def index_data(request, id, repo=None):
'''Return the fields and values to be indexed for a single object
as JSON. Index content is generated via
:meth:`eulfedora.models.DigitalObject.index_data`.
:param id: id of the object to be indexed; in this case a Fedora pid
'''
#Ensure permission to this resource is allowed. Currently based on IP only.
if _permission_denied_check(request):
return HttpResponseForbidden('Access to this web service was denied.',
content_type='text/html')
if repo is None:
repo_opts = {}
# if credentials are specified via Basic Auth, use them for Fedora access
auth_info = request.META.get('HTTP_AUTHORIZATION', None)
basic = 'Basic '
if auth_info and auth_info.startswith(basic):
basic_info = auth_info[len(basic):]
basic_info_decoded = codecs.decode(force_bytes(basic_info),
'base64')
u, p = force_text(basic_info_decoded).split(':')
repo_opts.update({'username': u, 'password': p})
repo = TypeInferringRepository(**repo_opts)
try:
obj = repo.get_object(id)
return HttpResponse(json.dumps(obj.index_data()),
content_type='application/json')
except RequestFailed:
# for now, treat any failure getting the object from Fedora as a 404
# (could also potentially be a permission error)
raise Http404
def index_data(request, id, repo=None):
"""Return the fields and values to be indexed for a single object
as JSON. Index content is generated via
:meth:`eulfedora.models.DigitalObject.index_data`.
:param id: id of the object to be indexed; in this case a Fedora pid
"""
# Ensure permission to this resource is allowed. Currently based on IP only.
if _permission_denied_check(request):
return HttpResponseForbidden("Access to this web service was denied.", content_type="text/html")
if repo is None:
repo_opts = {}
# if credentials are specified via Basic Auth, use them for Fedora access
auth_info = request.META.get("HTTP_AUTHORIZATION", None)
basic = "Basic "
if auth_info and auth_info.startswith(basic):
basic_info = auth_info[len(basic) :]
basic_info_decoded = codecs.decode(force_bytes(basic_info), "base64")
u, p = force_text(basic_info_decoded).split(":")
repo_opts.update({"username": u, "password": p})
repo = TypeInferringRepository(**repo_opts)
try:
obj = repo.get_object(id)
return HttpResponse(json.dumps(obj.index_data()), content_type="application/json")
except RequestFailed:
# for now, treat any failure getting the object from Fedora as a 404
# (could also potentially be a permission error)
raise Http404
def curl_upload_file(filename):
print("curl upload")
conn = pycurl.Curl()
auth = base64.b64encode(force_bytes("%s:%s" % (testsettings.FEDORA_USER, testsettings.FEDORA_PASSWORD)))
headers = {"Authorization": "Basic %s" % force_text(auth)}
conn.setopt(conn.URL, "%supload" % testsettings.FEDORA_ROOT_NONSSL)
conn.setopt(pycurl.VERBOSE, 1)
conn.setopt(pycurl.HTTPHEADER, ["%s: %s" % t for t in headers.items()])
filesize = os.path.getsize(filename)
widgets = [
"Upload: ",
progressbar.widgets.Percentage(),
" ",
progressbar.widgets.Bar(),
" ",
progressbar.widgets.ETA(),
" ",
progressbar.widgets.FileTransferSpeed(),
]
# set initial progressbar size based on file; will be slightly larger because
# of multipart boundary content
pbar = progressbar.ProgressBar(widgets=widgets, maxval=filesize).start()
def progress(dl_total, dl, up_total, up):
# update the progressbar to actual maxval (content + boundary)
pbar.max_value = up_total
# update current status
pbar.update(up)
conn.setopt(
conn.HTTPPOST,
[
(
"file",
(
# upload the contents of this file
conn.FORM_FILE,
filename,
# specify a different file name for the upload
conn.FORM_FILENAME,
"file",
# specify a different content type
# conn.FORM_CONTENTTYPE, 'application/x-python',
),
)
],
)
# conn.setopt(conn.CURLOPT_READFUNCTION)
conn.setopt(conn.XFERINFOFUNCTION, progress)
conn.setopt(conn.NOPROGRESS, False)
conn.perform()
