X-Git-Url: https://git.arvados.org/arvados.git/blobdiff_plain/f69e114cf6dbe51525e98ca30a52ef3dd4341167..fcf3ca5baf89bdd944e3a7dcdc1b65f8ff4945ca:/sdk/python/arvados/commands/run.py diff --git a/sdk/python/arvados/commands/run.py b/sdk/python/arvados/commands/run.py index 2f3e0427d9..831e496a29 100644 --- a/sdk/python/arvados/commands/run.py +++ b/sdk/python/arvados/commands/run.py @@ -1,7 +1,12 @@ -#!/usr/bin/env python +# Copyright (C) The Arvados Authors. All rights reserved. +# +# SPDX-License-Identifier: Apache-2.0 from __future__ import print_function from __future__ import absolute_import +from builtins import range +from past.builtins import basestring +from builtins import object import arvados import arvados.commands.ws as ws import argparse @@ -107,7 +112,7 @@ def determine_project(root, current_user): # ArvFile() (file already exists in a collection), UploadFile() (file needs to # be uploaded to a collection), or simply returns prefix+fn (which yields the # original parameter string). -def statfile(prefix, fn, fnPattern="$(file %s/%s)", dirPattern="$(dir %s/%s/)"): +def statfile(prefix, fn, fnPattern="$(file %s/%s)", dirPattern="$(dir %s/%s/)", raiseOSError=False): absfn = os.path.abspath(fn) try: st = os.stat(absfn) @@ -124,7 +129,7 @@ def statfile(prefix, fn, fnPattern="$(file %s/%s)", dirPattern="$(dir %s/%s/)"): # trim leading '/' for path prefix test later return UploadFile(prefix, absfn[1:]) except OSError as e: - if e.errno == errno.ENOENT: + if e.errno == errno.ENOENT and not raiseOSError: pass else: raise @@ -143,43 +148,46 @@ def write_file(collection, pathprefix, fn): def uploadfiles(files, api, dry_run=False, num_retries=0, project=None, fnPattern="$(file %s/%s)", - name=None): + name=None, + collection=None): # Find the smallest path prefix that includes all the files that need to be uploaded. # This starts at the root and iteratively removes common parent directory prefixes # until all file paths no longer have a common parent. - n = True - pathprefix = "/" - while n: - pathstep = None - for c in files: - if pathstep is None: - sp = c.fn.split('/') - if len(sp) < 2: - # no parent directories left - n = False - break - # path step takes next directory - pathstep = sp[0] + "/" - else: - # check if pathstep is common prefix for all files - if not c.fn.startswith(pathstep): - n = False - break - if n: - # pathstep is common parent directory for all files, so remove the prefix - # from each path - pathprefix += pathstep + if files: + n = True + pathprefix = "/" + while n: + pathstep = None for c in files: - c.fn = c.fn[len(pathstep):] - - logger.info("Upload local files: \"%s\"", '" "'.join([c.fn for c in files])) + if pathstep is None: + sp = c.fn.split('/') + if len(sp) < 2: + # no parent directories left + n = False + break + # path step takes next directory + pathstep = sp[0] + "/" + else: + # check if pathstep is common prefix for all files + if not c.fn.startswith(pathstep): + n = False + break + if n: + # pathstep is common parent directory for all files, so remove the prefix + # from each path + pathprefix += pathstep + for c in files: + c.fn = c.fn[len(pathstep):] + + logger.info("Upload local files: \"%s\"", '" "'.join([c.fn for c in files])) if dry_run: logger.info("$(input) is %s", pathprefix.rstrip('/')) pdh = "$(input)" else: files = sorted(files, key=lambda x: x.fn) - collection = arvados.collection.Collection(api_client=api, num_retries=num_retries) + if collection is None: + collection = arvados.collection.Collection(api_client=api, num_retries=num_retries) prev = "" for f in files: localpath = os.path.join(pathprefix, f.fn) @@ -198,8 +206,9 @@ def uploadfiles(files, api, dry_run=False, num_retries=0, for src in iterfiles: write_file(collection, pathprefix, os.path.join(root, src)) - filters=[["portable_data_hash", "=", collection.portable_data_hash()], - ["name", "like", name+"%"]] + filters=[["portable_data_hash", "=", collection.portable_data_hash()]] + if name: + filters.append(["name", "like", name+"%"]) if project: filters.append(["owner_uuid", "=", project]) @@ -209,7 +218,7 @@ def uploadfiles(files, api, dry_run=False, num_retries=0, item = exists["items"][0] pdh = item["portable_data_hash"] logger.info("Using collection %s (%s)", pdh, item["uuid"]) - else: + elif len(collection) > 0: collection.save_new(name=name, owner_uuid=project, ensure_unique_name=True) pdh = collection.portable_data_hash() logger.info("Uploaded to %s (%s)", pdh, collection.manifest_locator()) @@ -301,7 +310,7 @@ def main(arguments=None): if files: uploadfiles(files, api, dry_run=args.dry_run, num_retries=args.retries, project=project) - for i in xrange(1, len(slots)): + for i in range(1, len(slots)): slots[i] = [("%s%s" % (c.prefix, c.fn)) if isinstance(c, ArvFile) else c for c in slots[i]] component = { @@ -321,8 +330,8 @@ def main(arguments=None): group_parser.add_argument('-b', '--batch-size', type=int) group_parser.add_argument('args', nargs=argparse.REMAINDER) - for s in xrange(2, len(slots)): - for i in xrange(0, len(slots[s])): + for s in range(2, len(slots)): + for i in range(0, len(slots[s])): if slots[s][i] == '--': inp = "input%i" % (s-2) groupargs = group_parser.parse_args(slots[2][i+1:])