def gms(goals, noMagic, factGraph, ruleSet, defaultBasePreds,
defaultDerivedPreds, network, workingMemory, strictCheck):
magicRuleNo = 0
magicSeeds = []
for goal in goals:
goalSeed = AdornLiteral(goal).makeMagicPred()
print("Magic seed fact (used in bottom-up evaluation)", goalSeed)
magicSeeds.append(goalSeed.toRDFTuple())
if noMagic:
print("Predicates whose magic sets will not be calculated")
for p in noMagic:
print("\t", factGraph.qname(p))
for rule in MagicSetTransformation(
factGraph, ruleSet, goals, derivedPreds=[],
strictCheck=strictCheck,
defaultPredicates=(defaultBasePreds, defaultDerivedPreds),
noMagic=noMagic):
magicRuleNo += 1
network.buildNetworkFromClause(rule)
if len(list(ruleSet)):
print("reduction in size of program: %s (%s -> %s clauses)" % (
100 - (float(magicRuleNo) / float(len(ruleSet))) * 100,
len(list(ruleSet)), magicRuleNo))
start = time.time()
network.feedFactsToAdd(generateTokenSet(magicSeeds))
if not [rule for rule in factGraph.adornedProgram if len(rule.sip)]:
warnings.warn("Using GMS sideways information strategy with no " +
"information to pass from query. Falling back to " +
"naive method over given facts and rules")
network.feedFactsToAdd(workingMemory)
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
print("Time to calculate closure on working memory: ", sTimeStr)
def main():
from optparse import OptionParser
op = OptionParser(
'usage: %prog [options] factFile1 factFile2 ... factFileN')
op.add_option(
'--why',
default=None,
help='Specifies the goals to solve for using the non-naive methods' +
'see --method')
op.add_option(
'--closure',
action='store_true',
default=False,
help='Whether or not to serialize the inferred triples' +
' along with the original triples. Otherwise ' +
'(the default behavior), serialize only the inferred triples')
op.add_option(
'--imports',
action='store_true',
default=False,
help='Whether or not to follow owl:imports in the fact graph')
op.add_option(
'--output',
default='n3',
metavar='RDF_FORMAT',
choices=[
'xml', 'TriX', 'n3', 'pml', 'proof-graph', 'nt', 'rif', 'rif-xml',
'conflict', 'man-owl'
],
help=
"Serialize the inferred triples and/or original RDF triples to STDOUT "
+
"using the specified RDF syntax ('xml', 'pretty-xml', 'nt', 'turtle', "
+
"or 'n3') or to print a summary of the conflict set (from the RETE " +
"network) if the value of this option is 'conflict'. If the the " +
" value is 'rif' or 'rif-xml', Then the rules used for inference " +
"will be serialized as RIF. If the value is 'pml' and --why is used, "
+ " then the PML RDF statements are serialized. If output is " +
"'proof-graph then a graphviz .dot file of the proof graph is printed. "
+
"Finally if the value is 'man-owl', then the RDF facts are assumed " +
"to be OWL/RDF and serialized via Manchester OWL syntax. The default is %default"
)
op.add_option(
'--class',
dest='classes',
action='append',
default=[],
metavar='QNAME',
help='Used with --output=man-owl to determine which ' +
'classes within the entire OWL/RDF are targetted for serialization' +
'. Can be used more than once')
op.add_option(
'--hybrid',
action='store_true',
default=False,
help='Used with with --method=bfp to determine whether or not to ' +
'peek into the fact graph to identify predicates that are both ' +
'derived and base. This is expensive for large fact graphs' +
'and is explicitely not used against SPARQL endpoints')
op.add_option(
'--property',
action='append',
dest='properties',
default=[],
metavar='QNAME',
help='Used with --output=man-owl or --extract to determine which ' +
'properties are serialized / extracted. Can be used more than once')
op.add_option(
'--normalize',
action='store_true',
default=False,
help=
"Used with --output=man-owl to attempt to determine if the ontology is 'normalized' [Rector, A. 2003]"
+ "The default is %default")
op.add_option(
'--ddlGraph',
default=False,
help=
"The location of a N3 Data Description document describing the IDB predicates"
)
op.add_option(
'--input-format',
default='xml',
dest='inputFormat',
metavar='RDF_FORMAT',
choices=['xml', 'trix', 'n3', 'nt', 'rdfa'],
help=
"The format of the RDF document(s) which serve as the initial facts " +
" for the RETE network. One of 'xml', 'n3', 'trix', 'nt', " +
"or 'rdfa'. The default is %default")
op.add_option(
'--safety',
default='none',
metavar='RULE_SAFETY',
choices=['loose', 'strict', 'none'],
help="Determines how to handle RIF Core safety. A value of 'loose' " +
" means that unsafe rules will be ignored. A value of 'strict' " +
" will cause a syntax exception upon any unsafe rule. A value of " +
"'none' (the default) does nothing")
op.add_option(
'--pDSemantics',
action='store_true',
default=False,
help=
'Used with --dlp to add pD semantics ruleset for semantics not covered '
+ 'by DLP but can be expressed in definite Datalog Logic Programming' +
' The default is %default')
op.add_option(
'--stdin',
action='store_true',
default=False,
help=
'Parse STDIN as an RDF graph to contribute to the initial facts. The default is %default '
)
op.add_option(
'--ns',
action='append',
default=[],
metavar="PREFIX=URI",
help='Register a namespace binding (QName prefix to a base URI). This '
+ 'can be used more than once')
op.add_option(
'--rules',
default=[],
action='append',
metavar='PATH_OR_URI',
help='The Notation 3 documents to use as rulesets for the RETE network'
+ '. Can be specified more than once')
op.add_option('-d',
'--debug',
action='store_true',
default=True,
help='Include debugging output')
op.add_option(
'--strictness',
default='defaultBase',
metavar='DDL_STRICTNESS',
choices=['loose', 'defaultBase', 'defaultDerived', 'harsh'],
help=
'Used with --why to specify whether to: *not* check if predicates are '
+
' both derived and base (loose), if they are, mark as derived (defaultDerived) '
+
'or as base (defaultBase) predicates, else raise an exception (harsh)')
op.add_option(
'--method',
default='naive',
metavar='reasoning algorithm',
choices=['gms', 'bfp', 'naive'],
help='Used with --why to specify how to evaluate answers for query. '
+ 'One of: gms, sld, bfp, naive')
op.add_option(
'--firstAnswer',
default=False,
action='store_true',
help=
'Used with --why to determine whether to fetch all answers or just ' +
'the first')
op.add_option(
'--edb',
default=[],
action='append',
metavar='EXTENSIONAL_DB_PREDICATE_QNAME',
help=
'Used with --why/--strictness=defaultDerived to specify which clashing '
+ 'predicate will be designated as a base predicate')
op.add_option(
'--idb',
default=[],
action='append',
metavar='INTENSIONAL_DB_PREDICATE_QNAME',
help=
'Used with --why/--strictness=defaultBase to specify which clashing ' +
'predicate will be designated as a derived predicate')
op.add_option(
'--hybridPredicate',
default=[],
action='append',
metavar='PREDICATE_QNAME',
help=
'Used with --why to explicitely specify a hybrid predicate (in both ' +
' IDB and EDB) ')
op.add_option(
'--noMagic',
default=[],
action='append',
metavar='DB_PREDICATE_QNAME',
help='Used with --why to specify that the predicate shouldnt have its '
+ 'magic sets calculated')
op.add_option(
'--filter',
action='append',
default=[],
metavar='PATH_OR_URI',
help=
'The Notation 3 documents to use as a filter (entailments do not particpate in network)'
)
op.add_option(
'--ruleFacts',
action='store_true',
default=False,
help="Determines whether or not to attempt to parse initial facts from "
+ "the rule graph. The default is %default")
op.add_option(
'--builtins',
default=False,
metavar='PATH_TO_PYTHON_MODULE',
help="The path to a python module with function definitions (and a " +
"dicitonary called ADDITIONAL_FILTERS) to use for builtins implementations"
)
op.add_option(
'--dlp',
action='store_true',
default=False,
help=
'Use Description Logic Programming (DLP) to extract rules from OWL/RDF. The default is %default'
)
op.add_option(
'--sparqlEndpoint',
action='store_true',
default=False,
help=
'Indicates that the sole argument is the URI of a SPARQL endpoint to query'
)
op.add_option(
'--ontology',
action='append',
default=[],
metavar='PATH_OR_URI',
help=
'The path to an OWL RDF/XML graph to use DLP to extract rules from ' +
'(other wise, fact graph(s) are used) ')
op.add_option(
'--ontologyFormat',
default='xml',
dest='ontologyFormat',
metavar='RDF_FORMAT',
choices=['xml', 'trix', 'n3', 'nt', 'rdfa'],
help=
"The format of the OWL RDF/XML graph specified via --ontology. The default is %default"
)
op.add_option(
'--builtinTemplates',
default=None,
metavar='N3_DOC_PATH_OR_URI',
help=
'The path to an N3 document associating SPARQL FILTER templates to ' +
'rule builtins')
op.add_option('--negation',
action='store_true',
default=False,
help='Extract negative rules?')
