def test_raw_datastream_version(self):
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META = {}
self.obj.text.content = 'second version content'
self.obj.text.save()
# retrieve the view for each version and compare
for version in self.obj.text.history().versions:
# get the datastream version to compare with the response
dsversion = self.obj.getDatastreamObject(self.obj.text.id,
as_of_date=version.created)
response = raw_datastream(rqst, self.obj.pid, self.obj.text.id,
as_of_date=version.created)
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream as of %s' \
% (expected, got, version.created))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream as of %s' \
% (expected, got, version.created))
# should use version-specific checksum and size
self.assertEqual(dsversion.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response')
self.assertEqual(dsversion.size, int(response['Content-Length']))
# should retrieve appropriate version of the content
self.assertEqual(dsversion.content, response.content)
def file(request, pid):
dsid = FileObject.file.id
repo = Repository()
obj = repo.get_object(pid, type=FileObject)
filename = os.path.basename(obj.dc.content.title)
extra_headers = {
'Content-Disposition': "attachment; filename=%s" % filename,
}
return raw_datastream(request, pid, dsid, type=FileObject, headers=extra_headers)
def rdfxml(request, aggId):
dsid = AggregationObject.rdfxml.id
repo = Repository()
obj = repo.get_object(aggId, type=AggregationObject)
filename = os.path.basename(obj.dc.content.title)
extra_headers = {
'Content-Disposition': "attachment; filename=%s" % filename,
}
return raw_datastream(request, aggId, dsid, type=AggregationObject, headers=extra_headers)
def download(request, pid):
'Download disk image datastream contents'
repo = Repository(request=request)
obj = repo.get_object(pid, type=DiskImage)
extra_headers = {
'Content-Disposition': "attachment; filename=%s.%s" % \
(obj.noid, obj.provenance.content.object.latest_format.name)
}
return raw_datastream(request, pid, DiskImage.content.id,
repo=repo, headers=extra_headers)
def view_datastream(request, pid, dsid):
'Access raw object datastreams'
# initialize local repo with logged-in user credentials & call generic view
# use type-inferring repo to pick up rushdie file or generic arrangement
response = raw_datastream(request, pid, dsid,
repo=TypeInferringRepository(request=request))
# work-around for email MIME data : display as plain text so it
# can be viewed in the browser
if response['Content-Type'] == 'message/rfc822':
response['Content-Type'] = 'text/plain'
return response
def view_datastream(request, pid, dsid):
'Access raw object datastreams'
# initialize local repo with logged-in user credentials & call generic view
# use type-inferring repo to pick up rushdie file or generic arrangement
response = raw_datastream(request, pid, dsid,
repo=TypeInferringRepository(request=request))
# work-around for email MIME data : display as plain text so it
# can be viewed in the browser
if response.get('Content-Type') == 'message/rfc822':
response['Content-Type'] = 'text/plain'
return response
def download_file(request, pid):
'''Download the master file datastream associated with a
:class:`~genrepo.file.models.FileObject`'''
repo = Repository(request=request)
obj = init_by_cmodel(pid, request)
# use original or edited filename as download filename
extra_headers = {
'Content-Disposition': "attachment; filename=%s" % obj.master.label
}
# use generic raw datastream view from eulcore
# - use the datastream id and digital object type returned by cmodel init
return raw_datastream(request, pid, obj.master.id, type=obj.__class__,
repo=repo, headers=extra_headers)
def download(request, pid):
'Download disk image datastream contents'
repo = Repository(request=request)
obj = repo.get_object(pid, type=DiskImage)
extra_headers = {
'Content-Disposition': "attachment; filename=%s.%s" % \
(obj.noid, obj.provenance.content.object.latest_format.name)
}
return raw_datastream(request,
pid,
DiskImage.content.id,
repo=repo,
headers=extra_headers)
def download_file(request, pid):
'''Download the master file datastream associated with a
:class:`~genrepo.file.models.FileObject`'''
repo = Repository(request=request)
obj = init_by_cmodel(pid, request)
# use original or edited filename as download filename
extra_headers = {
'Content-Disposition': "attachment; filename=%s" % obj.master.label
}
# use generic raw datastream view from eulcore
# - use the datastream id and digital object type returned by cmodel init
return raw_datastream(request,
pid,
obj.master.id,
type=obj.__class__,
repo=repo,
headers=extra_headers)
def test_raw_datastream_version(self):
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META = {}
self.obj.text.content = 'second version content'
self.obj.text.save()
# retrieve the view for each version and compare
for version in self.obj.text.history().versions:
# get the datastream version to compare with the response
dsversion = self.obj.getDatastreamObject(
self.obj.text.id, as_of_date=version.created)
response = raw_datastream(rqst,
self.obj.pid,
self.obj.text.id,
