cosminbasca / surfrdf Goto Github PK

here is the code snippet from the federation.py example

Movie = session.get_class(surf.ns.SURF['Movie'], store = 'local')
m1 = Movie('http://baseuri/m1')
m1.surf_title = "Movie 1"

trying to create a blank node with rdf.BNode but couldn't figure out how. any pointer would be appreciated!

ciao, i'm currently evaluating rdf-related libraries for Python. surf is feature- and documentation-wise my favorite atm. unfortunately it lacks Python 3 support.

if i'd give it a shot and refactored the code to support Python 2 and 3, would that be accepted by the maintainers? if so, would using six be okay?

When writing to Openlink Virtuoso (07.20.3217) with the current sparql_protocol writer, a 400 bad request is returned with the following error:

Virtuoso 37000 Error SP031: SPARQL compiler: No plain default graph specified in the preamble, but it is needed for triple constructor in DELETE {...} without GRAPH {...}

Examples of the sparql_protocol feature a default_graph attribute:

store = surf.Store(reader = 'sparql_protocol',
                   endpoint = 'http://dbpedia.org/sparql',
                   default_graph = 'http://dbpedia.org')

However, in the code there seems no sign of usage of this attribute, which is needed for triple stores like Virtuoso.

I will create a pull request for the fix.

The code below throws AttributeError: 'ReaderPlugin' object has no attribute 'log'

import surf
from surf.log import setup_logger, set_logger_level
import logging

setup_logger()
set_logger_level(logging.DEBUG)

store = surf.Store(reader="rdflib",
                   writer="rdflib",
                   rdflib_store="IOMemory")

session = surf.Session(store)

print "Load RDF data"
store.load_triples(source="http://www.w3.org/People/Berners-Lee/card.rdf")
print len(store)

Person = session.get_class(surf.ns.FOAF["Person"])

all_persons = Person.all()

print "Found %d persons in Tim Berners-Lee's FOAF document" % (len(all_persons))
for person in all_persons:
    print person.foaf_name.first

---------------------------------------------------------------------------
AttributeError                            Traceback (most recent call last)
<ipython-input-7-fffd51a6eb68> in <module>()
     20 all_persons = Person.all()
     21 
---> 22 print "Found %d persons in Tim Berners-Lee's FOAF document" % (len(all_persons))
     23 for person in all_persons:
     24     print person.foaf_name.first

/usr/local/lib/python2.7/dist-packages/surf/resource/result_proxy.pyc in __len__(self)
    233         """ Return count of resources in this collection. """
    234 
--> 235         _, get_by_response = self.__execute_get_by()
    236         return len(get_by_response)
    237 

/usr/local/lib/python2.7/dist-packages/surf/resource/result_proxy.pyc in __execute_get_by(self)
    214 
    215             store = self._params['store']
--> 216             self._get_by_response = store.get_by(self.__get_by_args)
    217 
    218         return self.__get_by_args, self._get_by_response

/usr/local/lib/python2.7/dist-packages/surf/store.pyc in get_by(self, params)
    148     def get_by(self, params):
    149         params["context"] = self.__add_default_context(params.get("context"))
--> 150         return self.reader.get_by(params)
    151 
    152     def execute(self, query):

/usr/local/lib/python2.7/dist-packages/surf/plugin/reader.pyc in get_by(self, params)
    165 
    166     def get_by(self, params):
--> 167         return self._get_by(params)
    168 
    169 

/usr/local/lib/python2.7/dist-packages/surf/plugin/query_reader.pyc in _get_by(self, params)
    249 
    250         # Load just subjects and their types
--> 251         table = self._to_table(self._execute(query))
    252 
    253         # Create response structure, preserve order, don't include

/usr/local/lib/python2.7/dist-packages/surf.rdflib-1.0.0_r338-py2.7.egg/surf_rdflib/reader.pyc in _execute(self, query)
     78     def _execute(self, query):
     79         q_string = unicode(query)
---> 80         self.log.debug(q_string)
     81         return self.__graph.query(q_string)
     82 

AttributeError: 'ReaderPlugin' object has no attribute 'log'

pip install fails - unable to find a suitable SuRF package on PyPi:

pip install --upgrade SuRF
Collecting SuRF
  Could not find a version that satisfies the requirement SuRF (from versions: )
No matching distribution found for SuRF

environment: Python 2.7 on Mac OS X 10.9.5 (the error does not seem to be Mac-specific though).

It is not clear to me which licensing is used in this library. Could you please clarify?

Thank you!

graph = store.reader.graph
print graph.serialize(format='n3')

yields something like

@prefix ns2: <http://xmlns.com/foaf/0.1/> .

where ns2 is used instead of the predefined namespace prefix foaf:. tried different serialization formats with no luck.

is it possible to ask surf to use the predefined namespace prefixes in serialization? thanks!

Background: I'm developing a python library that depends on surf, called hucitlib. I use surf to read and modify resources from a SPARQL endpoint.

Problem: the tests for hucitlib work fine with surf==1.1.9, but many of them fail if I upgrade to 1.2.0.

For example I have a bit that looks like the following:

id = Identifier(id_uri)
id.rdfs_label = Literal(urn)
id.ecrm_P2_has_type = Type(BASE_URI_TYPES % "CTS_URN")
id.save()
return True

and it fails now with this exception

E           ValueError: Statement of type [list, tuple] does not
E                                            have all the (s,p,o) members (the length of the
E                                            supplied arguemnt must be at least 3)

The bug seems to be related to surf.plugin.sparql_protocol.writer and surf.plugin.sparql_protocol.writer. It is raised whenever (in my tests) the method save() or is_present() are called. From what I understand it seems related to some of the underlying SPARQL statements being invalid.

in plugins/surf.rdflib/surf_rdflib/reader.py there is an issue related to the JSON serialization of the query result -- I believe this is due to changes in RDFLib 4.x as opposed to 3.x.

Hi @cosminbasca,

reading the documentation I understood how to change the resource's context using surf.

But would it be possible to change its URI programmatically? I need to change several thousands of URIs in a triples store, and I'm trying to figure out what's the best way of doing so.

Thanks in advance for your help!
matteo

As far as I can see in the documentation, SuRF is more focused on retrieving RDF resources and map them into Python objects, right @cosminbasca?

I guess the opposite use case is also somehow supported: serialize Python object to RDF. I wonder if there is any documentation or source example covering that aspect. What do the Python classes need to provide?

For instance, for my tries I know they need to provide some kind of logic in the find() method. I could continue with a fail-and-error approach, but I guess is not very productive...

Thanks in advance.

python 3.7.6 issue error for the Moduel surf.query

class Filter(unicode):
NameError: name 'unicode' is not defined

It will be very helpful, if a work around is provided.