def __init__(self, connection):
self.api = API(connection)
class FedoraClient(object):
def __init__(self, connection):
self.api = API(connection)
def getNextPID(self, namespace, numPIDs=1, format=u'text/xml'):
request = self.api.getNextPID()
response = request.submit(namespace=namespace,
numPIDs=numPIDs,
format=format)
xml = response.read()
response.close()
'''
The following use of default namespace is to make this code work in 3.4 and 3.5 where the namespacing is different
'''
doc = etree.fromstring(xml)
fake_namespace_dictionary = {}
try:
fake_namespace_dictionary['default'] = doc.nsmap[None]
ids = [id.decode('utf8') for id in doc.xpath('/default:pidList/default:pid/text()', namespaces = fake_namespace_dictionary)]
except KeyError:
ids = [id.decode('utf8') for id in doc.xpath('/pidList/pid/text()')]
if len(ids) == 1:
return ids[0]
return ids
def createObject(self, pid, label, state=u'A'):
foxml = ElementMaker(namespace=NSMAP['foxml'], nsmap=NSMAP)
foxml_state = {'A': u'Active',
'I': u'Inactive',
'D': u'Deleted'}[state]
doc = foxml.digitalObject(
foxml.objectProperties(
foxml.property(NAME="info:fedora/fedora-system:def/model#state",
VALUE=state),
foxml.property(NAME="info:fedora/fedora-system:def/model#label",
VALUE=label)
),
VERSION='1.1', PID=pid)
body = etree.tostring(doc, encoding="UTF-8", xml_declaration=False)
request = self.api.createObject(pid=pid)
request.headers['Content-Type'] = 'text/xml; charset=utf-8'
response = request.submit(body, state=state[0], label=label)
return self.getObject(pid)
def getObject(self, pid):
return FedoraObject(pid, self)
def getObjectProfile(self, pid):
request = self.api.getObjectProfile(pid=pid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = {'ownerId': u''}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = {'objLabel': 'label',
'objOwnerId': 'ownerId',
'objCreateDate': 'createdDate',
'objLastModDate': 'lastModifiedDate',
'objState': 'state'}.get(child.tag.split('}')[-1])
if name is None or child.text is None:
continue
value = child.text
if not isinstance(value, unicode):
value = value.decode('utf8')
result[name] = value
return result
def updateObject(self, pid, body='', **params):
request = self.api.updateObject(pid=pid)
response = request.submit(body, **params)
def deleteObject(self, pid, **params):
request = self.api.deleteObject(pid=pid)
response = request.submit(**params)
def listDatastreams(self, pid):
request = self.api.listDatastreams(pid=pid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
return [child.attrib['dsid'] for child in doc]
def addDatastream(self, pid, dsid, body='', **params):
if dsid == 'RELS-EXT' and not body:
body = ('<rdf:RDF xmlns:rdf="%s"/>' % NS.rdf)
params['mimeType'] = u'application/rdf+xml'
params['formatURI'] = (
u'info:fedora/fedora-system:FedoraRELSExt-1.0')
if params.get('controlGroup', u'X') == u'X':
if not 'mimeType' in params:
params['mimeType'] = u'text/xml'
if not 'mimeType' in params:
params['mimeType'] = u'application/binary'
if 'checksumType' not in params:
params['checksumType'] = u'MD5'
params = self._fix_ds_params(params)
request = self.api.addDatastream(pid=pid, dsID=dsid)
request.headers['Content-Type'] = params['mimeType']
response = request.submit(body, **params)
def _fix_ds_params(self, params):
for name, param in params.items():
newname = {'label': 'dsLabel',
'location': 'dsLocation',
'state': 'dsState'}.get(name, name)
if newname != name:
params[newname] = param
del params[name]
return params
def getDatastreamProfile(self, pid, dsid):
request = self.api.getDatastreamProfile(pid=pid, dsID=dsid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = {}
tags = {
'dsLabel': 'label',
'dsVerionId': 'versionId',
'dsCreateDate': 'createdDate',