# HTTP response code, e.g. 200.
print("Status: %d" % conn.getinfo(conn.RESPONSE_CODE))
# Elapsed time for the transfer.
print("Time: %f" % conn.getinfo(conn.TOTAL_TIME))
conn.close()
def test_ingest_utf8(self):
# ingest with unicode log message
obj = self.loadFixtureData('basic-object.foxml')
response = self.rest_api.ingest(obj, logMessage=self.unicode_test_str)
pid = response.text
self.assertTrue(pid)
response = self.rest_api.getObjectXML(pid)
response.encoding = 'utf-8' # ensure requests decodes as utf-8
self.assert_(u'<audit:justification>%s</audit:justification>' %
self.unicode_test_str in response.text)
self.rest_api.purgeObject(force_text(pid))
# ingest with unicode object label
# convert to text to replace string, then convert back to bytes
obj = force_bytes(force_text(obj).replace(
u"A test object", self.unicode_test_str))
response = self.rest_api.ingest(obj)
pid = response.text
self.assertTrue(pid)
# object label in profile should match the unicode sent
response = self.rest_api.getObjectProfile(pid)
response.encoding = 'utf-8' # ensure requests decodes as utf-8
self.assert_(u'<objLabel>%s</objLabel>' % self.unicode_test_str
in response.text)
self.rest_api.purgeObject(force_text(pid))
def index_data(request, id, repo=None):
'''Return the fields and values to be indexed for a single object
as JSON. Index content is generated via
:meth:`eulfedora.models.DigitalObject.index_data`.
:param id: id of the object to be indexed; in this case a Fedora pid
'''
# Ensure permission to this resource is allowed. Currently based on IP only.
if _permission_denied_check(request):
return HttpResponseForbidden('Access to this web service was denied.', content_type='text/html')
if repo is None:
repo_opts = {}
# if credentials are specified via Basic Auth, use them for Fedora access
auth_info = request.META.get('HTTP_AUTHORIZATION', None)
basic = 'Basic '
if auth_info and auth_info.startswith(basic):
basic_info = auth_info[len(basic):]
basic_info_decoded = base64.b64decode(force_bytes(basic_info))
# NOTE: codecs.decode works everywhere but python 3.3. which
# complains about an unknown encoding
# basic_info_decoded = codecs.decode(force_bytes(basic_info), 'base64')
u, p = force_text(basic_info_decoded).split(':')
repo_opts.update({'username': u, 'password': p})
repo = TypeInferringRepository(**repo_opts)
try:
obj = repo.get_object(id)
return HttpResponse(json.dumps(obj.index_data()),
content_type='application/json')
except RequestFailed:
# for now, treat any failure getting the object from Fedora as a 404
# (could also potentially be a permission error)
raise Http404
def test_raw_audit_trail(self):
rqst = Mock()
rqst.method = 'GET'
# created with no ingest message = no audit trail
self.assertRaises(Http404, raw_audit_trail, rqst, self.obj.pid)
# modify object so it will have an audit trail
self.obj.dc.content.title = 'audit this!'