op.add_option(
'--normalForm',
action='store_true',
default=False,
help='Whether or not to reduce DL axioms & LP rules to a normal form')
(options, facts) = op.parse_args()
nsBinds = {'iw': 'http://inferenceweb.stanford.edu/2004/07/iw.owl#'}
for nsBind in options.ns:
pref, nsUri = nsBind.split('=')
nsBinds[pref] = nsUri
namespace_manager = NamespaceManager(Graph())
if options.sparqlEndpoint:
factGraph = Graph(plugin.get('SPARQLStore', Store)(facts[0]))
options.hybrid = False
else:
factGraph = Graph()
ruleSet = Ruleset()
for fileN in options.rules:
if options.ruleFacts and not options.sparqlEndpoint:
factGraph.parse(fileN, format='n3')
print("Parsing RDF facts from ", fileN)
if options.builtins:
import imp
userFuncs = imp.load_source('builtins', options.builtins)
rs = HornFromN3(fileN,
additionalBuiltins=userFuncs.ADDITIONAL_FILTERS)
else:
rs = HornFromN3(fileN)
nsBinds.update(rs.nsMapping)
ruleSet.formulae.extend(rs)
#ruleGraph.parse(fileN, format='n3')
ruleSet.nsMapping = nsBinds
for prefix, uri in list(nsBinds.items()):
namespace_manager.bind(prefix, uri, override=False)
closureDeltaGraph = Graph()
closureDeltaGraph.namespace_manager = namespace_manager
factGraph.namespace_manager = namespace_manager
if not options.sparqlEndpoint:
for fileN in facts:
factGraph.parse(fileN, format=options.inputFormat)
if options.imports:
for owlImport in factGraph.objects(predicate=OWL_NS.imports):
factGraph.parse(owlImport)
print("Parsed Semantic Web Graph.. ", owlImport)
if not options.sparqlEndpoint and facts:
for pref, uri in factGraph.namespaces():
nsBinds[pref] = uri
if options.stdin:
assert not options.sparqlEndpoint, "Cannot use --stdin with --sparqlEndpoint"
factGraph.parse(sys.stdin, format=options.inputFormat)
#Normalize namespace mappings
#prune redundant, rdflib-allocated namespace prefix mappings
newNsMgr = NamespaceManager(factGraph)
from FuXi.Rete.Util import CollapseDictionary
for k, v in list(
CollapseDictionary(
dict([(k, v) for k, v in factGraph.namespaces()])).items()):
newNsMgr.bind(k, v)
factGraph.namespace_manager = newNsMgr
if options.normalForm:
NormalFormReduction(factGraph)
if not options.sparqlEndpoint:
workingMemory = generateTokenSet(factGraph)
if options.builtins:
import imp
userFuncs = imp.load_source('builtins', options.builtins)
rule_store, rule_graph, network = SetupRuleStore(
makeNetwork=True, additionalBuiltins=userFuncs.ADDITIONAL_FILTERS)
else:
rule_store, rule_graph, network = SetupRuleStore(makeNetwork=True)
network.inferredFacts = closureDeltaGraph
network.nsMap = nsBinds
if options.dlp:
from FuXi.DLP.DLNormalization import NormalFormReduction
if options.ontology:
ontGraph = Graph()
for fileN in options.ontology:
ontGraph.parse(fileN, format=options.ontologyFormat)
for prefix, uri in ontGraph.namespaces():
nsBinds[prefix] = uri
namespace_manager.bind(prefix, uri, override=False)
if options.sparqlEndpoint:
factGraph.store.bind(prefix, uri)
else:
ontGraph = factGraph
NormalFormReduction(ontGraph)
dlp = network.setupDescriptionLogicProgramming(
ontGraph,
addPDSemantics=options.pDSemantics,
constructNetwork=False,
ignoreNegativeStratus=options.negation,
safety=safetyNameMap[options.safety])
ruleSet.formulae.extend(dlp)
if options.output == 'rif' and not options.why:
for rule in ruleSet:
print(rule)
if options.negation:
for nRule in network.negRules:
print(nRule)
elif options.output == 'man-owl':
cGraph = network.closureGraph(factGraph, readOnly=False)
cGraph.namespace_manager = namespace_manager
Individual.factoryGraph = cGraph
if options.classes:
mapping = dict(namespace_manager.namespaces())
for c in options.classes:
pref, uri = c.split(':')
print(Class(URIRef(mapping[pref] + uri)).__repr__(True))
elif options.properties:
mapping = dict(namespace_manager.namespaces())
for p in options.properties:
pref, uri = p.split(':')
print(Property(URIRef(mapping[pref] + uri)))
else:
for p in AllProperties(cGraph):
print(p.identifier, first(p.label))
print(repr(p))
for c in AllClasses(cGraph):
if options.normalize:
if c.isPrimitive():
primAnc = [
sc for sc in c.subClassOf if sc.isPrimitive()
]
if len(primAnc) > 1:
warnings.warn(
"Branches of primitive skeleton taxonomy" +
" should form trees: %s has %s primitive parents: %s"
% (c.qname, len(primAnc), primAnc),
UserWarning, 1)
children = [desc for desc in c.subSumpteeIds()]
for child in children:
for otherChild in [
o for o in children if o is not child
]:
if not otherChild in [
c.identifier
for c in Class(child).disjointWith
]: # and \
warnings.warn(
"Primitive children (of %s) " % (c.qname) + \
"must be mutually disjoint: %s and %s" % (
Class(child).qname, Class(otherChild).qname), UserWarning, 1)
# if not isinstance(c.identifier, BNode):
print(c.__repr__(True))
if not options.why:
# Naive construction of graph
for rule in ruleSet:
network.buildNetworkFromClause(rule)
magicSeeds = []
if options.why:
builtinTemplateGraph = Graph()
if options.builtinTemplates:
builtinTemplateGraph = Graph().parse(options.builtinTemplates,
format='n3')
factGraph.templateMap = \
dict([(pred, template)
for pred, _ignore, template in
builtinTemplateGraph.triples(
(None,
TEMPLATES.filterTemplate,
None))])
goals = []
query = ParseSPARQL(options.why)
network.nsMap['pml'] = PML
network.nsMap['gmp'] = GMP_NS
network.nsMap['owl'] = OWL_NS
nsBinds.update(network.nsMap)
network.nsMap = nsBinds
if not query.prologue:
query.prologue = Prologue(None, [])
query.prologue.prefixBindings.update(nsBinds)
else:
for prefix, nsInst in list(nsBinds.items()):
if prefix not in query.prologue.prefixBindings:
query.prologue.prefixBindings[prefix] = nsInst
print("query.prologue", query.prologue)
print("query.query", query.query)
print("query.query.whereClause", query.query.whereClause)
print("query.query.whereClause.parsedGraphPattern",
query.query.whereClause.parsedGraphPattern)
goals.extend([(s, p, o) for s, p, o, c in ReduceGraphPattern(
query.query.whereClause.parsedGraphPattern,
query.prologue).patterns])
# dPreds=[]# p for s, p, o in goals ]
# print("goals", goals)
magicRuleNo = 0
bottomUpDerivedPreds = []
# topDownDerivedPreds = []
defaultBasePreds = []
defaultDerivedPreds = set()
hybridPredicates = []
mapping = dict(newNsMgr.namespaces())
for edb in options.edb:
pref, uri = edb.split(':')
defaultBasePreds.append(URIRef(mapping[pref] + uri))
noMagic = []
for pred in options.noMagic:
pref, uri = pred.split(':')
noMagic.append(URIRef(mapping[pref] + uri))
if options.ddlGraph:
ddlGraph = Graph().parse(options.ddlGraph, format='n3')
# @TODO: should also get hybrid predicates from DDL graph
defaultDerivedPreds = IdentifyDerivedPredicates(
ddlGraph, Graph(), ruleSet)
else:
for idb in options.idb:
pref, uri = idb.split(':')
defaultDerivedPreds.add(URIRef(mapping[pref] + uri))
defaultDerivedPreds.update(
set([p == RDF.type and o or p for s, p, o in goals]))
for hybrid in options.hybridPredicate:
pref, uri = hybrid.split(':')
hybridPredicates.append(URIRef(mapping[pref] + uri))
if options.method == 'gms':
for goal in goals:
goalSeed = AdornLiteral(goal).makeMagicPred()
print("Magic seed fact (used in bottom-up evaluation)",
goalSeed)
magicSeeds.append(goalSeed.toRDFTuple())
if noMagic:
print("Predicates whose magic sets will not be calculated")
for p in noMagic:
print("\t", factGraph.qname(p))
for rule in MagicSetTransformation(
factGraph,
ruleSet,
goals,
derivedPreds=bottomUpDerivedPreds,
strictCheck=nameMap[options.strictness],
defaultPredicates=(defaultBasePreds, defaultDerivedPreds),
noMagic=noMagic):
magicRuleNo += 1
network.buildNetworkFromClause(rule)
if len(list(ruleSet)):
print("reduction in size of program: %s (%s -> %s clauses)" %
(100 -
(float(magicRuleNo) / float(len(list(ruleSet)))) * 100,
len(list(ruleSet)), magicRuleNo))
start = time.time()
network.feedFactsToAdd(generateTokenSet(magicSeeds))
if not [
rule for rule in factGraph.adornedProgram if len(rule.sip)
]:
warnings.warn(
"Using GMS sideways information strategy with no " +
"information to pass from query. Falling back to " +
"naive method over given facts and rules")
network.feedFactsToAdd(workingMemory)
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
print("Time to calculate closure on working memory: ", sTimeStr)
if options.output == 'rif':
print("Rules used for bottom-up evaluation")
if network.rules:
for clause in network.rules:
print(clause)
else:
for clause in factGraph.adornedProgram:
print(clause)
if options.output == 'conflict':
network.reportConflictSet()
elif options.method == 'bfp':
topDownDPreds = defaultDerivedPreds
if options.builtinTemplates:
builtinTemplateGraph = Graph().parse(options.builtinTemplates,
format='n3')
builtinDict = dict([
(pred, template) for pred, _ignore, template in
builtinTemplateGraph.triples((None,
TEMPLATES.filterTemplate,
None))
])
else:
builtinDict = None
topDownStore = TopDownSPARQLEntailingStore(
factGraph.store,
factGraph,
idb=ruleSet,
DEBUG=options.debug,
derivedPredicates=topDownDPreds,
templateMap=builtinDict,
nsBindings=network.nsMap,
identifyHybridPredicates=options.hybrid
if options.method == 'bfp' else False,
hybridPredicates=hybridPredicates)
targetGraph = Graph(topDownStore)
for pref, nsUri in list(network.nsMap.items()):
targetGraph.bind(pref, nsUri)
start = time.time()
# queryLiteral = EDBQuery([BuildUnitermFromTuple(goal) for goal in goals],
# targetGraph)
# query = queryLiteral.asSPARQL()
# print("Goal to solve ", query)
sTime = time.time() - start
result = targetGraph.query(options.why, initNs=network.nsMap)
if result.askAnswer:
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
print("Time to reach answer ground goal answer of %s: %s" %
(result.askAnswer[0], sTimeStr))
else:
for rt in result:
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
if options.firstAnswer:
break
print(
"Time to reach answer %s via top-down SPARQL sip strategy: %s"
% (rt, sTimeStr))
if options.output == 'conflict' and options.method == 'bfp':
for _network, _goal in topDownStore.queryNetworks:
print(network, _goal)
_network.reportConflictSet(options.debug)
for query in topDownStore.edbQueries:
print(query.asSPARQL())
elif options.method == 'naive':
start = time.time()
network.feedFactsToAdd(workingMemory)
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
print("Time to calculate closure on working memory: ", sTimeStr)
print(network)
if options.output == 'conflict':
network.reportConflictSet()
for fileN in options.filter:
for rule in HornFromN3(fileN):
network.buildFilterNetworkFromClause(rule)
if options.negation and network.negRules and options.method in [
'both', 'bottomUp'
]:
now = time.time()
rt = network.calculateStratifiedModel(factGraph)
print(
"Time to calculate stratified, stable model (inferred %s facts): %s"
% (rt, time.time() - now))
if options.filter:
print("Applying filter to entailed facts")
network.inferredFacts = network.filteredFacts
if options.closure and options.output in RDF_SERIALIZATION_FORMATS:
cGraph = network.closureGraph(factGraph)
cGraph.namespace_manager = namespace_manager
print(
cGraph.serialize(destination=None,
format=options.output,
base=None))
elif options.output and options.output in RDF_SERIALIZATION_FORMATS:
print(
network.inferredFacts.serialize(destination=None,
format=options.output,
base=None))
def main():
from optparse import OptionParser
op = OptionParser(
'usage: %prog [options] factFile1 factFile2 ... factFileN')
op.add_option('--why',
default=None,
help='Specifies the goals to solve for using the non-niave methods' +
'see --method')
op.add_option('--closure',
action='store_true',
default=False,
help='Whether or not to serialize the inferred triples' +
' along with the original triples. Otherwise ' +
'(the default behavior), serialize only the inferred triples')
op.add_option('--imports',
action='store_true',
default=False,
help='Whether or not to follow owl:imports in the fact graph')
op.add_option('--output',
default='n3',
metavar='RDF_FORMAT',
choices=['xml',
'TriX',
'n3',
'pml',
'proof-graph',
'nt',
'rif',
'rif-xml',
'conflict',
'man-owl'],
help="Serialize the inferred triples and/or original RDF triples to STDOUT "+
"using the specified RDF syntax ('xml','pretty-xml','nt','turtle', "+
"or 'n3') or to print a summary of the conflict set (from the RETE "+
"network) if the value of this option is 'conflict'. If the the "+
" value is 'rif' or 'rif-xml', Then the rules used for inference "+
"will be serialized as RIF. If the value is 'pml' and --why is used, "+
" then the PML RDF statements are serialized. If output is "+
