Expose --metadata-directive for s3 commands

CHANGELOG.rst

@@ -2,6 +2,14 @@
 CHANGELOG
 =========
 
+Next Release (TBD)
+==================
+
+* feature:``aws s3``: Add support for ``--metadata-directive`` that allows
+  metadata to be copied or replaced for single part copies.
+  (`issue 1188 <https://github.com/aws/aws-cli/pull/1188>`__)
+
+
 1.7.13
 ======
 

awscli/customizations/s3/fileinfo.py

@@ -243,6 +243,11 @@ def _handle_object_params(self, params):
         if self.parameters['expires']:
             params['expires'] = self.parameters['expires'][0]
 
+    def _handle_metadata_directive(self, params):
+        if self.parameters['metadata_directive']:
+            params['metadata_directive'] = \
+                self.parameters['metadata_directive'][0]
+
     def upload(self, payload=None):
         """
         Redirects the file to the multipart upload function if the file is
@@ -294,6 +299,7 @@ def copy(self):
         params = {'endpoint': self.endpoint, 'bucket': bucket,
                   'copy_source': copy_source, 'key': key}
         self._handle_object_params(params)
+        self._handle_metadata_directive(params)
         response_data, http = operate(self.service, 'CopyObject', params)
 
     def delete(self):

awscli/customizations/s3/s3handler.py

@@ -63,7 +63,7 @@ def __init__(self, session, params, result_queue=None,
                        'content_language': None, 'expires': None,
                        'grants': None, 'only_show_errors': False,
                        'is_stream': False, 'paths_type': None,
-                       'expected_size': None}
+                       'expected_size': None, 'metadata_directive': None}
         self.params['region'] = params['region']
         for key in self.params.keys():
             if key in params:

awscli/customizations/s3/subcommands.py

@@ -216,6 +216,28 @@
 }
 
 
+METADATA_DIRECTIVE = {
+    'name': 'metadata-directive', 'nargs': 1, 'choices': ['COPY', 'REPLACE'],
+    'help_text': (
+        'Specifies whether the metadata is copied from the source object '
+        'or replaced with metadata provided when copying S3 objects. '
+        'Note that if the object is copied over in parts, the source '
+        'object\'s metadata will not be copied over, no matter the value for '
+        '``--metadata-directive``, and instead the desired metadata values '
+        'must be specified as parameters on the command line. '
+        'Valid values are ``COPY`` and ``REPLACE``. If this parameter is not '
+        'specified, ``COPY`` will be used by default. If ``REPLACE`` is used, '
+        'the copied object will only have the metadata values that were'
+        ' specified by the CLI command. Note that if you are '
+        'using any of the following parameters: ``--content-type``, '
+        '``content-language``, ``--content-encoding``, '
+        '``--content-disposition``, ``-cache-control``, or ``--expires``, you '
+        'will need to specify ``--metadata-directive REPLACE`` for '
+        'non-multipart copies if you want the copied objects to have the '
+        'specified metadata values.')
+}
+
+
 INDEX_DOCUMENT = {'name': 'index-document',
                   'help_text': (
                       'A suffix that is appended to a request that is for '
@@ -530,7 +552,7 @@ class CpCommand(S3TransferCommand):
             "or <S3Path> <S3Path>"
     ARG_TABLE = [{'name': 'paths', 'nargs': 2, 'positional_arg': True,
                   'synopsis': USAGE}] + TRANSFER_ARGS + \
-                [EXPECTED_SIZE, RECURSIVE]
+                [METADATA_DIRECTIVE, EXPECTED_SIZE, RECURSIVE]
     EXAMPLES = BasicCommand.FROM_FILE('s3/cp.rst')
 