changelog = 'I just changed the title'
self.obj.save(changelog)
response = raw_audit_trail(rqst, self.obj.pid)
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_audit_trail' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_audit_trail' \
% (expected, got))
self.assert_(b'<audit:auditTrail' in response.content)
self.assert_(
force_bytes('<audit:justification>%s</audit:justification>' %
changelog) in response.content)
self.assert_('Last-Modified' in response)
def test_ingest_utf8(self):
# ingest with unicode log message
obj = self.loadFixtureData('basic-object.foxml')
response = self.rest_api.ingest(obj, logMessage=self.unicode_test_str)
pid = response.text
self.assertTrue(pid)
response = self.rest_api.getObjectXML(pid)
response.encoding = 'utf-8' # ensure requests decodes as utf-8
self.assert_(u'<audit:justification>%s</audit:justification>' %
self.unicode_test_str in response.text)
self.rest_api.purgeObject(force_text(pid))
# ingest with unicode object label
# convert to text to replace string, then convert back to bytes
obj = force_bytes(force_text(obj).replace(
u"A test object", self.unicode_test_str))
response = self.rest_api.ingest(obj)
pid = response.text
self.assertTrue(pid)
# object label in profile should match the unicode sent
response = self.rest_api.getObjectProfile(pid)
response.encoding = 'utf-8' # ensure requests decodes as utf-8
self.assert_(u'<objLabel>%s</objLabel>' % self.unicode_test_str
in response.text)
self.rest_api.purgeObject(force_text(pid))
def upload(self, data, callback=None, content_type=None, size=None):
'''
Upload a multi-part file for content to ingest. Returns a
temporary upload id that can be used as a datstream location.
:param data: content string, file-like object, or iterable with
content to be uploaded
:param callback: optional callback method to monitor the upload;
see :mod:`requests-toolbelt` documentation for more
details: https://toolbelt.readthedocs.org/en/latest/user.html#uploading-data
:param content_type: optional content type of the data
:param size: optional size of the data; required when using an
iterable for the data
:returns: upload id on success
'''
url = 'upload'
# fedora only expects content uploaded as multipart file;
# make string content into a file-like object so requests.post
# sends it the way Fedora expects.
# NOTE: checking for both python 2.x next method and
# python 3.x __next__ to test if data is iteraable
if not hasattr(data, 'read') and \
not (hasattr(data, '__next__') or hasattr(data, 'next')):
data = six.BytesIO(force_bytes(data))
# if data is an iterable, wrap in a readable iterator that
# requests-toolbelt can read data from
elif not hasattr(data, 'read') and \
(hasattr(data, '__next__') or hasattr(data, 'next')):
if size is None:
raise Exception('Cannot upload iterable with unknown size')
data = ReadableIterator(data, size)
# use requests-toolbelt multipart encoder to avoid reading
# the full content of large files into memory
menc = MultipartEncoder(fields={'file': ('file', data, content_type)})
if callback is not None:
menc = MultipartEncoderMonitor(menc, callback)
headers = {'Content-Type': menc.content_type}
if size:
headers['Content-Length'] = size
try:
response = self.post(url, data=menc, headers=headers)
except OverflowError:
# Python __len__ uses integer so it is limited to system maxint,
# and requests and requests-toolbelt use len() throughout.
# This results in an overflow error when trying to upload a file
# larger than system maxint (2GB on 32-bit OSes).
# See http://bugs.python.org/issue12159
msg = 'upload content larger than system maxint (32-bit OS limitation)'
logger.error('OverflowError: %s', msg)
raise OverflowError(msg)
if response.status_code == requests.codes.accepted:
return response.text.strip()
def curl_upload_file(filename):
print('curl upload')
conn = pycurl.Curl()
auth = base64.b64encode(
force_bytes("%s:%s" %
(testsettings.FEDORA_USER, testsettings.FEDORA_PASSWORD)))
headers = {'Authorization': 'Basic %s' % force_text(auth)}
conn.setopt(conn.URL, '%supload' % testsettings.FEDORA_ROOT_NONSSL)
conn.setopt(pycurl.VERBOSE, 1)
conn.setopt(pycurl.HTTPHEADER, ["%s: %s" % t for t in headers.items()])
filesize = os.path.getsize(filename)
widgets = [
'Upload: ',
progressbar.widgets.Percentage(), ' ',
progressbar.widgets.Bar(), ' ',
progressbar.widgets.ETA(), ' ',
progressbar.widgets.FileTransferSpeed()
]
# set initial progressbar size based on file; will be slightly larger because
# of multipart boundary content
pbar = progressbar.ProgressBar(widgets=widgets, maxval=filesize).start()
def progress(dl_total, dl, up_total, up):
# update the progressbar to actual maxval (content + boundary)
pbar.max_value = up_total
# update current status
pbar.update(up)
conn.setopt(
conn.HTTPPOST,
[
(
'file',
(
# upload the contents of this file
conn.FORM_FILE,
filename,
# specify a different file name for the upload
conn.FORM_FILENAME,
'file',
# specify a different content type
# conn.FORM_CONTENTTYPE, 'application/x-python',
)),
])
# conn.setopt(conn.CURLOPT_READFUNCTION)
conn.setopt(conn.XFERINFOFUNCTION, progress)
conn.setopt(conn.NOPROGRESS, False)
conn.perform()