'dsState': 'state',
'dsMIME': 'mimeType',
'dsFormatURI': 'formatURI',
'dsControlGroup': 'controlGroup',
'dsSize': 'size',
'dsVersionable': 'versionable',
'dsInfoType': 'infoType',
'dsLocation': 'location',
'dsLocationType': 'locationType',
'dsChecksum': 'checksum',
'dsChecksumType': 'checksumType'
}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = tags.get(child.tag.rpartition('}')[2], None)
if name is not None:
value = child.text
if value and not isinstance(value, unicode):
value = value.decode('utf8')
result[name] = value
return result
def modifyDatastream(self, pid, dsid, body='', **params):
params = self._fix_ds_params(params)
request = self.api.modifyDatastream(pid=pid, dsID=dsid)
response = request.submit(body, **params)
def getDatastream(self, pid, dsid):
request = self.api.getDatastream(pid=pid, dsID=dsid)
return request.submit()
def deleteDatastream(self, pid, dsid, **params):
request = self.api.deleteDatastream(pid=pid, dsID=dsid)
return request.submit(**params)
def getAllObjectMethods(self, pid, **params):
params['format'] = u'text/xml'
request = self.api.getAllObjectMethods(pid=pid)
response = request.submit(**params)
xml = response.read()
doc = etree.fromstring(xml)
method_list = []
for sdef_el in doc:
sdef = sdef_el.attrib['pid']
for method_el in sdef_el:
method = method_el.attrib['name']
method_list.append((sdef, method))
response.close()
return method_list
def invokeSDefMethodUsingGET(self, pid, sdef, method, **params):
request = self.api.invokeSDefMethodUsingGET(pid=pid, sDef=sdef,
method=method)
return request.submit(**params)
def searchObjects(self, query, fields, terms=False, maxResults=10):
field_params = {}
assert isinstance(fields, list)
for field in fields:
field_params[field] = u'true'
token = True
NS = 'http://www.fedora.info/definitions/1/0/types/'
while token:
if token is True:
token = False
request = self.api.searchObjects()
request.undocumented_params = field_params
if token:
if terms:
response = request.submit(terms=query,
sessionToken=token,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
response = request.submit(query=query,
sessionToken=token,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
if terms:
response = request.submit(terms=query,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
response = request.submit(query=query,
maxResults=maxResults,
resultFormat=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
tokens = doc.xpath('//f:token/text()', namespaces={'f': NS})
if tokens:
token = tokens[0].decode('utf8')
else:
token = False
for result in doc.xpath('//f:objectFields', namespaces={'f': NS}):
data = defaultdict(list)
for child in result:
field_name = child.tag.split('}')[-1].decode('utf8')
value = child.text or u''
if not isinstance(value, unicode):
value = value.decode('utf8')
data[field_name].append(value)
yield data
def searchTriples(self, query, lang='sparql', format='Sparql',
limit=100, type='tuples', dt='on', flush=True):
flush = str(flush).lower()
URL_pramaters = {'query':query,
'lang':lang,
'flush': flush,
'format':format,
'type':type,
'dt':dt}
#conditionaly set limit if there is one. (so it can be set to None)
if limit:
URL_pramaters['limit'] = limit
# Encode in utf8 to let unicode pass through urlencode
for key in URL_pramaters:
URL_pramaters[key] = URL_pramaters[key].encode('utf8')
url = u'/risearch?%s' % urllib.urlencode(URL_pramaters)
#Fedora started needing authentication in 3.5 for RI, tested in 3.4 as well
headers = copy(self.api.connection.form_headers)
headers['Accept:'] = 'text/xml'
response = self.api.connection.open(url, '', headers, method='POST')
xml = response.read()
doc = etree.fromstring(xml)
NS = 'http://www.w3.org/2001/sw/DataAccess/rf1/result' # ouch, old!