as_of_date=version.created)
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream as of %s' \
% (expected, got, version.created))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream as of %s' \
% (expected, got, version.created))
# should use version-specific checksum and size
self.assertEqual(
dsversion.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response'
)
self.assertEqual(dsversion.size, int(response['Content-Length']))
# should retrieve appropriate version of the content
self.assertEqual(dsversion.content, response.content)
def test_raw_datastream(self):
# tests for new version of raw_datastream introduced in 1.5,
# based on old raw_datastream tests
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META = {}
# DC
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
content = b''.join(c for c in response.streaming_content)
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of DC' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of DC' \
% (expected, got))
self.assertEqual(
self.obj.dc.checksum, response['ETag'],
'datastream checksum should be set as ETag header in the response')
self.assertEqual(self.obj.dc.checksum, response['Content-MD5'])
self.assert_('<dc:title>%s</dc:title>' %
self.obj.dc.content.title in force_text(content))
# RELS-EXT
response = raw_datastream(rqst, self.obj.pid, 'RELS-EXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of RELS-EXT' \
% (expected, got))
expected, got = 'application/rdf+xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of RELS-EXT' \
% (expected, got))
# TEXT (non-xml content)
response = raw_datastream(rqst, self.obj.pid, 'TEXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of TEXT' \
% (expected, got))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of TEXT' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(
self.obj.text.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response'
)
self.assertEqual(len(self.obj.text.content),
int(response['Content-Length']))
# IMAGE (binary content)
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of IMAGE' \
% (expected, got))
expected, got = 'image/png', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of IMAGE' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(
self.obj.image.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response'
)
self.assertTrue(
response.has_header('Content-Length'),
'content-length header should be set in the response for binary datastreams'
)
self.assert_(isinstance(response, StreamingHttpResponse))
# non-existent datastream should 404
self.assertRaises(Http404, raw_datastream, rqst, self.obj.pid,
'BOGUS-DSID')
# non-existent record should 404
self.assertRaises(Http404, raw_datastream, rqst, 'bogus-pid:1', 'DC')
# set extra headers in the response
extra_headers = {'Content-Disposition': 'attachment; filename=foo.txt'}
response = raw_datastream_old(rqst,
self.obj.pid,
'TEXT',
headers=extra_headers)
self.assertTrue(response.has_header('Content-Disposition'))
self.assertEqual(response['Content-Disposition'],
extra_headers['Content-Disposition'])
# explicitly support GET and HEAD requests only
rqst.method = 'POST'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 405, response.status_code
self.assertEqual(expected, got,
'Expected %s (Method not Allowed) but returned %s for POST to raw_datastream view' \
% (expected, got))
# test HEAD request
rqst.method = 'HEAD'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for HEAD request on raw_datastream view' \
% (expected, got))
self.assert_(isinstance(response, HttpResponse))
self.assertEqual(b'', response.content)
# test that range requests are passed through to fedora
# use IMAGE for testing since it is binary content
# set range header in the request; bytes=0- : entire datastream
rqst.META['HTTP_RANGE'] = 'bytes=0-'
rqst.method = 'GET'
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content = b''.join(c for c in response.streaming_content)
self.assertEqual(self.obj.image.size, len(content),
'range request of bytes=0- should return entire content (expected %d, got %d)' \
% (self.obj.image.size, len(content)))
self.assertEqual(self.obj.image.size, int(response['Content-Length']),
'content-length header should be size of entire content (expected %d, got %d)' \
% (self.obj.image.size, int(response['Content-Length'])))
expected = 'bytes 0-%d/%d' % (self.obj.image.size - 1,
self.obj.image.size)
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
del response
# set range request for partial beginning content; bytes=0-150
bytes_requested = 'bytes=0-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 151
content = b''.join(c for c in response.streaming_content)
self.assertEqual(content_len, len(content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 0-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# complex ranges not yet supported
bytes_requested = 'bytes=1-10,30-50'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream_old(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream_old invalid range request %s' \
% (expected, got, bytes_requested))
def test_raw_datastream(self):
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META = {}
