def _setattr(req, *opts):
"""
Sets entity attributes on the working document
:param req: The request
:param opts: Options (not used)
:return: A modified working document
Transforms the working document by setting the specified attribute on all of the EntityDescriptor
elements of the active document.
**Examples**
.. code-block:: yaml
- setattr:
attr1: value1
attr2: value2
...
Normally this would be combined with the 'merge' feature of fork to add attributes to the working
document for later processing.
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
for e in iter_entities(req.t):
#log.debug("setting %s on %s" % (req.args,e.get('entityID')))
req.md.set_entity_attributes(e, req.args)
return req.t
def update(self, t, tid=None, ts=None, merge_strategy=None):
# log.debug("memory store update: %s: %s" % (repr(t), tid))
relt = root(t)
assert (relt is not None)
ne = 0
if relt.tag == "{%s}EntityDescriptor" % NS['md']:
# log.debug("memory store setting entity descriptor")
self._unindex(relt)
self._index(relt)
self.entities[relt.get('entityID')] = relt # TODO: merge?
if tid is not None:
self.md[tid] = [relt.get('entityID')]
ne += 1
# log.debug("keys %s" % self.md.keys())
elif relt.tag == "{%s}EntitiesDescriptor" % NS['md']:
if tid is None:
tid = relt.get('Name')
lst = []
for e in iter_entities(t):
self.update(e)
lst.append(e.get('entityID'))
ne += 1
self.md[tid] = lst
return ne
def first(req, *opts):
"""
If the working document is a single EntityDescriptor, strip the outer EntitiesDescriptor element and return it.
:param req: The request
:param opts: Options (unused)
:return: returns the first entity descriptor if the working document only contains one
Sometimes (eg when running an MDX pipeline) it is usually expected that if a single EntityDescriptor is being returned
then the outer EntitiesDescriptor is stripped. This method does exactly that:
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
gone = object() # sentinel
entities = iter_entities(req.t)
one = next(entities, gone)
if one is gone:
return req.t # empty tree - return it as is
two = next(entities, gone) # one EntityDescriptor in tree - return just that one
if two is gone:
return one
return req.t
def first(req, *opts):
"""
If the working document is a single EntityDescriptor, strip the outer EntitiesDescriptor element and return it.
:param req: The request
:param opts: Options (unused)
:return: returns the first entity descriptor if the working document only contains one
Sometimes (eg when running an MDX pipeline) it is usually expected that if a single EntityDescriptor is being returned
then the outer EntitiesDescriptor is stripped. This method does exactly that:
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
gone = object() # sentinel
entities = iter_entities(req.t)
one = next(entities, gone)
if one is gone:
return req.t # empty tree - return it as is
two = next(entities,
gone) # one EntityDescriptor in tree - return just that one
if two is gone:
return one
return req.t
def _reginfo(req, *opts):
"""
Sets registration info extension on EntityDescription element
:param req: The request
:param opts: Options (not used)
:return: A modified working document
Transforms the working document by setting the specified attribute on all of the EntityDescriptor
elements of the active document.
**Examples**
.. code-block:: yaml
- reginfo:
[policy:
<lang>: <registration policy URL>]
authority: <registrationAuthority URL>
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
for e in iter_entities(req.t):
req.md.set_reginfo(e, **req.args)
return req.t
def _setattr(req, *opts):
"""
Sets entity attributes on the working document
:param req: The request
:param opts: Options (not used)
:return: A modified working document
Transforms the working document by setting the specified attribute on all of the EntityDescriptor
elements of the active document.
**Examples**
.. code-block:: yaml
- setattr:
attr1: value1
attr2: value2
...
Normally this would be combined with the 'merge' feature of fork to add attributes to the working
document for later processing.
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
for e in iter_entities(req.t):
# log.debug("setting %s on %s" % (req.args,e.get('entityID')))
req.md.set_entity_attributes(e, req.args)
return req.t
def _reginfo(req, *opts):
"""
Sets registration info extension on EntityDescription element
:param req: The request
:param opts: Options (not used)
:return: A modified working document
Transforms the working document by setting the specified attribute on all of the EntityDescriptor
elements of the active document.
**Examples**
.. code-block:: yaml
- reginfo:
[policy:
<lang>: <registration policy URL>]
authority: <registrationAuthority URL>
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
for e in iter_entities(req.t):
req.md.set_reginfo(e, **req.args)
return req.t
def update(self, t, tid=None, ts=None, merge_strategy=None):
# log.debug("memory store update: %s: %s" % (repr(t), tid))
relt = root(t)
assert (relt is not None)
ne = 0
if relt.tag == "{%s}EntityDescriptor" % NS['md']:
# log.debug("memory store setting entity descriptor")
self._unindex(relt)
self._index(relt)
self.entities[relt.get('entityID')] = relt # TODO: merge?
if tid is not None:
self.md[tid] = [relt.get('entityID')]
ne += 1
# log.debug("keys %s" % self.md.keys())
elif relt.tag == "{%s}EntitiesDescriptor" % NS['md']:
if tid is None:
tid = relt.get('Name')
lst = []
for e in iter_entities(t):
self.update(e)
lst.append(e.get('entityID'))
ne += 1
self.md[tid] = lst
return ne
def update(self,
t,
tid=None,
ts=None,
merge_strategy=None): # TODO: merge ?
