async def download_folder_as_zip(self):
zipfile_name = self.path.name or '{}-archive'.format(self.provider.NAME)
self.set_header('Content-Type', 'application/zip')
self.set_header('Content-Disposition', make_disposition(zipfile_name + '.zip'))
result = await self.provider.zip(self.path)
await self.write_stream(result)
async def download_folder_as_zip(self):
zipfile_name = self.path.name or '{}-archive'.format(
self.provider.NAME)
self.set_header('Content-Type', 'application/zip')
self.set_header('Content-Disposition',
make_disposition(zipfile_name + '.zip'))
result = await self.provider.zip(self.path)
await self.write_stream(result)
async def download_file(self):
if 'Range' not in self.request.headers:
request_range = None
else:
logger.debug('Range header is: {}'.format(self.request.headers['Range']))
request_range = utils.parse_request_range(self.request.headers['Range'])
logger.debug('Range header parsed as: {}'.format(request_range))
version = self.requested_version
stream = await self.provider.download(
self.path,
revision=version,
range=request_range,
accept_url='direct' not in self.request.query_arguments,
mode=self.get_query_argument('mode', default=None),
display_name=self.get_query_argument('displayName', default=None),
)
if isinstance(stream, str):
return self.redirect(stream)
if getattr(stream, 'partial', None):
# Use getattr here as not all stream may have a partial attribute
# Plus it fixes tests
self.set_status(206)
self.set_header('Content-Range', stream.content_range)
if stream.content_type is not None:
self.set_header('Content-Type', stream.content_type)
logger.debug('stream size is: {}'.format(stream.size))
if stream.size is not None:
self.set_header('Content-Length', str(stream.size))
# Build `Content-Disposition` header from `displayName` override,
# headers of provider response, or file path, whichever is truthy first
name = (
self.get_query_argument('displayName', default=None) or
getattr(stream, 'name', None) or
self.path.name
)
self.set_header('Content-Disposition', make_disposition(name))
_, ext = os.path.splitext(name)
# If the file extention is in mime_types
# override the content type to fix issues with safari shoving in new file extensions
if ext in mime_types:
self.set_header('Content-Type', mime_types[ext])
await self.write_stream(stream)
if getattr(stream, 'partial', False) and isinstance(stream, ResponseStreamReader):
await stream.response.release()
logger.debug('bytes received is: {}'.format(self.bytes_downloaded))
async def download_file(self):
if 'Range' not in self.request.headers:
request_range = None
else:
logger.debug('Range header is: {}'.format(
self.request.headers['Range']))
request_range = utils.parse_request_range(
self.request.headers['Range'])
logger.debug('Range header parsed as: {}'.format(request_range))
version = self.requested_version
stream = await self.provider.download(
self.path,
revision=version,
range=request_range,
accept_url='direct' not in self.request.query_arguments,
mode=self.get_query_argument('mode', default=None),
display_name=self.get_query_argument('displayName', default=None),
)
if isinstance(stream, str):
return self.redirect(stream)
if getattr(stream, 'partial', None):
# Use getattr here as not all stream may have a partial attribute
# Plus it fixes tests
self.set_status(206)
self.set_header('Content-Range', stream.content_range)
if stream.content_type is not None:
self.set_header('Content-Type', stream.content_type)
logger.debug('stream size is: {}'.format(stream.size))
if stream.size is not None:
self.set_header('Content-Length', str(stream.size))
# Build `Content-Disposition` header from `displayName` override,
# headers of provider response, or file path, whichever is truthy first
name = (self.get_query_argument('displayName', default=None)
or getattr(stream, 'name', None) or self.path.name)
self.set_header('Content-Disposition', make_disposition(name))
_, ext = os.path.splitext(name)
# If the file extention is in mime_types
# override the content type to fix issues with safari shoving in new file extensions
if ext in mime_types:
self.set_header('Content-Type', mime_types[ext])
await self.write_stream(stream)
if getattr(stream, 'partial', False) and isinstance(
stream, ResponseStreamReader):
await stream.response.release()
logger.debug('bytes received is: {}'.format(self.bytes_downloaded))
async def download(
self,
path: WaterButlerPath,
accept_url=False,
range=None, # type: ignore
**kwargs) -> typing.Union[str, ResponseStreamReader]:
"""Download the object with the given path.