for result in doc.xpath('//sparql:result', namespaces={'sparql': NS}):
data = {}
for el in result:
name = el.tag.split('}')[-1]
value = {}
uri = el.attrib.get('uri')
if uri:
value['value'] = uri.decode('utf8')
value['type'] = 'uri'
else:
value['type'] = 'literal'
if isinstance(el.text, unicode):
value['value'] = el.text
elif el.text:
value['value'] = el.text.decode('utf8')
else:
value['value'] = u''
datatype = el.attrib.get('datatype')
lang = el.attrib.get('lang')
if datatype:
value['datatype'] = datatype
elif lang:
value['lang'] = lang
data[name] = value
yield data
def __init__(self, connection):
self.api = API(connection)
class FedoraClient(object):
def __init__(self, connection):
self.api = API(connection)
def getNextPID(self, namespace, numPIDs=1, format=u"text/xml"):
request = self.api.getNextPID()
response = request.submit(namespace=namespace, numPIDs=numPIDs, format=format)
xml = response.read()
response.close()
doc = etree.fromstring(xml)
ids = [id.decode("utf8") for id in doc.xpath("/pidList/pid/text()")]
if len(ids) == 1:
return ids[0]
return ids
def createObject(self, pid, label, state=u"A"):
foxml = ElementMaker(namespace=NSMAP["foxml"], nsmap=NSMAP)
foxml_state = {"A": u"Active", "I": u"Inactive", "D": u"Deleted"}[state]
doc = foxml.digitalObject(
foxml.objectProperties(
foxml.property(NAME="info:fedora/fedora-system:def/model#state", VALUE=state),
foxml.property(NAME="info:fedora/fedora-system:def/model#label", VALUE=label),
),
VERSION="1.1",
PID=pid,
)
body = etree.tostring(doc, encoding="UTF-8", xml_declaration=False)
request = self.api.createObject(pid=pid)
request.headers["Content-Type"] = "text/xml; charset=utf-8"
response = request.submit(body, state=state[0], label=label)
return self.getObject(pid)
def getObject(self, pid):
return FedoraObject(pid, self)
def getObjectProfile(self, pid):
request = self.api.getObjectProfile(pid=pid)
response = request.submit(format=u"text/xml")
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = {"ownerId": u""}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = {
"objLabel": "label",
"objOwnerId": "ownerId",
"objCreateDate": "createdDate",
"objLastModDate": "lastModifiedDate",
"objState": "state",
}.get(child.tag.split("}")[-1])
if name is None or child.text is None:
continue
value = child.text
if not isinstance(value, unicode):
value = value.decode("utf8")
result[name] = value
return result
def updateObject(self, pid, body="", **params):
request = self.api.updateObject(pid=pid)
response = request.submit(body, **params)
def deleteObject(self, pid, **params):
request = self.api.deleteObject(pid=pid)
response = request.submit(**params)
def listDatastreams(self, pid):
request = self.api.listDatastreams(pid=pid)
response = request.submit(format=u"text/xml")
xml = response.read()
response.close()
doc = etree.fromstring(xml)
return [child.attrib["dsid"] for child in doc]
def addDatastream(self, pid, dsid, body="", **params):
if dsid == "RELS-EXT" and not body:
body = '<rdf:RDF xmlns:rdf="%s"/>' % NS.rdf
params["mimeType"] = u"application/rdf+xml"
params["formatURI"] = u"info:fedora/fedora-system:FedoraRELSExt-1.0"
if params.get("controlGroup", u"X") == u"X":
if not "mimeType" in params:
params["mimeType"] = u"text/xml"
if not "mimeType" in params:
params["mimeType"] = u"application/binary"
if "checksumType" not in params:
params["checksumType"] = u"MD5"
params = self._fix_ds_params(params)
request = self.api.addDatastream(pid=pid, dsID=dsid)
request.headers["Content-Type"] = params["mimeType"]
response = request.submit(body, **params)
def _fix_ds_params(self, params):
for name, param in params.items():
newname = {"label": "dsLabel", "location": "dsLocation", "state": "dsState"}.get(name, name)
if newname != name:
params[newname] = param
del params[name]
return params
def getDatastreamProfile(self, pid, dsid):
request = self.api.getDatastreamProfile(pid=pid, dsID=dsid)
response = request.submit(format=u"text/xml")
xml = response.read()
response.close()
doc = etree.fromstring(xml)