# rqst.META.get.return_value = None
# DC
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
content = response.content
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of DC' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of DC' \
% (expected, got))
self.assertEqual(self.obj.dc.checksum, response['ETag'],
'datastream checksum should be set as ETag header in the response')
self.assertEqual(self.obj.dc.checksum, response['Content-MD5'])
self.assert_('<dc:title>%s</dc:title>' % self.obj.dc.content.title in content)
# RELS-EXT
response = raw_datastream(rqst, self.obj.pid, 'RELS-EXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of RELS-EXT' \
% (expected, got))
expected, got = 'application/rdf+xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of RELS-EXT' \
% (expected, got))
# TEXT (non-xml content)
response = raw_datastream(rqst, self.obj.pid, 'TEXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of TEXT' \
% (expected, got))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of TEXT' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(self.obj.text.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response')
self.assertEqual(len(self.obj.text.content), int(response['Content-Length']))
# IMAGE (binary content)
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of IMAGE' \
% (expected, got))
expected, got = 'image/png', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of IMAGE' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(self.obj.image.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response')
self.assertTrue(response.has_header('Content-Length'),
'content-length header should be set in the response for binary datastreams')
# non-existent datastream should 404
self.assertRaises(Http404, raw_datastream, rqst, self.obj.pid, 'BOGUS-DSID')
# non-existent record should 404
self.assertRaises(Http404, raw_datastream, rqst, 'bogus-pid:1', 'DC')
# check type handling?
# set extra headers in the response
extra_headers = {'Content-Disposition': 'attachment; filename=foo.txt'}
response = raw_datastream(rqst, self.obj.pid, 'TEXT',
headers=extra_headers)
self.assertTrue(response.has_header('Content-Disposition'))
self.assertEqual(response['Content-Disposition'], extra_headers['Content-Disposition'])
# explicitly support GET and HEAD requests only
rqst.method = 'POST'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 405, response.status_code
self.assertEqual(expected, got,
'Expected %s (Method not Allowed) but returned %s for POST to raw_datastream view' \
% (expected, got))
# HEAD request is handled internally, for efficiency
rqst.method = 'HEAD'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for HEAD request on raw_datastream view' \
% (expected, got))
self.assertEqual('', response.content)
def test_raw_datastream_range(self):
# test http range requests
rqst = Mock()
rqst.method = 'GET'
rqst.META = {}
# use IMAGE for testing since it is binary content
# set range header in the request; bytes=0- : entire datastream
rqst.META['HTTP_RANGE'] = 'bytes=0-'
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content = response.content
self.assertEqual(self.obj.image.size, len(content),
'range request of bytes=0- should return entire content (expected %d, got %d)' \
% (self.obj.image.size, len(content)))
self.assertEqual(self.obj.image.size, int(response['Content-Length']),
'content-length header should be size of entire content (expected %d, got %d)' \
% (self.obj.image.size, int(response['Content-Length'])))
expected = 'bytes 0-%d/%d' % (self.obj.image.size - 1, self.obj.image.size)
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
del response
# set range request for partial beginning content; bytes=0-150
bytes_requested = 'bytes=0-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 150
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 0-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# set range request for partial middle content; bytes=10-150
bytes_requested = 'bytes=10-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 150 - 10
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 10-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# set range request for partial end content; bytes=2000-3118
bytes_requested = 'bytes=2000-3118'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 3118 - 2000
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 2000-3118/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# invalid or unsupported ranges should return 416, range not satisfiable
bytes_requested = 'bytes=10-9' # start > end
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream invalid range request %s' \
% (expected, got, bytes_requested))
# complex ranges not yet supported
bytes_requested = 'bytes=1-10,30-50'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream invalid range request %s' \
% (expected, got, bytes_requested))
def test_raw_datastream_range(self):
# test http range requests
rqst = Mock()
rqst.method = 'GET'
rqst.META = {}
# use IMAGE for testing since it is binary content
# set range header in the request; bytes=0- : entire datastream
rqst.META['HTTP_RANGE'] = 'bytes=0-'