"'proof-graph then a graphviz .dot file of the proof graph is printed. "+
"Finally if the value is 'man-owl', then the RDF facts are assumed "+
"to be OWL/RDF and serialized via Manchester OWL syntax. The default is %default")
op.add_option('--class',
dest='classes',
action='append',
default=[],
metavar='QNAME',
help='Used with --output=man-owl to determine which '+
'classes within the entire OWL/RDF are targetted for serialization'+
'. Can be used more than once')
op.add_option('--hybrid',
action='store_true',
default=False,
help='Used with with --method=bfp to determine whether or not to '+
'peek into the fact graph to identify predicates that are both '+
'derived and base. This is expensive for large fact graphs'+
'and is explicitely not used against SPARQL endpoints')
op.add_option('--property',
action='append',
dest='properties',
default=[],
metavar='QNAME',
help='Used with --output=man-owl or --extract to determine which '+
'properties are serialized / extracted. Can be used more than once')
op.add_option('--normalize',
action='store_true',
default=False,
help="Used with --output=man-owl to attempt to determine if the ontology is 'normalized' [Rector, A. 2003]"+
"The default is %default")
op.add_option('--ddlGraph',
default=False,
help="The location of a N3 Data Description document describing the IDB predicates")
op.add_option('--input-format',
default='xml',
dest='inputFormat',
metavar='RDF_FORMAT',
choices=['xml', 'trix', 'n3', 'nt', 'rdfa'],
help="The format of the RDF document(s) which serve as the initial facts "+
" for the RETE network. One of 'xml','n3','trix', 'nt', "+
"or 'rdfa'. The default is %default")
op.add_option('--safety',
default='none',
metavar='RULE_SAFETY',
choices=['loose', 'strict', 'none'],
help="Determines how to handle RIF Core safety. A value of 'loose' "+
" means that unsafe rules will be ignored. A value of 'strict' "+
" will cause a syntax exception upon any unsafe rule. A value of "+
"'none' (the default) does nothing")
op.add_option('--pDSemantics',
action='store_true',
default=False,
help='Used with --dlp to add pD semantics ruleset for semantics not covered '+
'by DLP but can be expressed in definite Datalog Logic Programming'+
' The default is %default')
op.add_option('--stdin',
action='store_true',
default=False,
help='Parse STDIN as an RDF graph to contribute to the initial facts. The default is %default ')
op.add_option('--ns',
action='append',
default=[],
metavar="PREFIX=URI",
help='Register a namespace binding (QName prefix to a base URI). This '+
'can be used more than once')
op.add_option('--rules',
default=[],
action='append',
metavar='PATH_OR_URI',
help='The Notation 3 documents to use as rulesets for the RETE network'+
'. Can be specified more than once')
op.add_option('-d', '--debug', action='store_true', default=False,
help='Include debugging output')
op.add_option('--strictness',
default='defaultBase',
metavar='DDL_STRICTNESS',
choices=['loose',
'defaultBase',
'defaultDerived',
'harsh'],
help='Used with --why to specify whether to: *not* check if predicates are '+
' both derived and base (loose), if they are, mark as derived (defaultDerived) '+
'or as base (defaultBase) predicates, else raise an exception (harsh)')
op.add_option('--method',
default='naive',
metavar='reasoning algorithm',
choices=['gms', 'bfp', 'naive'],
help='Used with --why to specify how to evaluate answers for query. '+
'One of: gms,bfp,naive')
op.add_option('--firstAnswer',
default=False,
action='store_true',
help='Used with --why to determine whether to fetch all answers or just '+
'the first')
op.add_option('--edb',
default=[],
action='append',
metavar='EXTENSIONAL_DB_PREDICATE_QNAME',
help='Used with --why/--strictness=defaultDerived to specify which clashing '+
'predicate will be designated as a base predicate')
op.add_option('--idb',
default=[],
action='append',
metavar='INTENSIONAL_DB_PREDICATE_QNAME',
help='Used with --why/--strictness=defaultBase to specify which clashing '+
'predicate will be designated as a derived predicate')
op.add_option('--hybridPredicate',
default=[],
action='append',
metavar='PREDICATE_QNAME',
help='Used with --why to explicitely specify a hybrid predicate (in both '+
' IDB and EDB) ')
op.add_option('--noMagic',
default=[],
action='append',
metavar='DB_PREDICATE_QNAME',
help='Used with --why to specify that the predicate shouldnt have its '+
'magic sets calculated')
op.add_option('--filter',
action='append',
default=[],
metavar='PATH_OR_URI',
help='The Notation 3 documents to use as a filter (entailments do not particpate in network)')
op.add_option('--ruleFacts',
action='store_true',
default=False,
help="Determines whether or not to attempt to parse initial facts from "+
"the rule graph. The default is %default")
op.add_option('--builtins',
default=False,
metavar='PATH_TO_PYTHON_MODULE',
help="The path to a python module with function definitions (and a "+
"dicitonary called ADDITIONAL_FILTERS) to use for builtins implementations")
op.add_option('--dlp',
action='store_true',
default=False,
help='Use Description Logic Programming (DLP) to extract rules from OWL/RDF. The default is %default')
op.add_option('--sparqlEndpoint',
action='store_true',
default=False,
help='Indicates that the sole argument is the URI of a SPARQL endpoint to query')
op.add_option('--ontology',
action='append',
default=[],
metavar='PATH_OR_URI',
help='The path to an OWL RDF/XML graph to use DLP to extract rules from '+
'(other wise, fact graph(s) are used) ')
op.add_option('--ruleFormat',
default='n3',
dest='ruleFormat',
metavar='RULE_FORMAT',
choices=['n3', 'rif'],
help="The format of the rules to parse ('n3', 'rif'). The default is %default")
op.add_option('--ontologyFormat',
default='xml',
dest='ontologyFormat',
metavar='RDF_FORMAT',
choices=['xml', 'trix', 'n3', 'nt', 'rdfa'],
help="The format of the OWL RDF/XML graph specified via --ontology. The default is %default")
op.add_option('--builtinTemplates',
default=None,
metavar='N3_DOC_PATH_OR_URI',
help='The path to an N3 document associating SPARQL FILTER templates to '+
'rule builtins')
op.add_option('--negation',
action='store_true',
default=False,
help='Extract negative rules?')