return self.__parseDatastreamProfile(doc)
def __parseDatastreamProfile(self, doc):
result = {}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = {
"dsLabel": "label",
"dsVerionId": "versionId",
"dsCreateDate": "createdDate",
"dsState": "state",
"dsMIME": "mimeType",
"dsFormatURI": "formatURI",
"dsControlGroup": "controlGroup",
"dsSize": "size",
"dsVersionable": "versionable",
"dsInfoType": "infoType",
"dsLocation": "location",
"dsLocationType": "locationType",
"dsChecksum": "checksum",
"dsChecksumType": "checksumType",
}.get(child.tag.split("}")[-1])
if name is None or child.text is None:
continue
value = child.text
if not isinstance(value, unicode):
value = value.decode("utf8")
result[name] = value
return result
def modifyDatastream(self, pid, dsid, body="", **params):
params = self._fix_ds_params(params)
request = self.api.modifyDatastream(pid=pid, dsID=dsid)
response = request.submit(body, **params)
v = response.read() # 2012-11-22 (whikloj) Need to read the whole response or your next GET
# throws a httplib.ResponseNotReady
def getDatastream(self, pid, dsid):
request = self.api.getDatastream(pid=pid, dsID=dsid)
return request.submit()
def getDatastreamHistory(self, pid, dsid):
request = self.api.getDatastreamHistory(pid=pid, dsID=dsid)
response = request.submit(format=u"text/xml")
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = []
for child in doc:
if child.tag.split("}")[-1] == "datastreamProfile":
result.append(self.__parseDatastreamProfile(child))
return result
def deleteDatastream(self, pid, dsid, **params):
request = self.api.deleteDatastream(pid=pid, dsID=dsid)
return request.submit(**params)
def getAllObjectMethods(self, pid, **params):
params["format"] = u"text/xml"
request = self.api.getAllObjectMethods(pid=pid)
response = request.submit(**params)
xml = response.read()
doc = etree.fromstring(xml)
method_list = []
for sdef_el in doc:
sdef = sdef_el.attrib["pid"]
for method_el in sdef_el:
method = method_el.attrib["name"]
method_list.append((sdef, method))
response.close()
return method_list
def invokeSDefMethodUsingGET(self, pid, sdef, method, **params):
request = self.api.invokeSDefMethodUsingGET(pid=pid, sDef=sdef, method=method)
return request.submit(**params)
def searchObjects(self, query, fields, terms=False, maxResults=10):
field_params = {}
assert isinstance(fields, list)
for field in fields:
field_params[field] = u"true"
token = True
NS = "http://www.fedora.info/definitions/1/0/types/"
while token:
if token is True:
token = False
request = self.api.searchObjects()
request.undocumented_params = field_params
if token:
if terms:
response = request.submit(
terms=query, sessionToken=token, maxResults=maxResults, resultFormat=u"text/xml"
)
else:
response = request.submit(
query=query, sessionToken=token, maxResults=maxResults, resultFormat=u"text/xml"
)
else:
if terms:
response = request.submit(terms=query, maxResults=maxResults, resultFormat=u"text/xml")
else:
response = request.submit(query=query, maxResults=maxResults, resultFormat=u"text/xml")
xml = response.read()
response.close()
doc = etree.fromstring(xml)
tokens = doc.xpath("//f:token/text()", namespaces={"f": NS})
if tokens:
token = tokens[0].decode("utf8")
else:
token = False
for result in doc.xpath("//f:objectFields", namespaces={"f": NS}):
data = defaultdict(list)
for child in result:
field_name = child.tag.split("}")[-1].decode("utf8")
value = child.text or u""
if not isinstance(value, unicode):
value = value.decode("utf8")
data[field_name].append(value)
yield data
def searchTriples(self, query, lang="sparql", format="Sparql", limit=100, type="tuples", dt="on", flush=True):
flush = str(flush).lower()
url = u"/risearch?%s" % urllib.urlencode(
{"query": query, "lang": lang, "flush": flush, "format": format, "limit": limit, "type": type, "dt": dt}
)
headers = {"Accept:": "text/xml"}
response = self.api.connection.open(url, "", headers, method="POST")
xml = response.read()
doc = etree.fromstring(xml)
NS = "http://www.w3.org/2001/sw/DataAccess/rf1/result" # ouch, old!