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content = response.content
self.assertEqual(self.obj.image.size, len(content),
'range request of bytes=0- should return entire content (expected %d, got %d)' \
% (self.obj.image.size, len(content)))
self.assertEqual(self.obj.image.size, int(response['Content-Length']),
'content-length header should be size of entire content (expected %d, got %d)' \
% (self.obj.image.size, int(response['Content-Length'])))
expected = 'bytes 0-%d/%d' % (self.obj.image.size - 1,
self.obj.image.size)
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
del response
# set range request for partial beginning content; bytes=0-150
bytes_requested = 'bytes=0-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 150
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 0-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# set range request for partial middle content; bytes=10-150
bytes_requested = 'bytes=10-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 150 - 10
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 10-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# set range request for partial end content; bytes=2000-3118
bytes_requested = 'bytes=2000-3118'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 3118 - 2000
self.assertEqual(content_len, len(response.content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(response.content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 2000-3118/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# invalid or unsupported ranges should return 416, range not satisfiable
bytes_requested = 'bytes=10-9' # start > end
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream invalid range request %s' \
% (expected, got, bytes_requested))
# complex ranges not yet supported
bytes_requested = 'bytes=1-10,30-50'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst,
self.obj.pid,
'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream invalid range request %s' \
% (expected, got, bytes_requested))
def download_audio(request, pid, type, extension=None):
'''Serve out an audio datastream for the fedora object specified by pid.
Can be used to download original (WAV) audio file or the access copy (MP3).
:param pid: pid of the :class:`~keep.audio.models.AudioObject` instance
from which the audio datastream should be returned
:param type: which audio datastream to return - should be one of 'original'
or 'access'
:param extension: optional filename extension for access copy to
distinguish between different types of access copies (currently MP3 or M4A)
The :class:`django.http.HttpResponse` returned will have a Content-Disposition
set to prompt the user to download the file with a filename based on the
object noid and an appropriate file extension for the type of audio requested.
'''
repo = Repository(request=request)
# retrieve the object so we can use it to set the download filename
obj = repo.get_object(pid, type=AudioObject)
# user needs either *play* or *download* permissions
# - could be any audio or researcher-accessible only, which additionally
# requires checking object is researcher-accessible
# for now, use presence of 'HTTP_RANGE' in request to differentiate
# jplayer requests from straight downloads
# NOTE: this would not be too difficult for a savvy user to circumvent
# (if they know what we are checking), but is intended mainly to prevent
# unwanted access by staff and researchers in the reading room
# if http range is present in request, check for play permissions
# (also requires that request is for access copy, not original)
if 'HTTP_RANGE' in request.META:
if not (request.user.has_perm('audio.play_audio') and type == 'access') and \
not (request.user.has_perm('audio.play_researcher_audio') and \
bool(obj.researcher_access) and type == 'access'):
return prompt_login_or_403(request)
# otherwise, check for download permissions
else:
# user either needs download audio permissions OR
# if they can download researcher audio and object must be researcher-accessible
if not request.user.has_perm('audio.download_audio') and \
not (request.user.has_perm('audio.download_researcher_audio') and \
bool(obj.researcher_access)):
return prompt_login_or_403(request)
# determine which datastream is requsted & set datastream id & file extension
if type == 'original':
dsid = AudioObject.audio.id
file_ext = 'wav'
elif type == 'access':
dsid = AudioObject.compressed_audio.id
# make sure the requested file extension matches the datastream
if (obj.compressed_audio.mimetype == 'audio/mp4' and \
extension != 'm4a') or \
(obj.compressed_audio.mimetype == 'audio/mpeg' and \
extension != 'mp3'):
raise Http404
file_ext = extension
else:
# any other type is not supported
raise Http404
extra_headers = {
'Content-Disposition': 'attachment; filename="%s.%s"' % (obj.noid, file_ext)
}
# use generic raw datastream view from eulfedora
return raw_datastream(request, pid, dsid, repo=repo, headers=extra_headers)
def test_raw_datastream(self):
# tests for new version of raw_datastream introduced in 1.5,
# based on old raw_datastream tests
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META = {}
