com.bigdata.rdf.sparql.ast.eval.service.TestSearch Maven / Gradle / Ivy
Show all versions of bigdata-rdf-test Show documentation
/**
Copyright (C) SYSTAP, LLC DBA Blazegraph 2006-2016. All rights reserved.
Contact:
SYSTAP, LLC DBA Blazegraph
2501 Calvert ST NW #106
Washington, DC 20008
[email protected]
This program is free software; you can redistribute it and/or modify
it under the terms of the GNU General Public License as published by
the Free Software Foundation; version 2 of the License.
This program is distributed in the hope that it will be useful,
but WITHOUT ANY WARRANTY; without even the implied warranty of
MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the
GNU General Public License for more details.
You should have received a copy of the GNU General Public License
along with this program; if not, write to the Free Software
Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA
*/
/*
* Created on Sep 4, 2011
*/
package com.bigdata.rdf.sparql.ast.eval.service;
import com.bigdata.rdf.sparql.ast.eval.AbstractDataDrivenSPARQLTestCase;
import com.bigdata.rdf.sparql.ast.service.ServiceNode;
import com.bigdata.rdf.store.BD;
/**
* Data driven test suite for full text search.
*
* @author Bryan Thompson
* @version $Id$
*
* TODO Correct rejection test when a magic predicate for search
* appears outside of a named query clause.
*
* TODO Correct rejection test when the subject of a magic predicate
* for search is a constant.
*
* TODO Correct rejection test when an unknown magic predicate in the
* {@link BD#SEARCH_NAMESPACE} is observed.
*
* TODO Add unit tests for slicing by rank and relevance.
*/
public class TestSearch extends AbstractDataDrivenSPARQLTestCase {
/**
*
*/
public TestSearch() {
}
/**
* @param name
*/
public TestSearch(String name) {
super(name);
}
/**
* A simple full text search query. The query is transformed from
*
*
* PREFIX bd:
*
* SELECT ?subj ?score
*
* WITH {
* SELECT ?subj ?score
* WHERE {
* ?lit bd:search "mike" .
* ?lit bd:relevance ?score .
* ?subj ?p ?lit .
* }
* ORDER BY DESC(?score)
* LIMIT 10
* OFFSET 0
* } as %searchSet1
*
* WHERE {
*
* include %searchSet1
*
* }
*
*
*
* PREFIX bd:
* WITH {
* QueryType: SELECT
* SELECT ( VarNode(subj) AS VarNode(subj) ) ( VarNode(score) AS VarNode(score) )
* JoinGroupNode {
* SERVICE {
* JoinGroupNode {
* StatementPatternNode(VarNode(lit), ConstantNode(TermId(0U)[http://www.bigdata.com/rdf/search#search]), ConstantNode(TermId(0L)[mike]), DEFAULT_CONTEXTS)
* StatementPatternNode(VarNode(lit), ConstantNode(TermId(0U)[http://www.bigdata.com/rdf/search#relevance]), VarNode(score), DEFAULT_CONTEXTS)
* }
* }
* StatementPatternNode(VarNode(subj), VarNode(p), VarNode(lit), DEFAULT_CONTEXTS)
* com.bigdata.rdf.sparql.ast.eval.AST2BOpBase.estimatedCardinality=5
* com.bigdata.rdf.sparql.ast.eval.AST2BOpBase.originalIndex=SPOC
* }
* order by com.bigdata.rdf.sparql.ast.OrderByExpr(VarNode(score))[ ascending=false]
* slice(limit=10)
* } AS %searchSet1 JOIN ON () DEPENDS ON ()
* QueryType: SELECT
* SELECT ( VarNode(subj) AS VarNode(subj) ) ( VarNode(score) AS VarNode(score) )
* JoinGroupNode {
* INCLUDE %searchSet1 JOIN ON ()
* }
*
*/
public void test_search_1() throws Exception {
new TestHelper("search-1").runTest();
}
/**
* A simple full text search query (variant of the above using the
* SPARQL 1.1 SERVICE syntax).
*/
public void test_search_service_1() throws Exception {
new TestHelper("search-service-1").runTest();
}
/**
* A full text search query based on the in-depth example in the full
* text index core package ("child proofing" example).
*/
public void test_search_2() throws Exception {
new TestHelper("search-2").runTest();
}
/**
* An example with an additional join in the named query and a join in the
* main query as well.