for result in doc.xpath("//sparql:result", namespaces={"sparql": NS}):
data = {}
for el in result:
name = el.tag.split("}")[-1]
value = {}
uri = el.attrib.get("uri")
if uri:
value["value"] = uri.decode("utf8")
value["type"] = "uri"
else:
value["type"] = "literal"
if isinstance(el.text, unicode):
value["value"] = el.text
elif el.text:
value["value"] = el.text.decode("utf8")
else:
value["value"] = u""
datatype = el.attrib.get("datatype")
lang = el.attrib.get("lang")
if datatype:
value["datatype"] = datatype
elif lang:
value["lang"] = lang
data[name] = value
yield data
class FedoraClient(object):
def __init__(self, connection):
self.api = API(connection)
def getNextPID(self, namespace, numPIDs=1, format=u'text/xml'):
request = self.api.getNextPID()
response = request.submit(namespace=namespace,
numPIDs=numPIDs,
format=format)
xml = response.read()
response.close()
'''
The following use of default namespace is to make this code work in 3.4 and 3.5 where the namespacing is different
'''
doc = etree.fromstring(xml)
fake_namespace_dictionary = {}
try:
fake_namespace_dictionary['default'] = doc.nsmap[None]
ids = [
id.decode('utf8')
for id in doc.xpath('/default:pidList/default:pid/text()',
namespaces=fake_namespace_dictionary)
]
except KeyError:
ids = [
id.decode('utf8') for id in doc.xpath('/pidList/pid/text()')
]
if len(ids) == 1:
return ids[0]
return ids
def createObject(self, pid, label, state=u'A'):
foxml = ElementMaker(namespace=NSMAP['foxml'], nsmap=NSMAP)
foxml_state = {
'A': u'Active',
'I': u'Inactive',
'D': u'Deleted'
}[state]
doc = foxml.digitalObject(foxml.objectProperties(
foxml.property(NAME="info:fedora/fedora-system:def/model#state",
VALUE=state),
foxml.property(NAME="info:fedora/fedora-system:def/model#label",
VALUE=label)),
VERSION='1.1',
PID=pid)
body = etree.tostring(doc, encoding="UTF-8", xml_declaration=False)
request = self.api.createObject(pid=pid)
request.headers['Content-Type'] = 'text/xml; charset=utf-8'
response = request.submit(body, state=state[0], label=label)
return self.getObject(pid)
def getObject(self, pid):
return FedoraObject(pid, self)
def getObjectProfile(self, pid):
request = self.api.getObjectProfile(pid=pid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = {'ownerId': u''}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = {
'objLabel': 'label',
'objOwnerId': 'ownerId',
'objCreateDate': 'createdDate',
'objLastModDate': 'lastModifiedDate',
'objState': 'state'
}.get(child.tag.split('}')[-1])
if name is None or child.text is None:
continue
value = child.text
if not isinstance(value, unicode):
value = value.decode('utf8')
result[name] = value
return result
def updateObject(self, pid, body='', **params):
request = self.api.updateObject(pid=pid)
response = request.submit(body, **params)
def deleteObject(self, pid, **params):
request = self.api.deleteObject(pid=pid)
response = request.submit(**params)
def listDatastreams(self, pid):
request = self.api.listDatastreams(pid=pid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
return [child.attrib['dsid'] for child in doc]
def addDatastream(self, pid, dsid, body='', **params):
if dsid == 'RELS-EXT' and not body:
body = ('<rdf:RDF xmlns:rdf="%s"/>' % NS.rdf)
params['mimeType'] = u'application/rdf+xml'
params['formatURI'] = (
u'info:fedora/fedora-system:FedoraRELSExt-1.0')
if params.get('controlGroup', u'X') == u'X':
if not 'mimeType' in params:
params['mimeType'] = u'text/xml'
if not 'mimeType' in params:
params['mimeType'] = u'application/binary'
if 'checksumType' not in params:
params['checksumType'] = u'MD5'
params = self._fix_ds_params(params)
request = self.api.addDatastream(pid=pid, dsID=dsid)
request.headers['Content-Type'] = params['mimeType']
response = request.submit(body, **params)
def _fix_ds_params(self, params):
for name, param in params.items():
newname = {
'label': 'dsLabel',
'location': 'dsLocation',
'state': 'dsState'
}.get(name, name)
if newname != name:
params[newname] = param
del params[name]
return params
def getDatastreamProfile(self, pid, dsid):
request = self.api.getDatastreamProfile(pid=pid, dsID=dsid)
response = request.submit(format=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
result = {}
tags = {
'dsLabel': 'label',
'dsVerionId': 'versionId',
'dsCreateDate': 'createdDate',
'dsState': 'state',
'dsMIME': 'mimeType',
'dsFormatURI': 'formatURI',
'dsControlGroup': 'controlGroup',