# DC
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
content = b''.join(c for c in response.streaming_content)
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of DC' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of DC' \
% (expected, got))
self.assertEqual(self.obj.dc.checksum, response['ETag'],
'datastream checksum should be set as ETag header in the response')
self.assertEqual(self.obj.dc.checksum, response['Content-MD5'])
self.assert_('<dc:title>%s</dc:title>' % self.obj.dc.content.title in force_text(content))
# RELS-EXT
response = raw_datastream(rqst, self.obj.pid, 'RELS-EXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of RELS-EXT' \
% (expected, got))
expected, got = 'application/rdf+xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of RELS-EXT' \
% (expected, got))
# TEXT (non-xml content)
response = raw_datastream(rqst, self.obj.pid, 'TEXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of TEXT' \
% (expected, got))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of TEXT' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(self.obj.text.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response')
self.assertEqual(len(self.obj.text.content), int(response['Content-Length']))
# IMAGE (binary content)
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of IMAGE' \
% (expected, got))
expected, got = 'image/png', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of IMAGE' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(self.obj.image.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response')
self.assertTrue(response.has_header('Content-Length'),
'content-length header should be set in the response for binary datastreams')
self.assert_(isinstance(response, StreamingHttpResponse))
# non-existent datastream should 404
self.assertRaises(Http404, raw_datastream, rqst, self.obj.pid, 'BOGUS-DSID')
# non-existent record should 404
self.assertRaises(Http404, raw_datastream, rqst, 'bogus-pid:1', 'DC')
# set extra headers in the response
extra_headers = {'Content-Disposition': 'attachment; filename=foo.txt'}
response = raw_datastream_old(rqst, self.obj.pid, 'TEXT',
headers=extra_headers)
self.assertTrue(response.has_header('Content-Disposition'))
self.assertEqual(response['Content-Disposition'], extra_headers['Content-Disposition'])
# explicitly support GET and HEAD requests only
rqst.method = 'POST'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 405, response.status_code
self.assertEqual(expected, got,
'Expected %s (Method not Allowed) but returned %s for POST to raw_datastream view' \
% (expected, got))
# test HEAD request
rqst.method = 'HEAD'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for HEAD request on raw_datastream view' \
% (expected, got))
self.assert_(isinstance(response, HttpResponse))
self.assertEqual(b'', response.content)
# test that range requests are passed through to fedora
# use IMAGE for testing since it is binary content
# set range header in the request; bytes=0- : entire datastream
rqst.META['HTTP_RANGE'] = 'bytes=0-'
rqst.method = 'GET'
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content = b''.join(c for c in response.streaming_content)
self.assertEqual(self.obj.image.size, len(content),
'range request of bytes=0- should return entire content (expected %d, got %d)' \
% (self.obj.image.size, len(content)))
self.assertEqual(self.obj.image.size, int(response['Content-Length']),
'content-length header should be size of entire content (expected %d, got %d)' \
% (self.obj.image.size, int(response['Content-Length'])))
expected = 'bytes 0-%d/%d' % (self.obj.image.size - 1, self.obj.image.size)
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
del response
# set range request for partial beginning content; bytes=0-150
bytes_requested = 'bytes=0-150'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 206, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream range request' \
% (expected, got))
content_len = 151
content = b''.join(c for c in response.streaming_content)
self.assertEqual(content_len, len(content),
'range request of %s should return %d bytes, got %d' \
% (bytes_requested, content_len, len(content)))
self.assertEqual(content_len, int(response['Content-Length']),
'content-length header should be set to partial size %d (got %d)' \
% (content_len, int(response['Content-Length'])))
expected = 'bytes 0-150/%d' % self.obj.image.size
self.assertEqual(expected, response['Content-Range'],
'content range response header should indicate bytes returned (expected %s, got %s)' \
% (expected, response['Content-Range']))
# complex ranges not yet supported
bytes_requested = 'bytes=1-10,30-50'
rqst.META['HTTP_RANGE'] = bytes_requested
response = raw_datastream_old(rqst, self.obj.pid, 'IMAGE',
accept_range_request=True)
expected, got = 416, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream_old invalid range request %s' \
% (expected, got, bytes_requested))
def download_video(request, pid, type, extension=None):
'''Serve out an video datastream for the fedora object specified by pid.