*/
public void test_search_3() throws Exception {
new TestHelper("search-3").runTest();
}
/**
* Unit test for a prefix match ("mi*").
*
*
* # Search query.
* PREFIX bds:
*
* SELECT ?subj ?label
* WHERE {
* ?lit bds:search "mi*" .
* ?lit bds:relevance ?cosine .
* ?subj ?p ?label .
* }
*
*/
public void test_search_prefix_match() throws Exception {
new TestHelper(
"search-prefix-match",// testURI
"search-prefix-match.rq",// query
"search-prefix-match.trig",//data
"search-prefix-match.srx" // expected results
).runTest();
}
/**
* Unit test for a prefix match ("mi*").
*
*
* # Search query.
* PREFIX bds:
*
* SELECT ?subj ?label
* WITH {
* SELECT ?subj ( ?lit as ?label )
* WHERE {
* ?lit bds:search "mi*" .
* ?lit bds:relevance ?cosine .
* ?subj ?p ?lit .
* }
* } as %searchSet1
* WHERE {
* include %searchSet1
* }
*
*/
public void test_search_prefix_match2() throws Exception {
new TestHelper(
"search-prefix-match2",// testURI
"search-prefix-match2.rq",// query
"search-prefix-match.trig",//data
"search-prefix-match.srx" // expected results
).runTest();
}
/**
* Unit test for a search where all tokens in the query must match.
*/
public void test_search_match_all_terms() throws Exception {
new TestHelper("search-match-all-terms").runTest();
}
/**
* Unit test with named graphs (quads) using graph graph pattern and a
* subject join in the named subquery. In this case we do not need to do
* anything more to filter out literals matched which do not appear in
* statements for the named graph.
*
* This version has an unbound graph variable.
*/
public void test_search_named_graphs1() throws Exception {
new TestHelper("search-named-graphs1").runTest();
}
/**
* Variant of the test above in which the graph variable is bound to the
* graph in which the search result is visible.
*/
public void test_search_named_graphs2() throws Exception {
new TestHelper("search-named-graphs2").runTest();
}
/**
* Variant of the test above in which the graph variable is bound to a graph
* in which the search result is not visible.
*/
public void test_search_named_graphs3() throws Exception {
new TestHelper("search-named-graphs3").runTest();
}
/**
* Variant of the test above in there are search results in more than one
* graph.
*/
public void test_search_named_graphs4() throws Exception {
new TestHelper("search-named-graphs4").runTest();
}
/**
* Unit test with named graphs (quads) using a graph graph pattern but not
* having a subject join in the named subquery. In this case we need to take
* action in order to ensure that literal matches are filtered out if they
* do not appear in a statement for the named graph. Verify that a subject
* join is automatically added into the named query using an anonymous
* variable for the subject and predicate positions to impose that
* constraint.
*/
public void test_search_named_graphs5() throws Exception {
new TestHelper("search-named-graphs5").runTest();
}
/**
* Run the query with no graphs specified.
*
*
* SELECT ?s
* WHERE {
* ?s bd:search "Alice" .
* }
*
*
* Note: The first and second condition in the old test were identical. The
* first condition was modified to run with no graphs specified when the
* tests were migrated to this format.
*
* Note: This is the first of a series of tests which are a port of
* com.bigdata.rdf.sail.TestSearchQuery#testWithNamedGraphs(). Some errors
* in that test were corrected when the tests were migrated to this format.
*/
public void test_search_named_graphs10a() throws Exception {
new TestHelper(//
"search-named-graphs10", // testURI
"search-named-graphs10a.rq", // queryFileURI
"search-named-graphs10.trig",// dataFileURI
"search-named-graphs10a.srx"// resultFileURI
).runTest();
}
/**
* Run the query with graphA specified as the default graph.
*
*
* SELECT ?s
* FROM :graphA
* WHERE {
* ?s bd:search "Alice" .
* }
*
*
* @throws Exception
*/
public void test_search_named_graphs10b() throws Exception {
new TestHelper(//
"search-named-graphs10", // testURI
"search-named-graphs10b.rq", // queryFileURI
"search-named-graphs10.trig",// dataFileURI
"search-named-graphs10b.srx"// resultFileURI
).runTest();
}
/**
* Run the query with graphB specified as the default graph
*
*
* SELECT ?s
* FROM :graphB
* WHERE {
* ?s bd:search "Alice" .
