Merge pull request #567 from joernhees/fix_sparql_count_sample

fix sparql count & sample for unbound

rdflib/plugins/sparql/aggregates.py

@@ -1,6 +1,6 @@
 from rdflib import Literal, XSD
 
-from rdflib.plugins.sparql.evalutils import _eval
+from rdflib.plugins.sparql.evalutils import _eval, NotBoundError
 from rdflib.plugins.sparql.operators import numeric
 from rdflib.plugins.sparql.datatypes import type_promotion
 
@@ -88,7 +88,9 @@ def agg_Count(a, group, bindings):
     for x in group:
         try:
             if a.vars != '*':
-                _eval(a.vars, x)
+                val = _eval(a.vars, x)
+                if isinstance(val, NotBoundError):
+                    continue
             c += 1
         except:
             return  # error in aggregate => no binding
@@ -98,10 +100,11 @@ def agg_Count(a, group, bindings):
 
 
 def agg_Sample(a, group, bindings):
-    try:
-        bindings[a.res] = _eval(a.vars, iter(group).next())
-    except StopIteration:
-        pass  # no res
+    for ctx in group:
+        val = _eval(a.vars, ctx)
+        if not isinstance(val, NotBoundError):
+            bindings[a.res] = val
+            break
 
 
 def agg_GroupConcat(a, group, bindings):

test/test_issue563.py

@@ -0,0 +1,48 @@
+from rdflib import BNode, Literal, Graph, Namespace
+from rdflib.compare import isomorphic
+
+EX = Namespace("http://example.org/")
+QUERY = """
+PREFIX rdf: <http://www.w3.org/1999/02/22-rdf-syntax-ns#>
+PREFIX rdfs: <http://www.w3.org/2000/01/rdf-schema#>
+SELECT ?x (%s(?y) as ?ys) (%s(?z) as ?zs) WHERE {
+  VALUES (?x ?y ?z) {
+
+    (2 6 UNDEF)
+    (2 UNDEF 10)
+
+    (3 UNDEF 15)
+    (3 9 UNDEF)
+
+    (5 UNDEF 25)
+  }
+}
+GROUP BY ?x
+"""
+
+def test_sample():
+    g = Graph()
+    results = set(tuple(i) for i in g.query(QUERY % ("SAMPLE", "SAMPLE")))
+
+    assert results == set([
+        (Literal(2), Literal(6), Literal(10)),
+        (Literal(3), Literal(9), Literal(15)),
+        (Literal(5), None, Literal(25)),
+    ])
+
+def test_count():
+    g = Graph()
+    results = set(tuple(i) for i in g.query(QUERY % ("COUNT", "COUNT")))
+
+    assert results == set([
+        (Literal(2), Literal(1), Literal(1)),
+        (Literal(3), Literal(1), Literal(1)),
+        (Literal(5), Literal(0), Literal(1)),
+    ])
+
+if __name__ == "__main__":
+    test_sample()
+    test_count()
+
+
+