#%s' % obj.generation | |
287 elif uri.get_provider().name == 'aws' and obj.version_id: | |
288 if isinstance(obj, DeleteMarker): | |
289 version_info = '#<DeleteMarker>' + str(obj.version_id) | |
290 else: | |
291 version_info = '#' + str(obj.version_id) | |
292 else: | |
293 version_info = '' | |
294 return '%s://%s/%s%s' % (uri.scheme, obj.bucket.name, obj.name, | |
295 version_info) | |
296 | |
297 def _PrintInfoAboutBucketListingRef(self, bucket_listing_ref, listing_style): | |
298 """Print listing info for given bucket_listing_ref. | |
299 | |
300 Args: | |
301 bucket_listing_ref: BucketListing being listed. | |
302 listing_style: ListingStyle enum describing type of output desired. | |
303 | |
304 Returns: | |
305 Tuple (number of objects, | |
306 object length, if listing_style is one of the long listing formats) | |
307 | |
308 Raises: | |
309 Exception: if calling bug encountered. | |
310 """ | |
311 uri = bucket_listing_ref.GetUri() | |
312 obj = bucket_listing_ref.GetKey() | |
313 uri_str = self._UriStrForObj(uri, obj) | |
314 if listing_style == ListingStyle.SHORT: | |
315 print uri_str.encode('utf-8') | |
316 return (1, 0) | |
317 elif listing_style == ListingStyle.LONG: | |
318 # Exclude timestamp fractional secs (example: 2010-08-23T12:46:54.187Z). | |
319 timestamp = obj.last_modified[:19].decode('utf8').encode('ascii') | |
320 if not isinstance(obj, DeleteMarker): | |
321 if self.all_versions: | |
322 print '%10s %s %s meta_generation=%s' % ( | |
323 obj.size, timestamp, uri_str.encode('utf-8'), obj.meta_generation) | |
324 else: | |
325 print '%10s %s %s' % (obj.size, timestamp, uri_str.encode('utf-8')) | |
326 return (1, obj.size) | |
327 else: | |
328 if self.all_versions: | |
329 print '%10s %s %s meta_generation=%s' % ( | |
330 0, timestamp, uri_str.encode('utf-8'), obj.meta_generation) | |
331 else: | |
332 print '%10s %s %s' % (0, timestamp, uri_str.encode('utf-8')) | |
333 return (0, 1) | |
334 elif listing_style == ListingStyle.LONG_LONG: | |
335 # Run in a try/except clause so we can continue listings past | |
336 # access-denied errors (which can happen because user may have READ | |
337 # permission on object and thus see the bucket listing data, but lack | |
338 # FULL_CONTROL over individual objects and thus not be able to read | |
339 # their ACLs). | |
340 try: | |
341 print '%s:' % uri_str.encode('utf-8') | |
342 suri = self.suri_builder.StorageUri(uri_str) | |
343 obj = suri.get_key(False) | |
344 print '\tCreation time:\t%s' % obj.last_modified | |
345 if obj.cache_control: | |
346 print '\tCache-Control:\t%s' % obj.cache_control | |
347 if obj.content_disposition: | |
348 print '\tContent-Disposition:\t%s' % obj.content_disposition | |
349 if obj.content_encoding: | |
350 print '\tContent-Encoding:\t%s' % obj.content_encoding | |
351 if obj.content_language: | |
352 print '\tContent-Language:\t%s' % obj.content_language | |
353 print '\tContent-Length:\t%s' % obj.size | |
354 print '\tContent-Type:\t%s' % obj.content_type | |
355 if obj.metadata: | |
356 prefix = uri.get_provider().metadata_prefix | |
357 for name in obj.metadata: | |
358 print '\t%s%s:\t%s' % (prefix, name, obj.metadata[name]) | |
359 print '\tETag:\t\t%s' % obj.etag.strip('"\'') | |
360 print '\tACL:\t\t%s' % (suri.get_acl(False, self.headers)) | |
361 return (1, obj.size) | |
362 except boto.exception.GSResponseError as e: | |
363 if e.status == 403: | |
364 print ('\tACL:\t\tACCESS DENIED. Note: you need FULL_CONTROL ' | |
365 'permission\n\t\t\ton the object to read its ACL.') | |
366 return (1, obj.size) | |
367 else: | |
368 raise e | |
369 else: | |
370 raise Exception('Unexpected ListingStyle(%s)' % listing_style) | |
371 | |
372 def _ExpandUriAndPrintInfo(self, uri, listing_style, should_recurse=False): | |
373 """ | |
374 Expands wildcards and directories/buckets for uri as needed, and | |
375 calls _PrintInfoAboutBucketListingRef() on each. | |
376 | |
377 Args: | |
378 uri: StorageUri being listed. | |
379 listing_style: ListingStyle enum describing type of output desired. | |
380 should_recurse: bool indicator of whether to expand recursively. | |
381 | |
382 Returns: | |
383 Tuple (number of matching objects, number of bytes across these objects). | |
384 """ | |