# HTTP response code, e.g. 200.
print('Status: %d' % conn.getinfo(conn.RESPONSE_CODE))
# Elapsed time for the transfer.
print('Time: %f' % conn.getinfo(conn.TOTAL_TIME))
conn.close()
def test_compareDatastreamChecksum(self):
# create datastream with checksum
(added, ds) = self._add_text_datastream()
r = self.rest_api.compareDatastreamChecksum(self.pid, ds['id'])
mdsum = hashlib.md5()
mdsum.update(force_bytes(self.TEXT_CONTENT))
text_md5 = mdsum.hexdigest()
self.assert_('<dsChecksum>%s</dsChecksum>' % text_md5 in r.text)
# FIXME: how to test that checksum has actually been checked?
# check for log message in audit trail
r = self.rest_api.getObjectXML(self.pid)
self.assert_(ds['logMessage'] in r.text)
def test_compareDatastreamChecksum(self):
# create datastream with checksum
(added, ds) = self._add_text_datastream()
r = self.rest_api.compareDatastreamChecksum(self.pid, ds["id"])
mdsum = hashlib.md5()
mdsum.update(force_bytes(self.TEXT_CONTENT))
text_md5 = mdsum.hexdigest()
self.assert_("<dsChecksum>%s</dsChecksum>" % text_md5 in r.text)
# FIXME: how to test that checksum has actually been checked?
# check for log message in audit trail
r = self.rest_api.getObjectXML(self.pid)
self.assert_(ds["logMessage"] in r.text)
def test_ingest_without_pid(self):
obj = self.loadFixtureData('basic-object.foxml')
pid = self.repo.ingest(force_bytes(obj))
self.assertTrue(pid)
self.repo.purge_object(force_text(pid))
# test ingesting with log message
pid = self.repo.ingest(obj, "this is my test ingest message")
# ingest message is stored in AUDIT datastream
# - can currently only be accessed by retrieving entire object xml
r = self.repo.api.getObjectXML(force_text(pid))
self.assertTrue("this is my test ingest message" in r.text)
purged = self.repo.purge_object(force_text(pid), "removing test ingest object")
self.assertTrue(purged)
def curl_download_file(pid, dsid):
repo = Repository(testsettings.FEDORA_ROOT_NONSSL,
testsettings.FEDORA_USER, testsettings.FEDORA_PASSWORD)
obj = repo.get_object(pid)
ds = obj.getDatastreamObject(dsid)
tmpfile = tempfile.NamedTemporaryFile(prefix='%s-%s_' % (pid, dsid),
delete=False)
print('writing to ', tmpfile.name)
widgets = [
'Download: ',
progressbar.widgets.Percentage(), ' ',
progressbar.widgets.Bar(), ' ',
progressbar.widgets.ETA(), ' ',
progressbar.widgets.FileTransferSpeed()
]
# set initial progressbar size based on file; will be slightly larger because
# of multipart boundary content
pbar = progressbar.ProgressBar(widgets=widgets, max_value=ds.size).start()
def progress(dl_total, dl, up_total, up):
# update current status
pbar.update(dl)
c = pycurl.Curl()
auth = base64.b64encode(
force_bytes("%s:%s" %
(testsettings.FEDORA_USER, testsettings.FEDORA_PASSWORD)))
headers = {'Authorization': 'Basic %s' % force_text(auth)}
c.setopt(pycurl.VERBOSE, 1)
c.setopt(pycurl.HTTPHEADER, ["%s: %s" % t for t in headers.items()])
# /objects/{pid}/datastreams/{dsID}/content ? [asOfDateTime] [download]
c.setopt(c.URL, '%sobjects/%s/datastreams/%s/content' % \
(testsettings.FEDORA_ROOT_NONSSL, pid, dsid))
# c.setopt(c.WRITEDATA, buffer)
c.setopt(c.WRITEFUNCTION, tmpfile.write)
c.setopt(c.XFERINFOFUNCTION, progress)
c.setopt(c.NOPROGRESS, False)
c.perform()
