def get_deposit(self, auth=None, atom_only=False):
# FIXME: this does not deal with the Media Part headers on a multipart deposit
"""
Take a request object and extract from it the parameters and content required for a SWORD deposit. This
includes determining whether this is an Atom Multipart request or not, and extracting the atom/payload where
appropriate. It also includes extracting the HTTP headers which are relevant to deposit, and for those not
supplied providing their defaults in the returned DepositRequest object
"""
d = DepositRequest()
# map the webpy headers to something more standard
mapped_headers = self._map_webpy_headers(request.environ)
# get the headers that have been provided. Any headers which have not been provided will
# will have default values applied
h = HttpHeaders()
d.set_from_headers(h.get_sword_headers(mapped_headers))
if d.content_type.startswith("application/atom+xml"):
atom_only = True
empty_request = False
if d.content_length == 0:
ssslog.info("Received empty deposit request")
empty_request = True
if config.max_upload_size is not None and d.content_length > config.max_upload_size:
raise SwordError(
error_uri=Errors.max_upload_size_exceeded,
msg="Max upload size is " + str(config.max_upload_size) +
"; incoming content length was " + str(d.content_length))
# FIXME: this method does NOT support multipart
# find out if this is a multipart or not
is_multipart = False
# FIXME: these headers aren't populated yet, because the webpy api doesn't
# appear to have a mechanism to retrieve them. urgh.
#entry_part_headers = {}
#media_part_headers = {}
#webin = request.POST
#ssslog.debug(webin)
#if len(webin) == 2:
# ssslog.info("Received multipart deposit request")
# d.atom = webin['atom']
# FIXME: this reads the payload into memory, we need to sort that out
# read the zip file from the base64 encoded string
# d.content = base64.decodestring(webin['payload'])
# is_multipart = True
#elif not empty_request:
if not empty_request:
# for this section, we have to reset the file pointer in the body_file
# part of the request back to the start, since it may have
# already been read once
f = request.body_file
f.seek(0, 0)
# if this wasn't a multipart, and isn't an empty request, then read the
# data from the body_file
if atom_only:
# we don't worry about scalability here - the entries should be
# generally small
ssslog.info("Received Entry deposit request")
d.atom = f.read()
else:
ssslog.info("Received Binary deposit request")
# FIXME: this is reading everything in, and should be re-evaluated for performance/scalability
d.content_file = f
if is_multipart:
d.filename = h.extract_filename(media_part_headers)
else:
d.filename = h.extract_filename(mapped_headers)
# now just attach the authentication data and return
d.auth = auth
return d
def get_deposit(self, web, auth=None, atom_only=False):
# FIXME: this reads files into memory, and therefore does not scale
# FIXME: this does not deal with the Media Part headers on a multipart deposit
"""
Take a web.py web object and extract from it the parameters and content required for a SWORD deposit. This
includes determining whether this is an Atom Multipart request or not, and extracting the atom/payload where
appropriate. It also includes extracting the HTTP headers which are relevant to deposit, and for those not
supplied providing their defaults in the returned DepositRequest object
"""
d = DepositRequest()
# map the webpy headers to something more standard
mapped_headers = self._map_webpy_headers(web.ctx.environ)
# get the headers that have been provided. Any headers which have not been provided will
# will have default values applied
h = HttpHeaders()
d.set_from_headers(h.get_sword_headers(mapped_headers))
if d.content_type.startswith("application/atom+xml"):
atom_only=True
empty_request = False
if d.content_length == 0:
ssslog.info("Received empty deposit request")
empty_request = True
if d.content_length > config.max_upload_size:
raise SwordError(error_uri=Errors.max_upload_size_exceeded,
msg="Max upload size is " + str(config.max_upload_size) +
"; incoming content length was " + str(d.content_length))
# find out if this is a multipart or not
is_multipart = False
# FIXME: these headers aren't populated yet, because the webpy api doesn't
# appear to have a mechanism to retrieve them. urgh.
entry_part_headers = {}
media_part_headers = {}
webin = web.input()
if len(webin) == 2:
ssslog.info("Received multipart deposit request")
d.atom = webin['atom']
# FIXME: this reads the payload into memory, we need to sort that out
# read the zip file from the base64 encoded string
d.content = base64.decodestring(webin['payload'])
is_multipart = True
elif not empty_request:
# if this wasn't a multipart, and isn't an empty request, then the data is in web.data(). This could be a binary deposit or
# an atom entry deposit - reply on the passed/determined argument to determine which
if atom_only:
ssslog.info("Received Entry deposit request")
d.atom = web.data()
else:
ssslog.info("Received Binary deposit request")
d.content = web.data()
if is_multipart:
d.filename = h.extract_filename(media_part_headers)
else:
d.filename = h.extract_filename(mapped_headers)
# now just attach the authentication data and return
d.auth = auth
return d
def get_deposit(self, auth=None, atom_only=False):
# FIXME: this does not deal with the Media Part headers on a multipart deposit
"""
Take a request object and extract from it the parameters and content required for a SWORD deposit. This
includes determining whether this is an Atom Multipart request or not, and extracting the atom/payload where
appropriate. It also includes extracting the HTTP headers which are relevant to deposit, and for those not
supplied providing their defaults in the returned DepositRequest object
"""
d = DepositRequest()
# map the webpy headers to something more standard
mapped_headers = self._map_webpy_headers(request.environ)
# get the headers that have been provided. Any headers which have not been provided will
# will have default values applied
h = HttpHeaders()
d.set_from_headers(h.get_sword_headers(mapped_headers))
if d.content_type.startswith("application/atom+xml"):
atom_only=True
empty_request = False
if d.content_length == 0:
ssslog.info("Received empty deposit request")
empty_request = True
if config.max_upload_size is not None and d.content_length > config.max_upload_size:
raise SwordError(error_uri=Errors.max_upload_size_exceeded,
msg="Max upload size is " + str(config.max_upload_size) +
"; incoming content length was " + str(d.content_length))
# FIXME: this method does NOT support multipart
# find out if this is a multipart or not
is_multipart = False
# FIXME: these headers aren't populated yet, because the webpy api doesn't
# appear to have a mechanism to retrieve them. urgh.
#entry_part_headers = {}
#media_part_headers = {}
#webin = request.POST
#ssslog.debug(webin)
#if len(webin) == 2:
# ssslog.info("Received multipart deposit request")
# d.atom = webin['atom']
# FIXME: this reads the payload into memory, we need to sort that out
# read the zip file from the base64 encoded string
# d.content = base64.decodestring(webin['payload'])
# is_multipart = True
#elif not empty_request:
if not empty_request:
# for this section, we have to reset the file pointer in the body_file
# part of the request back to the start, since it may have
# already been read once
f = request.body_file
f.seek(0, 0)
# if this wasn't a multipart, and isn't an empty request, then read the
# data from the body_file
if atom_only:
# we don't worry about scalability here - the entries should be
# generally small
ssslog.info("Received Entry deposit request")
d.atom = f.read()
else:
ssslog.info("Received Binary deposit request")
# FIXME: this is reading everything in, and should be re-evaluated for performance/scalability
d.content_file = f
if is_multipart:
d.filename = h.extract_filename(media_part_headers)
else:
d.filename = h.extract_filename(mapped_headers)
# now just attach the authentication data and return
d.auth = auth
return d