 
@@ -541,7 +563,8 @@ class MvCommand(S3TransferCommand):
     USAGE = "<LocalPath> <S3Path> or <S3Path> <LocalPath> " \
             "or <S3Path> <S3Path>"
     ARG_TABLE = [{'name': 'paths', 'nargs': 2, 'positional_arg': True,
-                  'synopsis': USAGE}] + TRANSFER_ARGS + [RECURSIVE]
+                  'synopsis': USAGE}] + TRANSFER_ARGS + [METADATA_DIRECTIVE,
+                                                         RECURSIVE]
     EXAMPLES = BasicCommand.FROM_FILE('s3/mv.rst')
 
 

tests/integration/customizations/s3/test_plugin.py

@@ -26,6 +26,7 @@
 import socket
 import tempfile
 import shutil
+import copy
 
 import botocore.session
 from botocore.exceptions import ClientError
@@ -35,6 +36,7 @@
 from awscli.testutils import aws as _aws
 from tests.unit.customizations.s3 import create_bucket as _create_bucket
 from awscli.customizations.s3.transferconfig import DEFAULTS
+from awscli.customizations.scalarparse import add_scalar_parsers
 
 
 @contextlib.contextmanager
@@ -523,6 +525,55 @@ def test_copy_large_file_signature_v4(self):
         self.assert_no_errors(p)
         self.assertTrue(self.key_exists(bucket_name, key_name='foo.txt'))
 
+    def test_copy_metadata_directive(self):
+        # Copy the same style of parsing as the CLI session. This is needed
+        # For comparing expires timestamp.
+        add_scalar_parsers(self.session)
+        bucket_name = self.create_bucket()
+        original_key = 'foo.txt'
+        new_key = 'bar.txt'
+        metadata = {
+            'ContentType': 'foo',
+            'ContentDisposition': 'foo',
+            'ContentEncoding': 'foo',
+            'ContentLanguage': 'foo',
+            'CacheControl': '90',
+            'Expires': '0'
+        }
+        self.put_object(bucket_name, original_key, contents='foo',
+                        extra_args=metadata)
+        p = aws('s3 cp s3://%s/%s s3://%s/%s' %
+                (bucket_name, original_key, bucket_name, new_key))
+        self.assert_no_errors(p)
+        response = self.head_object(bucket_name, new_key)
+        # These values should have the metadata of the source object
+        metadata_ref = copy.copy(metadata)
+        metadata_ref['Expires'] = 'Thu, 01 Jan 1970 00:00:00 GMT'
+        for name, value in metadata_ref.items():
+            self.assertEqual(response[name], value)
+
+        # Use REPLACE to wipe out all of the metadata.
+        p = aws('s3 cp s3://%s/%s s3://%s/%s --metadata-directive REPLACE' %
+                (bucket_name, original_key, bucket_name, new_key))
+        self.assert_no_errors(p)
+        response = self.head_object(bucket_name, new_key)
+        # Make sure all of the original metadata is gone.
+        for name, value in metadata_ref.items():
+            self.assertNotEqual(response.get(name), value)
+
+        # Use REPLACE to wipe out all of the metadata but include a new
+        # metadata value.
+        p = aws('s3 cp s3://%s/%s s3://%s/%s --metadata-directive REPLACE '
+                '--content-type bar' %
+                (bucket_name, original_key, bucket_name, new_key))
+        self.assert_no_errors(p)
+        response = self.head_object(bucket_name, new_key)
+        # Make sure the content type metadata is included
+        self.assertEqual(response['ContentType'], 'bar')
+        # Make sure all of the original metadata is gone.
+        for name, value in metadata_ref.items():
+            self.assertNotEqual(response.get(name), value)
+
 
 class TestSync(BaseS3CLICommand):
     def test_sync_with_plus_chars_paginate(self):

tests/unit/customizations/s3/test_cp_command.py

@@ -101,6 +101,33 @@ def test_website_redirect_ignore_paramfile(self):
             'http://someserver'
         )
 