'dsSize': 'size',
'dsVersionable': 'versionable',
'dsInfoType': 'infoType',
'dsLocation': 'location',
'dsLocationType': 'locationType',
'dsChecksum': 'checksum',
'dsChecksumType': 'checksumType'
}
for child in doc:
# rename elementnames to match property names in foxml
# the xml data is namespaced in 3.4, but not in 3.3, so strip out
# the namespace, to be compatible with both
name = tags.get(child.tag.rpartition('}')[2], None)
if name is not None:
value = child.text
if value and not isinstance(value, unicode):
value = value.decode('utf8')
result[name] = value
return result
def modifyDatastream(self, pid, dsid, body='', **params):
params = self._fix_ds_params(params)
request = self.api.modifyDatastream(pid=pid, dsID=dsid)
response = request.submit(body, **params)
def getDatastream(self, pid, dsid):
request = self.api.getDatastream(pid=pid, dsID=dsid)
return request.submit()
def deleteDatastream(self, pid, dsid, **params):
request = self.api.deleteDatastream(pid=pid, dsID=dsid)
return request.submit(**params)
def getAllObjectMethods(self, pid, **params):
params['format'] = u'text/xml'
request = self.api.getAllObjectMethods(pid=pid)
response = request.submit(**params)
xml = response.read()
doc = etree.fromstring(xml)
method_list = []
for sdef_el in doc:
sdef = sdef_el.attrib['pid']
for method_el in sdef_el:
method = method_el.attrib['name']
method_list.append((sdef, method))
response.close()
return method_list
def invokeSDefMethodUsingGET(self, pid, sdef, method, **params):
request = self.api.invokeSDefMethodUsingGET(pid=pid,
sDef=sdef,
method=method)
return request.submit(**params)
def searchObjects(self, query, fields, terms=False, maxResults=10):
field_params = {}
assert isinstance(fields, list)
for field in fields:
field_params[field] = u'true'
token = True
NS = 'http://www.fedora.info/definitions/1/0/types/'
while token:
if token is True:
token = False
request = self.api.searchObjects()
request.undocumented_params = field_params
if token:
if terms:
response = request.submit(terms=query,
sessionToken=token,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
response = request.submit(query=query,
sessionToken=token,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
if terms:
response = request.submit(terms=query,
maxResults=maxResults,
resultFormat=u'text/xml')
else:
response = request.submit(query=query,
maxResults=maxResults,
resultFormat=u'text/xml')
xml = response.read()
response.close()
doc = etree.fromstring(xml)
tokens = doc.xpath('//f:token/text()', namespaces={'f': NS})
if tokens:
token = tokens[0].decode('utf8')
else:
token = False
for result in doc.xpath('//f:objectFields', namespaces={'f': NS}):
data = defaultdict(list)
for child in result:
field_name = child.tag.split('}')[-1].decode('utf8')
value = child.text or u''
if not isinstance(value, unicode):
value = value.decode('utf8')
data[field_name].append(value)
yield data
def searchTriples(self,
query,
lang='sparql',
format='Sparql',
limit=100,
type='tuples',
dt='on',
flush=True):
flush = str(flush).lower()
URL_pramaters = {
'query': query,
'lang': lang,
'flush': flush,
'format': format,
'type': type,
'dt': dt
}
#conditionaly set limit if there is one. (so it can be set to None)
if limit:
URL_pramaters['limit'] = limit
# Encode in utf8 to let unicode pass through urlencode
for key in URL_pramaters:
URL_pramaters[key] = URL_pramaters[key].encode('utf8')
url = u'/risearch?%s' % urllib.urlencode(URL_pramaters)
#Fedora started needing authentication in 3.5 for RI, tested in 3.4 as well
headers = copy(self.api.connection.form_headers)
headers['Accept:'] = 'text/xml'
response = self.api.connection.open(url, '', headers, method='POST')
xml = response.read()
doc = etree.fromstring(xml)
NS = 'http://www.w3.org/2001/sw/DataAccess/rf1/result' # ouch, old!
for result in doc.xpath('//sparql:result', namespaces={'sparql': NS}):
data = {}
for el in result:
name = el.tag.split('}')[-1]
value = {}
uri = el.attrib.get('uri')
if uri:
value['value'] = uri.decode('utf8')
value['type'] = 'uri'
else:
value['type'] = 'literal'
if isinstance(el.text, unicode):
value['value'] = el.text
elif el.text:
value['value'] = el.text.decode('utf8')
else:
value['value'] = u''
datatype = el.attrib.get('datatype')
lang = el.attrib.get('lang')
if datatype:
value['datatype'] = datatype
elif lang:
value['lang'] = lang
data[name] = value
yield data