API Docs:
GET Object: https://cloud.google.com/storage/docs/xml-api/get-object
Download an Object: https://cloud.google.com/storage/docs/xml-api/get-object-download
The behavior of download differs depending on the value of ``accept_url``. If
``accept_url == False``, WB makes a standard signed request and returns a
``ResponseStreamReader``. If ``accept_url == True``, WB builds and signs the ``GET``
request with an extra query parameter ``response-content-disposition`` to trigger the
download with the display name. The signed URL is returned.
:param path: the WaterButlerPath for the object to download
:type path: :class:`.WaterButlerPath`
:param bool accept_url: should return a direct time-limited download url from the provider
:param tuple range: the Range HTTP request header
:param dict kwargs: ``display_name`` - the display name of the file on OSF and for download
:rtype: str or :class:`.streams.ResponseStreamReader`
"""
if path.is_folder:
raise DownloadError('Cannot download folders',
code=HTTPStatus.BAD_REQUEST)
req_method = 'GET'
obj_name = utils.get_obj_name(path, is_folder=False)
if accept_url:
display_name = kwargs.get('display_name') or path.name
query = {
'response-content-disposition': make_disposition(display_name)
}
# There is no need to delay URL building and signing
signed_url = self._build_and_sign_url(req_method, obj_name,
**query) # type: ignore
return signed_url
signed_url = functools.partial(self._build_and_sign_url, req_method,
obj_name, **{})
resp = await self.make_request(req_method,
signed_url,
range=range,
expects=(HTTPStatus.OK,
HTTPStatus.PARTIAL_CONTENT),
throws=DownloadError)
return ResponseStreamReader(resp)
async def download(self,
path,
accept_url=False,
revision=None,
range=None,
**kwargs):
"""Returns a ResponseWrapper (Stream) for the specified path
raises FileNotFoundError if the status from S3 is not 200
:param str path: Path to the key you want to download
:param dict \*\*kwargs: Additional arguments that are ignored
:rtype: :class:`waterbutler.core.streams.ResponseStreamReader`
:raises: :class:`waterbutler.core.exceptions.DownloadError`
"""
await self._check_region()
if not path.is_file:
raise exceptions.DownloadError('No file specified for download',
code=400)
if not revision or revision.lower() == 'latest':
query_parameters = None
else:
query_parameters = {'versionId': revision}
display_name = kwargs.get('display_name') or path.name
response_headers = {
'response-content-disposition': make_disposition(display_name)
}
url = functools.partial(self.bucket.new_key(path.path).generate_url,
settings.TEMP_URL_SECS,
query_parameters=query_parameters,
response_headers=response_headers)
if accept_url:
return url()
resp = await self.make_request(
'GET',
url,
range=range,
expects=(
200,
206,
),
throws=exceptions.DownloadError,
)
return streams.ResponseStreamReader(resp)
async def get(self):
"""Download as a Zip archive."""
zipfile_name = self.path.name or '{}-archive'.format(self.provider.NAME)
self.set_header('Content-Type', 'application/zip')
self.set_header(
'Content-Disposition',
make_disposition(zipfile_name + '.zip')
)
result = await self.provider.zip(**self.arguments)
await self.write_stream(result)
self._send_hook('download_zip', path=self.path)
async def get(self):
"""Download a file."""
try:
self.arguments['accept_url'] = TRUTH_MAP[self.arguments.get(
'accept_url', 'true').lower()]
except KeyError:
raise tornado.web.HTTPError(status_code=400)
if 'Range' in self.request.headers:
request_range = utils.parse_request_range(
self.request.headers['Range'])
else:
request_range = None
result = await self.provider.download(range=request_range,
**self.arguments)
if isinstance(result, str):
self.redirect(result)
self._send_hook('download_file', path=self.path)
return
if getattr(result, 'partial', None):
# Use getattr here as not all stream may have a partial attribute
# Plus it fixes tests
self.set_status(206)
self.set_header('Content-Range', result.content_range)
if result.content_type is not None:
self.set_header('Content-Type', result.content_type)
if result.size is not None:
self.set_header('Content-Length', str(result.size))
# Build `Content-Disposition` header from `displayName` override,
# headers of provider response, or file path, whichever is truthy first
name = self.arguments.get('displayName') or getattr(
result, 'name', None) or self.path.name
self.set_header('Content-Disposition', make_disposition(name))
_, ext = os.path.splitext(name)
# If the file extention is in mime_types
# override the content type to fix issues with safari shoving in new file extensions
if ext in mime_types:
self.set_header('Content-Type', mime_types[ext])
await self.write_stream(result)
self._send_hook('download_file', path=self.path)
async def download(self, path: WaterButlerPath, accept_url=False, range=None, # type: ignore
**kwargs) -> typing.Union[str, ResponseStreamReader]:
"""Download the object with the given path.