# HTTP response code, e.g. 200.
print('Status: %d' % c.getinfo(c.RESPONSE_CODE))
# Elapsed time for the transfer.
print('Time: %f' % c.getinfo(c.TOTAL_TIME))
c.close()
def curl_download_file(pid, dsid):
repo = Repository(testsettings.FEDORA_ROOT_NONSSL, testsettings.FEDORA_USER,
testsettings.FEDORA_PASSWORD)
obj = repo.get_object(pid)
ds = obj.getDatastreamObject(dsid)
tmpfile = tempfile.NamedTemporaryFile(
prefix='%s-%s_' % (pid, dsid), delete=False)
print('writing to ', tmpfile.name)
widgets = ['Download: ', progressbar.widgets.Percentage(), ' ',
progressbar.widgets.Bar(), ' ', progressbar.widgets.ETA(),
' ', progressbar.widgets.FileTransferSpeed()]
# set initial progressbar size based on file; will be slightly larger because
# of multipart boundary content
pbar = progressbar.ProgressBar(widgets=widgets, max_value=ds.size).start()
def progress(dl_total, dl, up_total, up):
# update current status
pbar.update(dl)
c = pycurl.Curl()
auth = base64.b64encode(force_bytes("%s:%s" % (testsettings.FEDORA_USER, testsettings.FEDORA_PASSWORD)))
headers = {'Authorization' : 'Basic %s' % force_text(auth)}
c.setopt(pycurl.VERBOSE, 1)
c.setopt(pycurl.HTTPHEADER, ["%s: %s" % t for t in headers.items()])
# /objects/{pid}/datastreams/{dsID}/content ? [asOfDateTime] [download]
c.setopt(c.URL, '%sobjects/%s/datastreams/%s/content' % \
(testsettings.FEDORA_ROOT_NONSSL, pid, dsid))
# c.setopt(c.WRITEDATA, buffer)
c.setopt(c.WRITEFUNCTION, tmpfile.write)
c.setopt(c.XFERINFOFUNCTION, progress)
c.setopt(c.NOPROGRESS, False)
c.perform()
# HTTP response code, e.g. 200.
print('Status: %d' % c.getinfo(c.RESPONSE_CODE))
# Elapsed time for the transfer.
print('Time: %f' % c.getinfo(c.TOTAL_TIME))
c.close()
def test_raw_audit_trail(self):
rqst = Mock()
rqst.method = 'GET'
# created with no ingest message = no audit trail
self.assertRaises(Http404, raw_audit_trail, rqst, self.obj.pid)
# modify object so it will have an audit trail
self.obj.dc.content.title = 'audit this!'
changelog = 'I just changed the title'
self.obj.save(changelog)
response = raw_audit_trail(rqst, self.obj.pid)
expected, got = 200, response.status_code
self.assertEqual(expected, got,
'Expected %s but returned %s for raw_audit_trail' \
% (expected, got))
expected, got = 'text/xml', response['Content-Type']
self.assertEqual(expected, got,
'Expected %s but returned %s for mimetype on raw_audit_trail' \
% (expected, got))
self.assert_(b'<audit:auditTrail' in response.content)
self.assert_(force_bytes('<audit:justification>%s</audit:justification>' % changelog)
in response.content)
self.assert_('Last-Modified' in response)
def upload(self, data, callback=None, content_type=None,
size=None):
'''
Upload a multi-part file for content to ingest. Returns a
temporary upload id that can be used as a datstream location.