op.add_option('--normalForm',
action='store_true',
default=False,
help='Whether or not to reduce DL axioms & LP rules to a normal form')
(options, facts) = op.parse_args()
nsBinds = {'iw': 'http://inferenceweb.stanford.edu/2004/07/iw.owl#'}
for nsBind in options.ns:
pref, nsUri = nsBind.split('=')
nsBinds[pref]=nsUri
namespace_manager = NamespaceManager(Graph())
if options.sparqlEndpoint:
factGraph = Graph(plugin.get('SPARQL', Store)(facts[0]))
options.hybrid = False
else:
factGraph = Graph()
ruleSet = Ruleset()
for fileN in options.rules:
if options.ruleFacts and not options.sparqlEndpoint:
factGraph.parse(fileN, format='n3')
print("Parsing RDF facts from %s" % fileN)
if options.builtins:
import imp
userFuncs = imp.load_source('builtins', options.builtins)
rs = HornFromN3(fileN,
additionalBuiltins=userFuncs.ADDITIONAL_FILTERS)
nsBinds.update(rs.nsMapping)
elif options.ruleFormat == 'rif':
try:
from FuXi.Horn.RIFCore import RIFCoreParser
rif_parser = RIFCoreParser(location=fileN, debug=options.debug)
rs = rif_parser.getRuleset()
except ImportError:
raise Exception(
"Missing 3rd party libraries for RIF processing"
)
else:
rs = HornFromN3(fileN)
nsBinds.update(rs.nsMapping)
ruleSet.formulae.extend(rs)
#ruleGraph.parse(fileN,format='n3')
ruleSet.nsMapping = nsBinds
for prefix, uri in list(nsBinds.items()):
namespace_manager.bind(prefix, uri, override=False)
closureDeltaGraph = Graph()
closureDeltaGraph.namespace_manager = namespace_manager
factGraph.namespace_manager = namespace_manager
if not options.sparqlEndpoint:
for fileN in facts:
factGraph.parse(fileN, format=options.inputFormat)
if options.imports:
for owlImport in factGraph.objects(predicate=OWL_NS.imports):
factGraph.parse(owlImport)
print("Parsed Semantic Web Graph.. %s" % owlImport)
if not options.sparqlEndpoint and facts:
for pref, uri in factGraph.namespaces():
nsBinds[pref]=uri
if options.stdin:
assert not options.sparqlEndpoint, "Cannot use --stdin with --sparqlEndpoint"
factGraph.parse(sys.stdin, format=options.inputFormat)
#Normalize namespace mappings
#prune redundant, rdflib-allocated namespace prefix mappings
newNsMgr = NamespaceManager(factGraph)
from FuXi.Rete.Util import CollapseDictionary
for k, v in list(CollapseDictionary(dict([(k, v)
for k, v in factGraph.namespaces()])).items()):
newNsMgr.bind(k, v)
factGraph.namespace_manager = newNsMgr
if options.normalForm:
NormalFormReduction(factGraph)
if not options.sparqlEndpoint:
workingMemory = generateTokenSet(factGraph)
if options.builtins:
import imp
userFuncs = imp.load_source('builtins', options.builtins)
rule_store, rule_graph, network = SetupRuleStore(
makeNetwork=True,
additionalBuiltins=userFuncs.ADDITIONAL_FILTERS)
else:
rule_store, rule_graph, network = SetupRuleStore(makeNetwork=True)
network.inferredFacts = closureDeltaGraph
network.nsMap = nsBinds
if options.dlp:
from FuXi.DLP.DLNormalization import NormalFormReduction
if options.ontology:
ontGraph = Graph()
for fileN in options.ontology:
ontGraph.parse(fileN, format=options.ontologyFormat)
for prefix, uri in ontGraph.namespaces():
nsBinds[prefix] = uri
namespace_manager.bind(prefix, uri, override=False)
if options.sparqlEndpoint:
factGraph.store.bind(prefix, uri)
else:
ontGraph=factGraph
NormalFormReduction(ontGraph)
dlp=network.setupDescriptionLogicProgramming(
ontGraph,
addPDSemantics=options.pDSemantics,
constructNetwork=False,
ignoreNegativeStratus=options.negation,
safety=safetyNameMap[options.safety])
ruleSet.formulae.extend(dlp)
if options.output == 'rif' and not options.why:
for rule in ruleSet:
print(rule)
if options.negation:
for nRule in network.negRules:
print(nRule)
elif options.output == 'man-owl':
cGraph = network.closureGraph(factGraph, readOnly=False)
cGraph.namespace_manager = namespace_manager
Individual.factoryGraph = cGraph
if options.classes:
mapping = dict(namespace_manager.namespaces())
for c in options.classes:
pref, uri = c.split(':')
print(Class(URIRef(mapping[pref] + uri)).__repr__(True))
elif options.properties:
mapping = dict(namespace_manager.namespaces())
for p in options.properties:
pref, uri = p.split(':')
print(Property(URIRef(mapping[pref] + uri)))
else:
for p in AllProperties(cGraph):
print(p.identifier, first(p.label))
print(repr(p))
for c in AllClasses(cGraph):
if options.normalize:
if c.isPrimitive():
primAnc = [sc for sc in c.subClassOf if sc.isPrimitive()]
if len(primAnc) > 1:
warnings.warn("Branches of primitive skeleton taxonomy" +
" should form trees: %s has %s primitive parents: %s" % (
c.qname, len(primAnc), primAnc), UserWarning, 1)
children = [desc for desc in c.subSumpteeIds()]
for child in children:
for otherChild in [o for o in children if o is not child]:
if not otherChild in [c.identifier
for c in Class(child).disjointWith]: # and\
warnings.warn("Primitive children (of %s) " % (c.qname) +
"must be mutually disjoint: %s and %s" % (
Class(child).qname,
Class(otherChild).qname), UserWarning, 1)
# if not isinstance(c.identifier,BNode):
print(c.__repr__(True))
if not options.why:
#Naive construction of graph
for rule in ruleSet:
network.buildNetworkFromClause(rule)
magicSeeds=[]
if options.why:
builtinTemplateGraph = Graph()
if options.builtinTemplates:
builtinTemplateGraph = Graph().parse(options.builtinTemplates,
format='n3')
factGraph.templateMap = \
dict([(pred, template)
for pred, _ignore, template in
builtinTemplateGraph.triples(
(None,
TEMPLATES.filterTemplate,
None))])
goals = []
query = ParseSPARQL(options.why)
network.nsMap['pml'] = PML
network.nsMap['gmp'] = GMP_NS
network.nsMap['owl'] = OWL_NS
nsBinds.update(network.nsMap)
network.nsMap = nsBinds
if not query.prolog:
query.prolog = Prolog(None, [])
query.prolog.prefixBindings.update(nsBinds)
else:
for prefix, nsInst in list(nsBinds.items()):
if prefix not in query.prolog.prefixBindings:
query.prolog.prefixBindings[prefix] = nsInst
goals.extend([(s, p, o) for s, p, o, c in ReduceGraphPattern(
query.query.whereClause.parsedGraphPattern,
query.prolog).patterns])
# dPreds=[]# p for s,p,o in goals ]
magicRuleNo = 0
bottomUpDerivedPreds = []
# topDownDerivedPreds = []
defaultBasePreds = []
defaultDerivedPreds = set()
hybridPredicates = []
mapping = dict(newNsMgr.namespaces())
for edb in options.edb:
pref, uri = edb.split(':')
defaultBasePreds.append(URIRef(mapping[pref] + uri))
noMagic = []
for pred in options.noMagic:
pref, uri = pred.split(':')
noMagic.append(URIRef(mapping[pref] + uri))
if options.ddlGraph:
ddlGraph = Graph().parse(options.ddlGraph, format='n3')