* }
*
*/
public void test_search_named_graphs10c() throws Exception {
new TestHelper(//
"search-named-graphs10", // testURI
"search-named-graphs10c.rq", // queryFileURI
"search-named-graphs10.trig",// dataFileURI
"search-named-graphs10c.srx"// resultFileURI
).runTest();
}
/**
* Run the query with graphB specified as the default graph and also pull
* out the subject from the search variable.
*
*
* SELECT ?s ?o
* FROM :graphB
* WHERE {
* ?s rdfs:label ?o .
* ?o bd:search "Alice" .
* }
*
*/
public void test_search_named_graphs10d() throws Exception {
new TestHelper(//
"search-named-graphs10", // testURI
"search-named-graphs10d.rq", // queryFileURI
"search-named-graphs10.trig",// dataFileURI
"search-named-graphs10d.srx"// resultFileURI
).runTest();
}
/**
* Run the query with graphA specified as the default graph and also pull
* out the subject from the search variable.
*
*
* SELECT ?s ?o
* FROM :graphA
* WHERE {
* ?s rdfs:label ?o .
* ?o bd:search "Alice" .
* }
*
*
* Note: Note: this was running against graphB in the old test suite, but we
* do that in the case above. It has been modified to run against graphA in
* this test suite.
*/
public void test_search_named_graphs10e() throws Exception {
new TestHelper(//
"search-named-graphs10", // testURI
"search-named-graphs10e.rq", // queryFileURI
"search-named-graphs10.trig",// dataFileURI
"search-named-graphs10e.srx"// resultFileURI
).runTest();
}
/**
* Unit test for a query with magic search predicates in the main WHERE
* clause. These predicates should be translated into a {@link ServiceNode}
* which is then lifted into a named subquery.
*/
public void test_search_main_where_clause() throws Exception {
new TestHelper("search-main-where-clause").runTest();
}
/**
* Unit test with explicit {@link BD#MIN_RELEVANCE} of zero and a prefix
* match search.
*/
public void test_search_min_relevance() throws Exception {
new TestHelper("search-min-relevance").runTest();
}
/**
* Ported from TestNamedGraphs in the sails package.
*/
public void test_search_query() throws Exception {
if(!store.isQuads())
return;
new TestHelper(
"search-query",// testURI
"search-query.rq", // queryURI
"search-query.trig", // dataURI
"search-query.srx" // resultURI
).runTest();
}
// /**
// * TODO Unit test for search-in-search pattern:
// *
// *
// * select ?snippet
// * where {
// *
// * ?o bd:search "foo" .
// * ?snippet ?p ?o .
// *
// * ?snippet ?p1 ?o1 .
// * filter(isLiteral(?o1)) .
// * filter(regex(str(?o1),"bar")) .
// *
// * }
// * You could achieve the same result by running it like this:
// *
// * select ?snippet
// * where {
// *
// * ?o1 bd:search "foo" .
// * ?snippet ?p1 ?o1 .
// *
// * ?o2 bd:search "bar" .
// * ?snippet ?p2 ?o2 .
// *
// * }
// *
// * You'd want to do a hash join of the two groups of two predicates, joining on ?snippet.
// * there needs to be pagination on both sides of the two searches or the query will croak.
// *
// * So for something like snippet merge:
// *
// * select ?snippet1 ?snippet2 ?mergeVal
// * where {
// *
// * ?o1 bd:search "foo" .
// * ?o1 bd:rank ?rank1 .
// * ?snippet1 ?p1 ?o1 .
// * ?snippet1 ?mergeVal .
// *
// * optional {
// *
// * ?o2 bd:search "bar" .
// * ?o2 bd:rank ?rank2 .
// * ?snippet2 ?p2 ?o2 .
// * ?snippet2 ?mergeVal .
// *
// * }
// *
// * }
// *
// * I page through the right side of the hash join (the "bar" search) rank by rank.
// * When I hit the end of the right side I move one page forward on the left side
// * and then do the whole paging of the right side over again. Right now I have no
// * way to do an N-way merge, because the pagination would start to take forever.
// *
// * ?rank1, ?rank2
// * 1, 1
// * 1, 2
// * 1, 3
// * 1, 4
// * 2, 1
// * 2, 2
// * 2, 3
// * 2, 4
// *
// */
// public void test_search_in_search() {
// fail("write test");
// }
}