API Docs:
GET Object: https://cloud.google.com/storage/docs/xml-api/get-object
Download an Object: https://cloud.google.com/storage/docs/xml-api/get-object-download
The behavior of download differs depending on the value of ``accept_url``. If
``accept_url == False``, WB makes a standard signed request and returns a
``ResponseStreamReader``. If ``accept_url == True``, WB builds and signs the ``GET``
request with an extra query parameter ``response-content-disposition`` to trigger the
download with the display name. The signed URL is returned.
:param path: the WaterButlerPath for the object to download
:type path: :class:`.WaterButlerPath`
:param bool accept_url: should return a direct time-limited download url from the provider
:param tuple range: the Range HTTP request header
:param dict kwargs: ``display_name`` - the display name of the file on OSF and for download
:rtype: str or :class:`.streams.ResponseStreamReader`
"""
if path.is_folder:
raise DownloadError('Cannot download folders', code=HTTPStatus.BAD_REQUEST)
req_method = 'GET'
obj_name = utils.get_obj_name(path, is_folder=False)
if accept_url:
display_name = kwargs.get('display_name') or path.name
query = {'response-content-disposition': make_disposition(display_name)}
# There is no need to delay URL building and signing
signed_url = self._build_and_sign_url(req_method, obj_name, **query) # type: ignore
return signed_url
signed_url = functools.partial(self._build_and_sign_url, req_method, obj_name, **{})
resp = await self.make_request(
req_method,
signed_url,
range=range,
expects=(HTTPStatus.OK, HTTPStatus.PARTIAL_CONTENT),
throws=DownloadError
)
return ResponseStreamReader(resp)
async def get(self):
"""Download a file."""
try:
self.arguments['accept_url'] = TRUTH_MAP[self.arguments.get('accept_url', 'true').lower()]
except KeyError:
raise tornado.web.HTTPError(status_code=400)
if 'Range' in self.request.headers:
request_range = utils.parse_request_range(self.request.headers['Range'])
else:
request_range = None
result = await self.provider.download(range=request_range, **self.arguments)
if isinstance(result, str):
self.redirect(result)
self._send_hook('download_file', path=self.path)
return
if getattr(result, 'partial', None):
# Use getattr here as not all stream may have a partial attribute
# Plus it fixes tests
self.set_status(206)
self.set_header('Content-Range', result.content_range)
if result.content_type is not None:
self.set_header('Content-Type', result.content_type)
if result.size is not None:
self.set_header('Content-Length', str(result.size))
# Build `Content-Disposition` header from `displayName` override,
# headers of provider response, or file path, whichever is truthy first
name = self.arguments.get('displayName') or getattr(result, 'name', None) or self.path.name
self.set_header('Content-Disposition', make_disposition(name))
_, ext = os.path.splitext(name)
# If the file extention is in mime_types
# override the content type to fix issues with safari shoving in new file extensions
if ext in mime_types:
self.set_header('Content-Type', mime_types[ext])
await self.write_stream(result)
self._send_hook('download_file', path=self.path)
async def download(self, path, accept_url=False, revision=None, range=None, **kwargs):
"""Returns a ResponseWrapper (Stream) for the specified path
raises FileNotFoundError if the status from S3 is not 200
:param str path: Path to the key you want to download
:param dict \*\*kwargs: Additional arguments that are ignored
:rtype: :class:`waterbutler.core.streams.ResponseStreamReader`
:raises: :class:`waterbutler.core.exceptions.DownloadError`
"""
await self._check_region()
if not path.is_file:
raise exceptions.DownloadError('No file specified for download', code=400)
if not revision or revision.lower() == 'latest':
query_parameters = None
else:
query_parameters = {'versionId': revision}
display_name = kwargs.get('display_name') or path.name
response_headers = {
'response-content-disposition': make_disposition(display_name)
}
url = functools.partial(
self.bucket.new_key(path.path).generate_url,
settings.TEMP_URL_SECS,
query_parameters=query_parameters,
response_headers=response_headers
)
if accept_url:
return url()
resp = await self.make_request(
'GET',
url,
range=range,
expects=(200, 206, ),
throws=exceptions.DownloadError,
)
return streams.ResponseStreamReader(resp)
def test_content_disposition(self, filename, expected):
disposition = utils.make_disposition(filename)
assert disposition == expected
def test_content_disposition(self, filename, expected):
disposition = utils.make_disposition(filename)
assert disposition == expected