# @TODO: should also get hybrid predicates from DDL graph
defaultDerivedPreds=IdentifyDerivedPredicates(
ddlGraph,
Graph(),
ruleSet)
else:
for idb in options.idb:
pref, uri = idb.split(':')
defaultDerivedPreds.add(URIRef(mapping[pref] + uri))
defaultDerivedPreds.update(
set([p == RDF.type and o or p for s, p, o in goals]))
for hybrid in options.hybridPredicate:
pref, uri = hybrid.split(':')
hybridPredicates.append(URIRef(mapping[pref]+uri))
if options.method == 'gms':
for goal in goals:
goalSeed=AdornLiteral(goal).makeMagicPred()
print("Magic seed fact (used in bottom-up evaluation) %s" % goalSeed)
magicSeeds.append(goalSeed.toRDFTuple())
if noMagic:
print("Predicates whose magic sets will not be calculated")
for p in noMagic:
print("\t%s" % factGraph.qname(p))
for rule in MagicSetTransformation(
factGraph,
ruleSet,
goals,
derivedPreds=bottomUpDerivedPreds,
strictCheck=nameMap[options.strictness],
defaultPredicates=(defaultBasePreds,
defaultDerivedPreds),
noMagic=noMagic):
magicRuleNo+=1
network.buildNetworkFromClause(rule)
if len(list(ruleSet)):
print("reduction in size of program: %s (%s -> %s clauses)" % (
100 - (float(magicRuleNo) /
float(len(list(ruleSet)))
) * 100,
len(list(ruleSet)),
magicRuleNo))
start = time.time()
network.feedFactsToAdd(generateTokenSet(magicSeeds))
if not [
rule for rule in factGraph.adornedProgram if len(rule.sip)]:
warnings.warn(
"Using GMS sideways information strategy with no "+
"information to pass from query. Falling back to "+
"naive method over given facts and rules")
network.feedFactsToAdd(workingMemory)
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds"%sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds"%sTime
print("Time to calculate closure on working memory: %s" % sTimeStr)
if options.output == 'rif':
print("Rules used for bottom-up evaluation")
if network.rules:
for clause in network.rules:
print(clause)
else:
for clause in factGraph.adornedProgram:
print(clause)
if options.output == 'conflict':
network.reportConflictSet()
elif options.method == 'bfp':
topDownDPreds = defaultDerivedPreds
if options.builtinTemplates:
builtinTemplateGraph = Graph().parse(options.builtinTemplates,
format='n3')
builtinDict = dict([(pred, template)
for pred, _ignore, template in
builtinTemplateGraph.triples(
(None,
TEMPLATES.filterTemplate,
None))])
else:
builtinDict = None
topDownStore=TopDownSPARQLEntailingStore(
factGraph.store,
factGraph,
idb=ruleSet,
DEBUG=options.debug,
derivedPredicates=topDownDPreds,
templateMap=builtinDict,
nsBindings=network.nsMap,
identifyHybridPredicates=options.hybrid \
if options.method == 'bfp' else False,
hybridPredicates=hybridPredicates)
targetGraph = Graph(topDownStore)
for pref, nsUri in list(network.nsMap.items()):
targetGraph.bind(pref, nsUri)
start = time.time()
# queryLiteral = EDBQuery([BuildUnitermFromTuple(goal)
# for goal in goals],
# targetGraph)
# query = queryLiteral.asSPARQL()
# print >>sys.stderr, "Goal to solve ", query
sTime = time.time() - start
result = targetGraph.query(options.why, initNs=network.nsMap)
if result.askAnswer:
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds"%sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds"%sTime
print("Time to reach answer ground goal answer of %s: %s" % (
result.askAnswer[0], sTimeStr))
else:
for rt in result:
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
if options.firstAnswer:
break
print(
"Time to reach answer %s via top-down SPARQL sip strategy: %s" % (
rt, sTimeStr))
if options.output == 'conflict' and options.method == 'bfp':
for _network, _goal in topDownStore.queryNetworks:
print(_network, _goal)
_network.reportConflictSet(options.debug)
for query in topDownStore.edbQueries:
print(query.asSPARQL())
elif options.method == 'naive':
start = time.time()
network.feedFactsToAdd(workingMemory)
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds"%sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds"%sTime
print("Time to calculate closure on working memory: %s" % sTimeStr)
print(network)
if options.output == 'conflict':
network.reportConflictSet()
for fileN in options.filter:
for rule in HornFromN3(fileN):
network.buildFilterNetworkFromClause(rule)
if options.negation and network.negRules and options.method in ['both',
'bottomUp']:
now=time.time()
rt=network.calculateStratifiedModel(factGraph)
print("Time to calculate stratified, stable model (inferred %s facts): %s" % (
rt,
time.time()-now))
if options.filter:
print("Applying filter to entailed facts")
network.inferredFacts = network.filteredFacts
if options.closure \
and options.output in RDF_SERIALIZATION_FORMATS:
cGraph = network.closureGraph(factGraph)
cGraph.namespace_manager = namespace_manager
print(cGraph.serialize(destination=None,
format=options.output,
base=None))
elif options.output and options.output in RDF_SERIALIZATION_FORMATS:
print(network.inferredFacts.serialize(destination=None,
format=options.output,
base=None))
def MagicOWLProof(self, goals, rules, factGraph, conclusionFile):
progLen = len(rules)
magicRuleNo = 0
dPreds = []
for rule in AdditionalRules(factGraph):
rules.append(rule)
if not GROUND_QUERY and REASONING_STRATEGY != 'gms':
goalDict = dict([((Variable('SUBJECT'),goalP,goalO),goalS)
for goalS,goalP,goalO in goals])
goals = goalDict.keys()
assert goals
if REASONING_STRATEGY == 'gms':
for rule in MagicSetTransformation(factGraph,
rules,
goals,
dPreds):
magicRuleNo += 1
self.network.buildNetworkFromClause(rule)
self.network.rules.add(rule)
if DEBUG:
print("\t%s" % rule)
print("rate of reduction in the size of the program: %s " % (
100 - (float(magicRuleNo) / float(progLen)) * 100))
if REASONING_STRATEGY in ['bfp', 'sld']: # and not GROUND_QUERY:
reasoningAlg = TOP_DOWN_METHOD if REASONING_STRATEGY == 'sld' \
else BFP_METHOD
topDownStore = TopDownSPARQLEntailingStore(
factGraph.store,
factGraph,
idb=rules,
DEBUG=DEBUG,
nsBindings=nsMap,
decisionProcedure=reasoningAlg,
identifyHybridPredicates=REASONING_STRATEGY == 'bfp')
targetGraph = Graph(topDownStore)
for pref,nsUri in nsMap.items():
targetGraph.bind(pref,nsUri)
start = time.time()
for goal in goals:
queryLiteral = EDBQuery([BuildUnitermFromTuple(goal)],
factGraph,
None if GROUND_QUERY else [goal[0]])
query = queryLiteral.asSPARQL()
print "Goal to solve ", query
rt = targetGraph.query(query, initNs=nsMap)
if GROUND_QUERY:
self.failUnless(rt.askAnswer[0],"Failed top-down problem")
else:
if (goalDict[goal]) not in rt or DEBUG:
for network,_goal in topDownStore.queryNetworks:
print network,_goal
network.reportConflictSet(True)
for query in topDownStore.edbQueries:
print query.asSPARQL()
self.failUnless((goalDict[goal]) in rt,
"Failed top-down problem")
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s milli seconds" % sTime
return sTimeStr
elif REASONING_STRATEGY == 'gms':
for goal in goals:
adornedGoalSeed = AdornLiteral(goal).makeMagicPred()
goal = adornedGoalSeed.toRDFTuple()
if DEBUG:
print("Magic seed fact %s" % adornedGoalSeed)
factGraph.add(goal)
timing = self.calculateEntailments(factGraph)
for goal in goals:
# self.failUnless(goal in self.network.inferredFacts or goal in factGraph,
# "Failed GMS query")
if goal not in self.network.inferredFacts and goal not in factGraph:
print("missing triple %s" % (pformat(goal)))
pprint(list(factGraph.adornedProgram))
# from FuXi.Rete.Util import renderNetwork
# dot=renderNetwork(self.network,self.network.nsMap).write_jpeg('test-fail.jpeg')
self.network.reportConflictSet(True)
raise # Exception ("Failed test: "+feature)
else:
print("=== Passed! ===")
return timing
def MagicOWLProof(self, goals, rules, factGraph, conclusionFile):
progLen = len(rules)
magicRuleNo = 0
dPreds = []
for rule in AdditionalRules(factGraph):
rules.append(rule)
if not GROUND_QUERY and REASONING_STRATEGY != 'gms':
goalDict = dict([((Variable('SUBJECT'), goalP, goalO), goalS)
for goalS, goalP, goalO in goals])
goals = goalDict.keys()
assert goals
if REASONING_STRATEGY == 'gms':
for rule in MagicSetTransformation(factGraph,
rules,
goals,
dPreds):
magicRuleNo += 1
self.network.buildNetworkFromClause(rule)
self.network.rules.add(rule)
if DEBUG:
log.debug("\t", rule)
log.debug("rate of reduction in the size of the program: ",
(100 - (float(magicRuleNo) / float(progLen)) * 100))
if REASONING_STRATEGY in ['bfp', 'sld']: # and not GROUND_QUERY:
reasoningAlg = TOP_DOWN_METHOD if REASONING_STRATEGY == 'sld' \
else BFP_METHOD
topDownStore = TopDownSPARQLEntailingStore(
factGraph.store,
factGraph,
idb=rules,
DEBUG=DEBUG,
nsBindings=nsMap,
decisionProcedure=reasoningAlg,
identifyHybridPredicates=REASONING_STRATEGY == 'bfp')
targetGraph = Graph(topDownStore)
for pref, nsUri in nsMap.items():
targetGraph.bind(pref, nsUri)
start = time.time()
for goal in goals:
queryLiteral = EDBQuery([BuildUnitermFromTuple(goal)],
factGraph,
None if GROUND_QUERY else [goal[0]])
query = queryLiteral.asSPARQL()
log.debug("Goal to solve ", query)
rt = targetGraph.query(query, initNs=nsMap)
if GROUND_QUERY:
self.failUnless(rt.askAnswer[0], "Failed top-down problem")
else:
if (goalDict[goal]) not in rt or DEBUG:
for network, _goal in topDownStore.queryNetworks:
log.debug(network, _goal)
network.reportConflictSet(True)
for query in topDownStore.edbQueries:
log.debug(query.asSPARQL())
self.failUnless((goalDict[goal]) in rt,
"Failed top-down problem")
sTime = time.time() - start
if sTime > 1:
sTimeStr = "%s seconds" % sTime
else:
sTime = sTime * 1000
sTimeStr = "%s ms" % sTime
return sTimeStr
elif REASONING_STRATEGY == 'gms':
for goal in goals:
adornedGoalSeed = AdornLiteral(goal).makeMagicPred()
goal = adornedGoalSeed.toRDFTuple()
if DEBUG:
log.debug("Magic seed fact ", adornedGoalSeed)
factGraph.add(goal)
timing = self.calculateEntailments(factGraph)
for goal in goals:
# self.failUnless(goal in self.network.inferredFacts or goal in factGraph,
# "Failed GMS query")
if goal not in self.network.inferredFacts and goal not in factGraph:
log.debug("missing triple %s" % (pformat(goal)))
# print(list(factGraph.adornedProgram))
# from FuXi.Rete.Util import renderNetwork
# dot = renderNetwork(
# self.network,self.network.nsMap).write_jpeg('test-fail.jpeg')
self.network.reportConflictSet(True)
log.debug("=== Failed: %s ====" % pformat(goal))
else:
log.debug("=== Passed! ===")
return timing
def SipStrategy(query,
sipCollection,
factGraph,
derivedPreds,
bindings={},
processedRules=None,
network=None,
debug=False,
buildProof=False,
memoizeMemory=None,
proofLevel=1):
"""
Accordingly, we define a sip-strategy for computing the answers to a query
expressed using a set of Datalog rules, and a set of sips, one for each
adornment of a rule head, as follows...
Each evaluation uses memoization (via Python decorators) but also relies on well-formed
rewrites for using semi-naive bottom up method over large SPARQL data.
"""
memoizeMemory = memoizeMemory and memoizeMemory or {}
queryLiteral = BuildUnitermFromTuple(query)
processedRules = processedRules and processedRules or set()
if bindings:
#There are bindings. Apply them to the terms in the query
queryLiteral.ground(bindings)
if debug:
print >> sys.stderr, "%sSolving" % ('\t' *
proofLevel), queryLiteral, bindings
#Only consider ground triple pattern isomorphism with matching bindings
goalRDFStatement = queryLiteral.toRDFTuple()
if queryLiteral in memoizeMemory:
if debug:
print >> sys.stderr, "%sReturning previously calculated results for "%\
('\t'*proofLevel), queryLiteral
for answers in memoizeMemory[queryLiteral]:
yield answers
elif AlphaNode(goalRDFStatement).alphaNetworkHash(
True,
skolemTerms=bindings.values()) in\
[AlphaNode(r.toRDFTuple()).alphaNetworkHash(True,
skolemTerms=bindings.values())
for r in processedRules
if AdornLiteral(goalRDFStatement).adornment == \
r.adornment]:
if debug:
print >> sys.stderr, "%s Goal already processed..."%\
('\t'*proofLevel)
else:
isGround = literalIsGround(queryLiteral)
if buildProof:
ns = NodeSet(goalRDFStatement, network=network, identifier=BNode())
else:
ns = None
adornedProgram = factGraph.adornedProgram
queryPred = GetOp(queryLiteral)
if sipCollection is None:
rules = []
else:
#For every rule head matching the query, we invoke the rule,
#thus determining an adornment, and selecting a sip to follow
rules = sipCollection.headToRule.get(queryPred, set())
if None in sipCollection.headToRule:
#If there are second order rules, we add them
#since they are a 'wildcard'
rules.update(sipCollection.headToRule[None])
#maintained list of rules that haven't been processed before and
#match the query
validRules = []
#each subquery contains values for the bound arguments that are passed
#through the sip arcs entering the node corresponding to that literal. For
#each subquery generated, there is a set of answers.
answers = []
variableMapping = {}
#Some TBox queries can be 'joined' together into SPARQL queries against
#'base' predicates via an RDF dataset
#These atomic concept inclusion axioms can be evaluated together
#using a disjunctive operator at the body of a horn clause
#where each item is a query of the form uniPredicate(?X):
#Or( uniPredicate1(?X1), uniPredicate2(?X), uniPredicate3(?X),..)