+    def test_metadata_directive_copy(self):
+        self.parsed_responses = [
+            {"ContentLength": "100", "LastModified": "00:00:00Z"},
+            {'ETag': '"foo-1"'},
+        ]
+        cmdline = ('%s s3://bucket/key.txt s3://bucket/key2.txt'
+                   ' --metadata-directive REPLACE' % self.prefix)
+        self.run_cmd(cmdline, expected_rc=0)
+        self.assertEqual(len(self.operations_called), 2,
+                         self.operations_called)
+        self.assertEqual(self.operations_called[0][0].name, 'HeadObject')
+        self.assertEqual(self.operations_called[1][0].name, 'CopyObject')
+        self.assertEqual(self.operations_called[1][1]['metadata_directive'],
+                         'REPLACE')
+
+    def test_no_metadata_directive_for_non_copy(self):
+        full_path = self.files.create_file('foo.txt', 'mycontent')
+        cmdline = '%s %s s3://bucket --metadata-directive REPLACE' % \
+            (self.prefix, full_path)
+        self.parsed_responses = \
+            [{'ETag': '"c8afdb36c52cf4727836669019e69222"'}]
+        self.run_cmd(cmdline, expected_rc=0)
+        self.assertEqual(len(self.operations_called), 1,
+                         self.operations_called)
+        self.assertEqual(self.operations_called[0][0].name, 'PutObject')
+        self.assertNotIn('metadata_directive', self.operations_called[0][1])
+
 
 if __name__ == "__main__":
     unittest.main()

tests/unit/customizations/s3/test_mv_command.py

@@ -55,6 +55,34 @@ def test_website_redirect_ignore_paramfile(self):
             'http://someserver'
         )
 
+    def test_metadata_directive_copy(self):
+        self.parsed_responses = [
+            {"ContentLength": "100", "LastModified": "00:00:00Z"},
+            {'ETag': '"foo-1"'},
+            {'ETag': '"foo-2"'}
+        ]
+        cmdline = ('%s s3://bucket/key.txt s3://bucket/key2.txt'
+                   ' --metadata-directive REPLACE' % self.prefix)
+        self.run_cmd(cmdline, expected_rc=0)
+        self.assertEqual(len(self.operations_called), 3,
+                         self.operations_called)
+        self.assertEqual(self.operations_called[0][0].name, 'HeadObject')
+        self.assertEqual(self.operations_called[1][0].name, 'CopyObject')
+        self.assertEqual(self.operations_called[2][0].name, 'DeleteObject')
+        self.assertEqual(self.operations_called[1][1]['metadata_directive'],
+                         'REPLACE')
+
+    def test_no_metadata_directive_for_non_copy(self):
+        full_path = self.files.create_file('foo.txt', 'mycontent')
+        cmdline = '%s %s s3://bucket --metadata-directive REPLACE' % \
+            (self.prefix, full_path)
+        self.parsed_responses = \
+            [{'ETag': '"c8afdb36c52cf4727836669019e69222"'}]
+        self.run_cmd(cmdline, expected_rc=0)
+        self.assertEqual(len(self.operations_called), 1,
+                         self.operations_called)
+        self.assertEqual(self.operations_called[0][0].name, 'PutObject')
+        self.assertNotIn('metadata_directive', self.operations_called[0][1])
 
 if __name__ == "__main__":
     unittest.main()

tests/unit/test_completer.py

@@ -78,7 +78,8 @@
                              '--content-disposition', '--source-region',
                              '--content-encoding', '--content-language',
                              '--expires', '--grants', '--only-show-errors',
-                             '--expected-size', '--page-size']
+                             '--expected-size', '--page-size',
+                             '--metadata-directive']
                             + GLOBALOPTS)),
     ('aws s3 cp --quiet -', -1, set(['--no-guess-mime-type', '--dryrun',
                                      '--recursive', '--content-type',
@@ -88,7 +89,7 @@
                                      '--expires', '--website-redirect', '--acl',
                                      '--storage-class', '--sse',
                                      '--exclude', '--include',
-                                     '--source-region',
+                                     '--source-region', '--metadata-directive',
                                      '--grants', '--only-show-errors',
                                      '--expected-size', '--page-size']
                                     + GLOBALOPTS)),