:param data: content string, file-like object, or iterable with
content to be uploaded
:param callback: optional callback method to monitor the upload;
see :mod:`requests-toolbelt` documentation for more
details: https://toolbelt.readthedocs.org/en/latest/user.html#uploading-data
:param content_type: optional content type of the data
:param size: optional size of the data; required when using an
iterable for the data
:returns: upload id on success
'''
url = 'upload'
# fedora only expects content uploaded as multipart file;
# make string content into a file-like object so requests.post
# sends it the way Fedora expects.
# NOTE: checking for both python 2.x next method and
# python 3.x __next__ to test if data is iteraable
if not hasattr(data, 'read') and \
not (hasattr(data, '__next__') or hasattr(data, 'next')):
data = six.BytesIO(force_bytes(data))
# if data is an iterable, wrap in a readable iterator that
# requests-toolbelt can read data from
elif not hasattr(data, 'read') and \
(hasattr(data, '__next__') or hasattr(data, 'next')):
if size is None:
raise Exception('Cannot upload iterable with unknown size')
data = ReadableIterator(data, size)
# use requests-toolbelt multipart encoder to avoid reading
# the full content of large files into memory
menc = MultipartEncoder(fields={'file': ('file', data, content_type)})
if callback is not None:
menc = MultipartEncoderMonitor(menc, callback)
headers = {'Content-Type': menc.content_type}
if size:
headers['Content-Length'] = size
try:
response = self.post(url, data=menc, headers=headers)
except OverflowError:
# Python __len__ uses integer so it is limited to system maxint,
# and requests and requests-toolbelt use len() throughout.
# This results in an overflow error when trying to upload a file
# larger than system maxint (2GB on 32-bit OSes).
# See http://bugs.python.org/issue12159
msg = 'upload content larger than system maxint (32-bit OS limitation)'
logger.error('OverflowError: %s', msg)
raise OverflowError(msg)
if response.status_code == requests.codes.accepted:
return response.text.strip()
def load_fixture_data(fname):
with open(fixture_path(fname)) as f:
return force_bytes(f.read())
def object_data(self):
'''Process the archival export and return a buffer with foxml
content for ingest into the destination repository.
:returns: :class:`io.BytesIO` for ingest, with references
to uploaded datastream content or content location urls
'''
self.foxml_buffer = io.BytesIO()
if self.progress_bar:
self.progress_bar.start()
previous_section = None
while True:
try:
section = self.get_next_section()
except StopIteration:
break
if section == BINARY_CONTENT_START:
self.within_file = True
# get datastream info from the end of the section just before this one
# (needed to provide size to upload request)
dsinfo = self.get_datastream_info(previous_section)
if dsinfo:
'Found encoded datastream %(id)s (%(mimetype)s, size %(size)s, %(type)s %(digest)s)' % \
dsinfo
logger.info('Found encoded datastream %(id)s (%(mimetype)s, size %(size)s, %(type)s %(digest)s)',
dsinfo)
else:
# error if datastream info is not found, because either
# size or version date is required to handle content
raise Exception('Failed to find datastream information for %s from \n%s' \
% (self.obj.pid, previous_section))
if self.xml_only and not dsinfo['mimetype'] == 'text/xml': # possibly others?
try:
dsid, dsversion = dsinfo['id'].split('.')