#In this way massive, conjunctive joins can be 'mediated'
#between the stated facts and the top-down solver
@parameterizedPredicate([i for i in derivedPreds])
def IsAtomicInclusionAxiomRHS(rule, dPreds):
"""
This is an atomic inclusion axiom with
a variable (or bound) RHS: uniPred(?ENTITY)
"""
bodyList = list(iterCondition(rule.formula.body))
body = first(bodyList)
return GetOp(body) not in dPreds and \
len(bodyList) == 1 and \
body.op == RDF.type
atomicInclusionAxioms = list(ifilter(IsAtomicInclusionAxiomRHS, rules))
if atomicInclusionAxioms and len(atomicInclusionAxioms) > 1:
if debug:
print >> sys.stderr, "\tCombining atomic inclusion axioms: "
pprint(atomicInclusionAxioms, sys.stderr)
if buildProof:
factStep = InferenceStep(ns, source='some RDF graph')
ns.steps.append(factStep)
axioms = [rule.formula.body for rule in atomicInclusionAxioms]
#attempt to exaustively apply any available substitutions
#and determine if query if fully ground
vars = [
v for v in GetArgs(queryLiteral, secondOrder=True)
if isinstance(v, Variable)
]
openVars,axioms,_bindings = \
normalizeBindingsAndQuery(vars,
bindings,
axioms)
if openVars:
mappings = {}
#See if we need to do any variable mappings from the query literals
#to the literals in the applicable rules
query, rt = EDBQuery(axioms, factGraph, openVars,
_bindings).evaluate(
debug, symmAtomicInclusion=True)
if buildProof:
factStep.groundQuery = subquery
for ans in rt:
if buildProof:
factStep.bindings.update(ans)
memoizeMemory.setdefault(queryLiteral, set()).add(
(prepMemiozedAns(ans), ns))
yield ans, ns
else:
#All the relevant derivations have been explored and the result
#is a ground query we can directly execute against the facts
if buildProof:
factStep.bindings.update(bindings)
query, rt = EDBQuery(axioms, factGraph, _bindings).evaluate(
debug, symmAtomicInclusion=True)
if buildProof:
factStep.groundQuery = subquery
memoizeMemory.setdefault(queryLiteral, set()).add(
(prepMemiozedAns(rt), ns))
yield rt, ns
rules = ifilter(lambda i: not IsAtomicInclusionAxiomRHS(i), rules)
for rule in rules:
#An exception is the special predicate ph; it is treated as a base
#predicate and the tuples in it are those supplied for qb by unification.
headBindings = getBindingsFromLiteral(goalRDFStatement,
rule.formula.head)
comboBindings = dict([(k, v) for k, v in itertools.chain(
bindings.items(), headBindings.items())])
varMap = rule.formula.head.getVarMapping(queryLiteral)
if headBindings and\
[term for term in rule.formula.head.getDistinguishedVariables(True)
if varMap.get(term,term) not in headBindings]:
continue
subQueryAnswers = []
dontStop = True
projectedBindings = comboBindings.copy()
if debug:
print >> sys.stderr, "%sProcessing rule"%\
('\t'*proofLevel), rule.formula
if debug and sipCollection:
print >> sys.stderr, "Sideways Information Passing (sip) graph for %s: " % queryLiteral
print >> sys.stdout, sipCollection.serialize(format='n3')
for sip in SIPRepresentation(sipCollection):
print >> sys.stderr, sip
try:
#Invoke the rule
if buildProof:
step = InferenceStep(ns, rule.formula)
else:
step = None
for rt,step in\
invokeRule([headBindings],
iter(iterCondition(rule.formula.body)),
rule.sip,
(proofLevel + 1,
memoizeMemory,
sipCollection,
factGraph,
derivedPreds,
processedRules.union([
AdornLiteral(query)])),
step=step,
debug = debug):
if rt:
if isinstance(rt, dict):
#We received a mapping and must rewrite it via
#correlation between the variables in the rule head
#and the variables in the original query (after applying
#bindings)
varMap = rule.formula.head.getVarMapping(
queryLiteral)
if varMap:
rt = MakeImmutableDict(
refactorMapping(varMap, rt))
if buildProof:
step.bindings = rt
else:
if buildProof:
step.bindings = headBindings
validRules.append(rule)
if buildProof:
ns.steps.append(step)
if isGround:
yield True, ns
else:
memoizeMemory.setdefault(queryLiteral, set()).add(
(prepMemiozedAns(rt), ns))
yield rt, ns
except RuleFailure, e:
#Clean up failed antecedents
if buildProof:
if ns in step.antecedents:
step.antecedents.remove(ns)
if not validRules:
#No rules matching, query factGraph for answers
successful = False
if buildProof:
factStep = InferenceStep(ns, source='some RDF graph')
ns.steps.append(factStep)
if not isGround:
subquery, rt = EDBQuery([queryLiteral], factGraph, [
v for v in GetArgs(queryLiteral, secondOrder=True)
if isinstance(v, Variable)
], bindings).evaluate(debug)
if buildProof:
factStep.groundQuery = subquery
for ans in rt:
successful = True
if buildProof:
factStep.bindings.update(ans)
memoizeMemory.setdefault(queryLiteral, set()).add(
(prepMemiozedAns(ans), ns))
yield ans, ns
if not successful and queryPred not in derivedPreds:
#Open query didn't return any results and the predicate
#is ostensibly marked as derived predicate, so we have failed
memoizeMemory.setdefault(queryLiteral, set()).add(
(False, ns))
yield False, ns
else:
#All the relevant derivations have been explored and the result
#is a ground query we can directly execute against the facts
if buildProof:
factStep.bindings.update(bindings)
subquery, rt = EDBQuery([queryLiteral], factGraph,
bindings).evaluate(debug)
if buildProof:
factStep.groundQuery = subquery
memoizeMemory.setdefault(queryLiteral, set()).add(
(prepMemiozedAns(rt), ns))
yield rt, ns
from FuXi.Rete.RuleStore import SetupRuleStore
from FuXi.Rete.Util import generateTokenSet
from FuXi.Horn.HornRules import HornFromN3
from FuXi.Rete.Magic import MagicSetTransformation, AdornLiteral
from FuXi.SPARQL import RDFTuplesToSPARQL
exNs = Namespace('http://dev.w3.org/2000/10/swap/test/cwm/fam.n3#')
rules = HornFromN3('http://dev.w3.org/2000/10/swap/test/cwm/fam-rules.n3')
factGraph = Graph().parse('http://dev.w3.org/2000/10/swap/test/cwm/fam.n3',format='n3')
factGraph.bind(u'ex',exNs)
dPreds = [exNs.ancestor]
rule_store, rule_graph, network = SetupRuleStore(makeNetwork=True)
network.nsMap = {u'ex':exNs}
closureDeltaGraph=Graph()
network.inferredFacts = closureDeltaGraph
goals = [(exNs.david,exNs.ancestor,Variable('ANCESTOR'))]
for rule in MagicSetTransformation(factGraph,rules,goals,dPreds):
network.buildNetworkFromClause(rule)
# network.rules.add(rule)
print ("\t", rule)
goalLit = AdornLiteral(goals[0])
adornedGoalSeed = goalLit.makeMagicPred()
goal=adornedGoalSeed.toRDFTuple()
print (RDFTuplesToSPARQL([goalLit],factGraph,vars=[Variable('ANCESTOR')])
SELECT ?ANCESTOR { <http://dev.w3.org/2000/10/swap/test/cwm/fam.n3#david> <http://dev.w3.org/2000/10/swap/test/cwm/fam.n3#ancestor> ?ANCESTOR })