log.debug("redis store update: %s: %s" % (t, tid))
relt = root(t)
ne = 0
if ts is None:
ts = int(
_now() +
3600 * 24 * 4) # 4 days is the arbitrary default expiration
if relt.tag == "{%s}EntityDescriptor" % NS['md']:
if tid is None:
tid = relt.get('entityID')
with self.rc.pipeline() as p:
self.update_entity(relt, t, tid, ts, p)
entity_id = relt.get("entityID")
if entity_id is not None:
self.membership("entities", entity_id, ts, p)
for ea, eav in entity_attribute_dict(relt).iteritems():
for v in eav:
# log.debug("%s=%s" % (ea, v))
self.membership("{%s}%s" % (ea, v), tid, ts, p)
p.zadd("%s#values" % ea, v, ts)
p.sadd("#attributes", ea)
for hn in ('sha1', 'sha256', 'md5'):
tid_hash = hex_digest(tid, hn)
p.set("{%s}%s#alias" % (hn, tid_hash), tid)
if ts is not None:
p.expireat(tid_hash, ts)
p.execute()
ne += 1
elif relt.tag == "{%s}EntitiesDescriptor" % NS['md']:
if tid is None:
tid = relt.get('Name')
ts = self._expiration(relt)
with self.rc.pipeline() as p:
self.update_entity(relt, t, tid, ts, p)
for e in iter_entities(t):
ne += self.update(e, ts=ts)
entity_id = e.get("entityID")
if entity_id is not None:
self.membership(tid, entity_id, ts, p)
self.membership("entities", entity_id, ts, p)
p.execute()
else:
raise ValueError("Bad metadata top-level element: '%s'" %
root(t).tag)
return ne
def discojson(req, *opts):
"""
Return a discojuice-compatible json representation of the tree
:param req: The request
:param opts: Options (unusued)
:return: returns a JSON array
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
return json.dumps([req.md.discojson(e) for e in iter_entities(req.t)])
def discojson(req, *opts):
"""
Return a discojuice-compatible json representation of the tree
:param req: The request
:param opts: Options (unusued)
:return: returns a JSON array
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
res = [req.md.discojson(e) for e in iter_entities(req.t)]
res.sort(key=operator.itemgetter('title'))
return json.dumps(res)
def update(self, t, tid=None, ts=None, merge_strategy=None): # TODO: merge ?
log.debug("redis store update: %s: %s" % (t, tid))
relt = root(t)
ne = 0
if ts is None:
ts = int(_now() + 3600 * 24 * 4) # 4 days is the arbitrary default expiration
if relt.tag == "{%s}EntityDescriptor" % NS['md']:
if tid is None:
tid = relt.get('entityID')
with self.rc.pipeline() as p:
self.update_entity(relt, t, tid, ts, p)
entity_id = relt.get("entityID")
if entity_id is not None:
self.membership("entities", entity_id, ts, p)
for ea, eav in entity_attribute_dict(relt).iteritems():
for v in eav:
# log.debug("%s=%s" % (ea, v))
self.membership("{%s}%s" % (ea, v), tid, ts, p)
p.zadd("%s#values" % ea, v, ts)
p.sadd("#attributes", ea)
for hn in ('sha1', 'sha256', 'md5'):
tid_hash = hex_digest(tid, hn)
p.set("{%s}%s#alias" % (hn, tid_hash), tid)
if ts is not None:
p.expireat(tid_hash, ts)
p.execute()
ne += 1
elif relt.tag == "{%s}EntitiesDescriptor" % NS['md']:
if tid is None:
tid = relt.get('Name')
ts = self._expiration(relt)
with self.rc.pipeline() as p:
self.update_entity(relt, t, tid, ts, p)
for e in iter_entities(t):
ne += self.update(e, ts=ts)
entity_id = e.get("entityID")
if entity_id is not None:
self.membership(tid, entity_id, ts, p)
self.membership("entities", entity_id, ts, p)
p.execute()
else:
raise ValueError("Bad metadata top-level element: '%s'" % root(t).tag)
return ne
def store(req, *opts):
"""
Save the working document as separate files
:param req: The request
:param opts: Options (unused)
:return: always returns the unmodified working document
Split the working document into EntityDescriptor-parts and save in directory/sha1(@entityID).xml. Note that
this does not erase files that may already be in the directory. If you want a "clean" directory, remove it
before you call store.
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
if not req.args:
raise PipeException("store requires an argument")
target_dir = None
if type(req.args) is dict:
target_dir = req.args.get('directory', None)
else:
target_dir = req.args[0]
if target_dir is not None:
if not os.path.isdir(target_dir):
os.makedirs(target_dir)
for e in iter_entities(req.t):
eid = e.get('entityID')
if eid is None or len(eid) == 0:
raise PipeException("Missing entityID in %s" % e)
m = hashlib.sha1()
m.update(eid)
d = m.hexdigest()
safe_write("%s.xml" % os.path.join(target_dir, d),
dumptree(e, pretty_print=True))
return req.t
def store(req, *opts):
"""
Save the working document as separate files
:param req: The request
:param opts: Options (unused)
:return: always returns the unmodified working document
Split the working document into EntityDescriptor-parts and save in directory/sha1(@entityID).xml. Note that
this does not erase files that may already be in the directory. If you want a "clean" directory, remove it
before you call store.
"""
if req.t is None:
raise PipeException("Your pipeline is missing a select statement.")
if not req.args:
raise PipeException("store requires an argument")
target_dir = None
if type(req.args) is dict:
target_dir = req.args.get('directory', None)
else:
target_dir = req.args[0]
if target_dir is not None:
if not os.path.isdir(target_dir):
os.makedirs(target_dir)
for e in iter_entities(req.t):
eid = e.get('entityID')
if eid is None or len(eid) == 0:
raise PipeException("Missing entityID in %s" % e)
m = hashlib.sha1()
m.update(eid)
d = m.hexdigest()
safe_write("%s.xml" % os.path.join(target_dir, d), dumptree(e, pretty_print=True))
return req.t