385 # We do a two-level loop, with the outer loop iterating level-by-level from | |
386 # blrs_to_expand, and the inner loop iterating the matches at the current | |
387 # level, printing them, and adding any new subdirs that need expanding to | |
388 # blrs_to_expand (to be picked up in the next outer loop iteration). | |
389 blrs_to_expand = [BucketListingRef(uri)] | |
390 num_objs = 0 | |
391 num_bytes = 0 | |
392 expanding_top_level = True | |
393 printed_one = False | |
394 num_expanded_blrs = 0 | |
395 while len(blrs_to_expand): | |
396 if printed_one: | |
397 print | |
398 blr = blrs_to_expand.pop(0) | |
399 if blr.HasKey(): | |
400 blr_iterator = iter([blr]) | |
401 elif blr.HasPrefix(): | |
402 # Bucket subdir from a previous iteration. Print "header" line only if | |
403 # we're listing more than one subdir (or if it's a recursive listing), | |
404 # to be consistent with the way UNIX ls works. | |
405 if num_expanded_blrs > 1 or should_recurse: | |
406 print '%s:' % blr.GetUriString().encode('utf-8') | |
407 printed_one = True | |
408 blr_iterator = self.WildcardIterator('%s/*' % | |
409 blr.GetRStrippedUriString(), | |
410 all_versions=self.all_versions) | |
411 elif blr.NamesBucket(): | |
412 blr_iterator = self.WildcardIterator('%s*' % blr.GetUriString(), | |
413 all_versions=self.all_versions) | |
414 else: | |
415 # This BLR didn't come from a bucket listing. This case happens for | |
416 # BLR's instantiated from a user-provided URI. | |
417 blr_iterator = PluralityCheckableIterator( | |
418 _UriOnlyBlrExpansionIterator( | |
419 self, blr, all_versions=self.all_versions)) | |
420 if blr_iterator.is_empty() and not ContainsWildcard(uri): | |
421 raise CommandException('No such object %s' % uri) | |
422 for cur_blr in blr_iterator: | |
423 num_expanded_blrs = num_expanded_blrs + 1 | |
424 if cur_blr.HasKey(): | |
425 # Object listing. | |
426 (no, nb) = self._PrintInfoAboutBucketListingRef( | |
427 cur_blr, listing_style) | |
428 num_objs += no | |
429 num_bytes += nb | |
430 printed_one = True | |
431 else: | |
432 # Subdir listing. If we're at the top level of a bucket subdir | |
433 # listing don't print the list here (corresponding to how UNIX ls | |
434 # dir just prints its contents, not the name followed by its | |
435 # contents). | |
436 if (expanding_top_level and not uri.names_bucket()) or should_recurse: | |
437 if cur_blr.GetUriString().endswith('//'): | |
438 # Expand gs://bucket// into gs://bucket//* so we don't infinite | |
439 # loop. This case happens when user has uploaded an object whose | |
440 # name begins with a /. | |
441 cur_blr = BucketListingRef(self.suri_builder.StorageUri( | |
442 '%s*' % cur_blr.GetUriString()), None, None, cur_blr.headers) | |
443 blrs_to_expand.append(cur_blr) | |
444 # Don't include the subdir name in the output if we're doing a | |
445 # recursive listing, as it will be printed as 'subdir:' when we get | |
446 # to the prefix expansion, the next iteration of the main loop. | |
447 else: | |
448 if listing_style == ListingStyle.LONG: | |
449 print '%-33s%s' % ( | |
450 '', cur_blr.GetUriString().encode('utf-8')) | |
451 else: | |
452 print cur_blr.GetUriString().encode('utf-8') | |
453 expanding_top_level = False | |
454 return (num_objs, num_bytes) | |
455 | |
456 # Command entry point. | |
457 def RunCommand(self): | |
458 got_nomatch_errors = False | |
459 listing_style = ListingStyle.SHORT | |
460 get_bucket_info = False | |
461 self.recursion_requested = False | |
462 self.all_versions = False | |
463 if self.sub_opts: | |
464 for o, a in self.sub_opts: | |
465 if o == '-a': | |
466 self.all_versions = True | |
467 elif o == '-b': | |
468 get_bucket_info = True | |
469 elif o == '-l': | |
470 listing_style = ListingStyle.LONG | |
471 elif o == '-L': | |
472 listing_style = ListingStyle.LONG_LONG | |
473 elif o == '-p': | |
474 self.proj_id_handler.SetProjectId(a) | |
475 elif o == '-r' or o == '-R': | |
476 self.recursion_requested = True | |
477 | |
478 if not self.args: | |
479 # default to listing all gs buckets | |
480 self.args = ['gs://'] | |
481 | |
482 total_objs = 0 | |
483 total_bytes = 0 | |
484 for uri_str in self.args: | |
485 uri = self.suri_builder.StorageUri(uri_str) | |
486 self.proj_id_handler.FillInProjectHeaderIfNeeded('ls', uri, self.headers) | |
487 | |