except ValueError:
# if dsid doesn't include a .# (for versioning),
# use the id as is.
dsid = dsinfo['id']
if self.url_credentials:
# if url credentials are set, parse the base fedora api
# url so they can be inserted at the right place
parsed_url = urlparse(self.obj.api.base_url)
# reassemble base url, adding in credentials
base_url = ''.join([parsed_url.scheme, '://',
self.url_credentials, parsed_url.netloc,
parsed_url.path])
else:
base_url = self.obj.api.base_url
# versioned datastream dissemination url
content_location = '%sobjects/%s/datastreams/%s/content?asOfDateTime=%s' % \
(base_url, self.obj.pid, dsid, dsinfo['created'])
else:
upload_args = {}
if self.progress_bar:
def upload_callback(monitor):
self.progress_bar.upload = monitor.bytes_read
upload_args = {'callback': upload_callback}
# use upload id as content location
content_location = self.dest_repo.api.upload(self.encoded_datastream(),
size=int(dsinfo['size']), **upload_args)
self.foxml_buffer.write(force_bytes('<foxml:contentLocation REF="%s" TYPE="URL"/>' \
% content_location))
elif section == BINARY_CONTENT_END:
# should not occur here; this section will be processed by
# encoded_datastream method
self.within_file = False
elif self.within_file:
# should not occur here; this section will be pulled by
# encoded_datastream method
# binary content within a file - ignore here
# (handled by encoded_datastream method)
next
else:
# not start or end of binary content, and not
# within a file, so yield as is (e.g., datastream tags
# between small files)
self.foxml_buffer.write(section)
previous_section = section
return self.foxml_buffer
def object_data(self):
'''Process the archival export and return a buffer with foxml
content for ingest into the destination repository.
:returns: :class:`io.BytesIO` for ingest, with references
to uploaded datastream content or content location urls
'''
self.foxml_buffer = io.BytesIO()
if self.progress_bar:
self.progress_bar.start()
previous_section = None
while True:
try:
section = self.get_next_section()
except StopIteration:
break
if section == BINARY_CONTENT_START:
self.within_file = True
# get datastream info from the end of the section just before this one
# (needed to provide size to upload request)
dsinfo = self.get_datastream_info(previous_section)
if dsinfo:
'Found encoded datastream %(id)s (%(mimetype)s, size %(size)s, %(type)s %(digest)s)' % \
dsinfo
logger.info(
'Found encoded datastream %(id)s (%(mimetype)s, size %(size)s, %(type)s %(digest)s)',
dsinfo)
else:
# error if datastream info is not found, because either
# size or version date is required to handle content
raise Exception(
'Failed to find datastream information from \n%s' %
previous_section)
if self.xml_only and not dsinfo[
'mimetype'] == 'text/xml': # possibly others?
try:
dsid, dsversion = dsinfo['id'].split('.')
except ValueError:
# if dsid doesn't include a .# (for versioning),
# use the id as is.
dsid = dsinfo['id']
if self.url_credentials:
# if url credentials are set, parse the base fedora api
# url so they can be inserted at the right place
parsed_url = urlparse(self.obj.api.base_url)
# reassemble base url, adding in credentials
base_url = ''.join([
parsed_url.scheme, '://', self.url_credentials,
parsed_url.netloc, parsed_url.path
])
else:
base_url = self.obj.api.base_url
# versioned datastream dissemination url
content_location = '%sobjects/%s/datastreams/%s/content?asOfDateTime=%s' % \
(base_url, self.obj.pid, dsid, dsinfo['created'])
else:
upload_args = {}
if self.progress_bar:
def upload_callback(monitor):
self.progress_bar.upload = monitor.bytes_read
upload_args = {'callback': upload_callback}
# use upload id as content location
content_location = self.dest_repo.api.upload(
self.encoded_datastream(),
size=int(dsinfo['size']),
**upload_args)
self.foxml_buffer.write(force_bytes('<foxml:contentLocation REF="%s" TYPE="URL"/>' \
% content_location))
elif section == BINARY_CONTENT_END:
# should not occur here; this section will be processed by
# encoded_datastream method
self.within_file = False
elif self.within_file:
# should not occur here; this section will be pulled by
# encoded_datastream method
# binary content within a file - ignore here
# (handled by encoded_datastream method)
next
else:
# not start or end of binary content, and not
# within a file, so yield as is (e.g., datastream tags
# between small files)
self.foxml_buffer.write(section)
previous_section = section
return self.foxml_buffer
