Merge branch '8784-dir-listings'

[arvados.git] / sdk / python / arvados / commands / get.py

diff --git a/sdk/python/arvados/commands/get.py b/sdk/python/arvados/commands/get.py
index c7254da1c970dee41ae8eee45023a366e8598d66..56d471888d65a8f200bb6ae7af39991286be8aab 100755 (executable)
--- a/sdk/python/arvados/commands/get.py
+++ b/sdk/python/arvados/commands/get.py
@@ -1,4 +1,7 @@
 #!/usr/bin/env python
+# Copyright (C) The Arvados Authors. All rights reserved.
+#
+# SPDX-License-Identifier: Apache-2.0
 
 import argparse
 import hashlib
@@ -10,6 +13,7 @@ import logging
 
 import arvados
 import arvados.commands._util as arv_cmd
+import arvados.util as util
 
 from arvados._version import __version__
 
@@ -84,6 +88,11 @@ write *anything* if any files exist that would have to be
 overwritten. This option causes even devices, sockets, and fifos to be
 skipped.
 """)
+group.add_argument('--strip-manifest', action='store_true', default=False,
+                   help="""
+When getting a collection manifest, strip its access tokens before writing
+it.
+""")
 
 def parse_arguments(arguments, stdout, stderr):
     args = parser.parse_args(arguments)
@@ -125,18 +134,22 @@ def parse_arguments(arguments, stdout, stderr):
 
 def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr):
     global api_client
-    
+
+    if stdout is sys.stdout and hasattr(stdout, 'buffer'):
+        # in Python 3, write to stdout as binary
+        stdout = stdout.buffer
+
     args = parse_arguments(arguments, stdout, stderr)
     if api_client is None:
         api_client = arvados.api('v1')
 
     r = re.search(r'^(.*?)(/.*)?$', args.locator)
-    collection = r.group(1)
+    col_loc = r.group(1)
     get_prefix = r.group(2)
     if args.r and not get_prefix:
         get_prefix = os.sep
     try:
-        reader = arvados.CollectionReader(collection, num_retries=args.retries)
+        reader = arvados.CollectionReader(col_loc, num_retries=args.retries)
     except Exception as error:
         logger.error("failed to read collection: {}".format(error))
         return 1
@@ -149,16 +162,16 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr):
                 open_flags |= os.O_EXCL
             try:
                 if args.destination == "-":
-                    stdout.write(reader.manifest_text())
+                    stdout.write(reader.manifest_text(strip=args.strip_manifest).encode())
                 else:
                     out_fd = os.open(args.destination, open_flags)
                     with os.fdopen(out_fd, 'wb') as out_file:
-                        out_file.write(reader.manifest_text())
+                        out_file.write(reader.manifest_text(strip=args.strip_manifest).encode())
             except (IOError, OSError) as error:
                 logger.error("can't write to '{}': {}".format(args.destination, error))
                 return 1
             except (arvados.errors.ApiError, arvados.errors.KeepReadError) as error:
-                logger.error("failed to download '{}': {}".format(collection, error))
+                logger.error("failed to download '{}': {}".format(col_loc, error))
                 return 1
         return 0
 
@@ -230,7 +243,7 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr):
         if args.hash:
             digestor = hashlib.new(args.hash)
         try:
-            with s.open(f.name, 'r') as file_reader:
+            with s.open(f.name, 'rb') as file_reader:
                 for data in file_reader.readall():
                     if outfile:
                         outfile.write(data)
@@ -261,10 +274,11 @@ def main(arguments=None, stdout=sys.stdout, stderr=sys.stderr):
 
     if args.progress:
         stderr.write('\n')
+    return 0
 
 def files_in_collection(c):
     # Sort first by file type, then alphabetically by file path.
-    for i in sorted(c.keys(),
+    for i in sorted(list(c.keys()),
                     key=lambda k: (
                         isinstance(c[k], arvados.collection.Subcollection),
                         k.upper())):