Can be used to download original file or the access copy.
:param pid: pid of the :class:`~keep.vidoe.models.Video` instance
from which the vidoe datastream should be returned
:param type: which video datastream to return - should be one of 'original'
or 'access'
:param extension: optional filename extension for access copy to
distinguish between different types of access copies
The :class:`django.http.HttpResponse` returned will have a Content-Disposition
set to prompt the user to download the file with a filename based on the
object noid and an appropriate file extension for the type of video requested.
'''
repo = Repository(request=request)
# retrieve the object so we can use it to set the download filename
obj = repo.get_object(pid, type=Video)
# user needs either *play* or *download* permissions
# - could be any video or researcher-accessible only, which additionally
# requires checking object is researcher-accessible
# for now, use presence of 'HTTP_RANGE' in request to differentiate
# jplayer requests from straight downloads
# NOTE: this would not be too difficult for a savvy user to circumvent
# (if they know what we are checking), but is intended mainly to prevent
# unwanted access by staff and researchers in the reading room
# if http range is present in request, check for play permissions
# (also requires that request is for access copy, not original)
if 'HTTP_RANGE' in request.META:
playable = (type == 'access' and
(request.user.has_perm('video.play_video')) or
(request.user.has_perm('video.play_researcher_video') and
bool(obj.researcher_access)))
if not playable:
return prompt_login_or_403(request)
# otherwise, check for download permissions
else:
# user either needs download video permissions OR
# if they can download researcher audio and object must be researcher-accessible
downloadable = request.user.has_perm('video.download_video') or \
(request.user.has_perm('video.download_researcher_video') and
bool(obj.researcher_access))
if not downloadable:
return prompt_login_or_403(request)
# determine which datastream is requsted & set datastream id & file extension
if type == 'original':
dsid = Video.content.id
# set file extension based on the datastream content type,
# with a fallback for generic binary (should not happen in production)
file_ext = Video.allowed_master_mimetypes.get(obj.content.mimetype, 'bin')
elif type == 'access':
dsid = Video.access_copy.id
# set file extension based on the datastream content
file_ext = Video.allowed_access_mimetypes[obj.access_copy.mimetype]
else:
# any other type is not supported
raise Http404
extra_headers = {
'Content-Disposition': 'attachment; filename="%s.%s"' % (obj.noid, file_ext)
}
# use generic raw datastream view from eulfedora
return raw_datastream(request, pid, dsid, repo=repo,
headers=extra_headers)
# errors accessing Fedora will fall through to default 500 error handling
def view_datastream(request, pid, dsid):
'Access raw object datastreams (MODS, RELS-EXT, DC, DigitalTech, SourceTech, JHOVE)'
# initialize local repo with logged-in user credentials & call generic view
return raw_datastream(request, pid, dsid, repo=Repository(request=request))
def download_audio(request, pid, type, extension=None):
'''Serve out an audio datastream for the fedora object specified by pid.
Can be used to download original (WAV) audio file or the access copy (MP3).
