X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/c203e53c2929c2ddf1b079ec077364f8f4d23c40..78700dbcc4f5d34f0a4cfee5c040e716d684ed62:/sdk/python/arvados/commands/put.py?ds=sidebyside diff --git a/sdk/python/arvados/commands/put.py b/sdk/python/arvados/commands/put.py index 00994872f3..95ba172801 100644 --- a/sdk/python/arvados/commands/put.py +++ b/sdk/python/arvados/commands/put.py @@ -3,7 +3,6 @@ # TODO: # --md5sum - display md5 of each file as read from disk -import apiclient.errors import argparse import arvados import base64 @@ -18,6 +17,7 @@ import signal import socket import sys import tempfile +from apiclient import errors as apiclient_errors import arvados.commands._util as arv_cmd @@ -27,11 +27,13 @@ api_client = None upload_opts = argparse.ArgumentParser(add_help=False) upload_opts.add_argument('paths', metavar='path', type=str, nargs='*', - help=""" + help=""" Local file or directory. Default: read from standard input. """) -upload_opts.add_argument('--max-manifest-depth', type=int, metavar='N', +_group = upload_opts.add_mutually_exclusive_group() + +_group.add_argument('--max-manifest-depth', type=int, metavar='N', default=-1, help=""" Maximum depth of directory tree to represent in the manifest structure. A directory structure deeper than this will be represented @@ -40,56 +42,62 @@ a single stream. Default: -1 (unlimited), i.e., exactly one manifest stream per filesystem directory that contains files. """) +_group.add_argument('--normalize', action='store_true', + help=""" +Normalize the manifest by re-ordering files and streams after writing +data. +""") + _group = upload_opts.add_mutually_exclusive_group() _group.add_argument('--as-stream', action='store_true', dest='stream', - help=""" + help=""" Synonym for --stream. """) _group.add_argument('--stream', action='store_true', - help=""" + help=""" Store the file content and display the resulting manifest on stdout. Do not write the manifest to Keep or save a Collection object in Arvados. """) _group.add_argument('--as-manifest', action='store_true', dest='manifest', - help=""" + help=""" Synonym for --manifest. """) _group.add_argument('--in-manifest', action='store_true', dest='manifest', - help=""" + help=""" Synonym for --manifest. """) _group.add_argument('--manifest', action='store_true', - help=""" + help=""" Store the file data and resulting manifest in Keep, save a Collection object in Arvados, and display the manifest locator (Collection uuid) on stdout. This is the default behavior. """) _group.add_argument('--as-raw', action='store_true', dest='raw', - help=""" + help=""" Synonym for --raw. """) _group.add_argument('--raw', action='store_true', - help=""" + help=""" Store the file content and display the data block locators on stdout, separated by commas, with a trailing newline. Do not store a manifest. """) upload_opts.add_argument('--use-filename', type=str, default=None, - dest='filename', help=""" + dest='filename', help=""" Synonym for --filename. """) upload_opts.add_argument('--filename', type=str, default=None, - help=""" + help=""" Use the given filename in the manifest, instead of the name of the local file. This is useful when "-" or "/dev/stdin" is given as an input file. It can be used only if there is exactly one path given and @@ -97,7 +105,7 @@ it is not a directory. Implies --manifest. """) upload_opts.add_argument('--portable-data-hash', action='store_true', - help=""" + help=""" Print the portable data hash instead of the Arvados UUID for the collection created by the upload. """) @@ -115,31 +123,31 @@ Save the collection with the specified name. _group = run_opts.add_mutually_exclusive_group() _group.add_argument('--progress', action='store_true', - help=""" + help=""" Display human-readable progress on stderr (bytes and, if possible, percentage of total data size). This is the default behavior when stderr is a tty. """) _group.add_argument('--no-progress', action='store_true', - help=""" + help=""" Do not display human-readable progress on stderr, even if stderr is a tty. """) _group.add_argument('--batch-progress', action='store_true', - help=""" + help=""" Display machine-readable progress on stderr (bytes and, if known, total data size). """) _group = run_opts.add_mutually_exclusive_group() _group.add_argument('--resume', action='store_true', default=True, - help=""" + help=""" Continue interrupted uploads from cached state (default). """) _group.add_argument('--no-resume', action='store_false', dest='resume', - help=""" + help=""" Do not continue interrupted uploads from cached state. """) @@ -287,12 +295,12 @@ class ArvPutCollectionWriter(arvados.ResumableCollectionWriter): def flush_data(self): start_buffer_len = self._data_buffer_len - start_block_count = self.bytes_written / self.KEEP_BLOCK_SIZE + start_block_count = self.bytes_written / arvados.config.KEEP_BLOCK_SIZE super(ArvPutCollectionWriter, self).flush_data() if self._data_buffer_len < start_buffer_len: # We actually PUT data. self.bytes_written += (start_buffer_len - self._data_buffer_len) self.report_progress() - if (self.bytes_written / self.KEEP_BLOCK_SIZE) > start_block_count: + if (self.bytes_written / arvados.config.KEEP_BLOCK_SIZE) > start_block_count: self.cache_state() def _record_new_input(self, input_type, source_name, dest_name): @@ -364,11 +372,11 @@ def desired_project_uuid(api_client, project_uuid, num_retries): def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr): global api_client - if api_client is None: - api_client = arvados.api('v1') - status = 0 args = parse_arguments(arguments) + status = 0 + if api_client is None: + api_client = arvados.api('v1') # Determine the name to use if args.name: @@ -390,7 +398,7 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr): try: project_uuid = desired_project_uuid(api_client, args.project_uuid, args.retries) - except (apiclient.errors.Error, ValueError) as error: + except (apiclient_errors.Error, ValueError) as error: print >>stderr, error sys.exit(1) @@ -447,16 +455,21 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr): if args.stream: output = writer.manifest_text() + if args.normalize: + output = CollectionReader(output).manifest_text(normalize=True) elif args.raw: output = ','.join(writer.data_locators()) else: try: + manifest_text = writer.manifest_text() + if args.normalize: + manifest_text = CollectionReader(manifest_text).manifest_text(normalize=True) # Register the resulting collection in Arvados. collection = api_client.collections().create( body={ 'owner_uuid': project_uuid, 'name': collection_name, - 'manifest_text': writer.manifest_text() + 'manifest_text': manifest_text }, ensure_unique_name=True ).execute(num_retries=args.retries) @@ -468,7 +481,7 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr): else: output = collection['uuid'] - except apiclient.errors.Error as error: + except apiclient_errors.Error as error: print >>stderr, ( "arv-put: Error creating Collection on project: {}.".format( error))