488 if uri.names_provider(): | |
489 # Provider URI: use bucket wildcard to list buckets. | |
490 for uri in self.WildcardIterator('%s://*' % uri.scheme).IterUris(): | |
491 (bucket_objs, bucket_bytes) = self._PrintBucketInfo(uri, | |
492 listing_style) | |
493 total_bytes += bucket_bytes | |
494 total_objs += bucket_objs | |
495 elif uri.names_bucket(): | |
496 # Bucket URI -> list the object(s) in that bucket. | |
497 if get_bucket_info: | |
498 # ls -b bucket listing request: List info about bucket(s). | |
499 for uri in self.WildcardIterator(uri).IterUris(): | |
500 (bucket_objs, bucket_bytes) = self._PrintBucketInfo(uri, | |
501 listing_style) | |
502 total_bytes += bucket_bytes | |
503 total_objs += bucket_objs | |
504 else: | |
505 # Not -b request: List objects in the bucket(s). | |
506 (no, nb) = self._ExpandUriAndPrintInfo(uri, listing_style, | |
507 should_recurse=self.recursion_requested) | |
508 if no == 0 and ContainsWildcard(uri): | |
509 got_nomatch_errors = True | |
510 total_objs += no | |
511 total_bytes += nb | |
512 else: | |
513 # URI names an object or object subdir -> list matching object(s) / | |
514 # subdirs. | |
515 (exp_objs, exp_bytes) = self._ExpandUriAndPrintInfo(uri, listing_style, | |
516 should_recurse=self.recursion_requested) | |
517 if exp_objs == 0 and ContainsWildcard(uri): | |
518 got_nomatch_errors = True | |
519 total_bytes += exp_bytes | |
520 total_objs += exp_objs | |
521 | |
522 if total_objs and listing_style != ListingStyle.SHORT: | |
523 print ('TOTAL: %d objects, %d bytes (%s)' % | |
524 (total_objs, total_bytes, MakeHumanReadable(float(total_bytes)))) | |
525 if got_nomatch_errors: | |
526 raise CommandException('One or more URIs matched no objects.') | |
527 | |
528 return 0 | |
529 | |
530 | |
531 class _UriOnlyBlrExpansionIterator: | |
532 """ | |
533 Iterator that expands a BucketListingRef that contains only a URI (i.e., | |
534 didn't come from a bucket listing), yielding BucketListingRefs to which it | |
535 expands. This case happens for BLR's instantiated from a user-provided URI. | |
536 | |
537 Note that we can't use NameExpansionIterator here because it produces an | |
538 iteration over the full object names (e.g., expanding "gs://bucket" to | |
539 "gs://bucket/dir/o1" and "gs://bucket/dir/o2"), while for the ls command | |
540 we need also to see the intermediate directories (like "gs://bucket/dir"). | |
541 """ | |
542 def __init__(self, command_instance, blr, all_versions=False): | |
543 self.command_instance = command_instance | |
544 self.blr = blr | |
545 self.all_versions=all_versions | |
546 | |
547 def __iter__(self): | |
548 """ | |
549 Args: | |
550 command_instance: calling instance of Command class. | |
551 blr: BucketListingRef to expand. | |
552 | |
553 Yields: | |
554 List of BucketListingRef to which it expands. | |
555 """ | |
556 # Do a delimited wildcard expansion so we get any matches along with | |
557 # whether they are keys or prefixes. That way if bucket contains a key | |
558 # 'abcd' and another key 'abce/x.txt' the expansion will return two BLRs, | |
559 # the first with HasKey()=True and the second with HasPrefix()=True. | |
560 rstripped_versionless_uri_str = self.blr.GetRStrippedUriString() | |
561 if ContainsWildcard(rstripped_versionless_uri_str): | |
562 for blr in self.command_instance.WildcardIterator( | |
563 rstripped_versionless_uri_str, all_versions=self.all_versions): | |
564 yield blr | |
565 return | |
566 # Build a wildcard to expand so CloudWildcardIterator will not just treat it | |
567 # as a key and yield the result without doing a bucket listing. | |
568 for blr in self.command_instance.WildcardIterator( | |
569 rstripped_versionless_uri_str + '*', all_versions=self.all_versions): | |
570 # Find the originally specified BucketListingRef in the expanded list (if | |
571 # present). Don't just use the expanded list, because it would also | |
572 # include objects whose name prefix matches the blr name (because of the | |
573 # wildcard match we did above). Note that there can be multiple matches, | |
574 # for the case where there's both an object and a subdirectory with the | |
575 # same name. | |
576 if (blr.GetRStrippedUriString() | |
577 == rstripped_versionless_uri_str): | |
578 yield blr | |
OLD | NEW |