:param pid: pid of the :class:`~keep.audio.models.AudioObject` instance
from which the audio datastream should be returned
:param type: which audio datastream to return - should be one of 'original'
or 'access'
:param extension: optional filename extension for access copy to
distinguish between different types of access copies (currently MP3 or M4A)
The :class:`django.http.HttpResponse` returned will have a Content-Disposition
set to prompt the user to download the file with a filename based on the
object noid and an appropriate file extension for the type of audio requested.
'''
repo = Repository(request=request)
# retrieve the object so we can use it to set the download filename
obj = repo.get_object(pid, type=AudioObject)
# user needs either *play* or *download* permissions
# - could be any audio or researcher-accessible only, which additionally
# requires checking object is researcher-accessible
# for now, use presence of 'HTTP_RANGE' in request to differentiate
# jplayer requests from straight downloads
# NOTE: this would not be too difficult for a savvy user to circumvent
# (if they know what we are checking), but is intended mainly to prevent
# unwanted access by staff and researchers in the reading room
# if http range is present in request, check for play permissions
# (also requires that request is for access copy, not original)
if 'HTTP_RANGE' in request.META:
if not (request.user.has_perm('audio.play_audio') and type == 'access') and \
not (request.user.has_perm('audio.play_researcher_audio') and \
bool(obj.researcher_access) and type == 'access'):
return prompt_login_or_403(request)
# otherwise, check for download permissions
else:
# user either needs download audio permissions OR
# if they can download researcher audio and object must be researcher-accessible
if not request.user.has_perm('audio.download_audio') and \
not (request.user.has_perm('audio.download_researcher_audio') and \
bool(obj.researcher_access)):
return prompt_login_or_403(request)
# determine which datastream is requsted & set datastream id & file extension
if type == 'original':
dsid = AudioObject.audio.id
file_ext = 'wav'
elif type == 'access':
dsid = AudioObject.compressed_audio.id
# make sure the requested file extension matches the datastream
if (obj.compressed_audio.mimetype == 'audio/mp4' and \
extension != 'm4a') or \
(obj.compressed_audio.mimetype == 'audio/mpeg' and \
extension != 'mp3'):
raise Http404
file_ext = extension
else:
# any other type is not supported
raise Http404
extra_headers = {
'Content-Disposition':
'attachment; filename="%s.%s"' % (obj.noid, file_ext)
}
# use generic raw datastream view from eulfedora
return raw_datastream(request, pid, dsid, repo=repo, headers=extra_headers)
def download_video(request, pid, type, extension=None):
'''Serve out an video datastream for the fedora object specified by pid.
Can be used to download original file or the access copy.
:param pid: pid of the :class:`~keep.vidoe.models.Video` instance
from which the vidoe datastream should be returned
:param type: which video datastream to return - should be one of 'original'
or 'access'
:param extension: optional filename extension for access copy to
distinguish between different types of access copies
The :class:`django.http.HttpResponse` returned will have a Content-Disposition
set to prompt the user to download the file with a filename based on the
object noid and an appropriate file extension for the type of video requested.
'''
repo = Repository(request=request)
# retrieve the object so we can use it to set the download filename
obj = repo.get_object(pid, type=Video)
# user needs either *play* or *download* permissions
# - could be any video or researcher-accessible only, which additionally
# requires checking object is researcher-accessible
# for now, use presence of 'HTTP_RANGE' in request to differentiate
# jplayer requests from straight downloads
# NOTE: this would not be too difficult for a savvy user to circumvent
# (if they know what we are checking), but is intended mainly to prevent
# unwanted access by staff and researchers in the reading room
# if http range is present in request, check for play permissions
# (also requires that request is for access copy, not original)
if 'HTTP_RANGE' in request.META:
playable = (type == 'access' and
(request.user.has_perm('video.play_video'))
or (request.user.has_perm('video.play_researcher_video')
and bool(obj.researcher_access)))
if not playable:
return prompt_login_or_403(request)
# otherwise, check for download permissions
else:
# user either needs download video permissions OR
# if they can download researcher audio and object must be researcher-accessible
downloadable = request.user.has_perm('video.download_video') or \
(request.user.has_perm('video.download_researcher_video') and
bool(obj.researcher_access))
if not downloadable:
return prompt_login_or_403(request)
# determine which datastream is requsted & set datastream id & file extension
if type == 'original':
dsid = Video.content.id
# set file extension based on the datastream content type,
# with a fallback for generic binary (should not happen in production)
file_ext = Video.allowed_master_mimetypes.get(obj.content.mimetype,
'bin')
elif type == 'access':
dsid = Video.access_copy.id
# set file extension based on the datastream content
file_ext = Video.allowed_access_mimetypes[obj.access_copy.mimetype]
else:
# any other type is not supported
raise Http404
extra_headers = {
'Content-Disposition':
'attachment; filename="%s.%s"' % (obj.noid, file_ext)
}
# use generic raw datastream view from eulfedora
return raw_datastream(request, pid, dsid, repo=repo, headers=extra_headers)
def view_datastream(request, pid, dsid):
'Access raw object datastreams (MODS, RELS-EXT, DC)'
# initialize local repo with logged-in user credentials & call generic view
return raw_datastream(request, pid, dsid, repo=Repository(request=request))
def test_raw_datastream(self):
rqst = Mock()
rqst.method = 'GET'
# return empty headers for ETag condition check
rqst.META.get.return_value = None
# DC
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of DC' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of DC' \
% (expected, got))
self.assertEqual(
self.obj.dc.checksum, response['ETag'],
'datastream checksum should be set as ETag header in the response')
self.assertEqual(self.obj.dc.checksum, response['Content-MD5'])
self.assert_('<dc:title>%s</dc:title>' %
self.obj.dc.content.title in response.content)
# RELS-EXT
response = raw_datastream(rqst, self.obj.pid, 'RELS-EXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of RELS-EXT' \
% (expected, got))
expected, got = 'application/rdf+xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of RELS-EXT' \
% (expected, got))
# TEXT (non-xml content)
response = raw_datastream(rqst, self.obj.pid, 'TEXT')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of TEXT' \
% (expected, got))
expected, got = 'text/plain', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of TEXT' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(
self.obj.text.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response'
)
self.assertEqual(len(self.obj.text.content),
int(response['Content-Length']))
# IMAGE (binary content)
response = raw_datastream(rqst, self.obj.pid, 'IMAGE')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_datastream view of IMAGE' \
% (expected, got))
expected, got = 'image/png', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_datastream view of IMAGE' \
% (expected, got))
# non-xml datastreams should have content-md5 & content-length headers
self.assertEqual(
self.obj.image.checksum, response['Content-MD5'],
'datastream checksum should be set as Content-MD5 header in the response'
)
self.assertTrue(
response.has_header('Content-Length'),
'content-length header should be set in the response for binary datastreams'
)
# non-existent datastream should 404
self.assertRaises(Http404, raw_datastream, rqst, self.obj.pid,
'BOGUS-DSID')
# non-existent record should 404
self.assertRaises(Http404, raw_datastream, rqst, 'bogus-pid:1', 'DC')
# check type handling?
# set extra headers in the response
extra_headers = {'Content-Disposition': 'attachment; filename=foo.txt'}
response = raw_datastream(rqst,
self.obj.pid,
'TEXT',
headers=extra_headers)
self.assertTrue(response.has_header('Content-Disposition'))
self.assertEqual(response['Content-Disposition'],
extra_headers['Content-Disposition'])
# explicitly support GET and HEAD requests only
rqst.method = 'POST'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 405, response.status_code
self.assertEqual(expected, got,
'Expected %s (Method not Allowed) but returned %s for POST to raw_datastream view' \
% (expected, got))
# HEAD request is handled internally, for efficiency
rqst.method = 'HEAD'
response = raw_datastream(rqst, self.obj.pid, 'DC')
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for HEAD request on raw_datastream view' \
% (expected, got))
self.assertEqual('', response.content)