You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@lucene.apache.org by no...@apache.org on 2016/06/16 12:08:34 UTC
[04/50] [abbrv] lucene-solr:apiv2: SOLR-8812: edismax: turn off mm
processing if no explicit mm spec is provided and there are explicit
operators (except for AND) - addresses problems caused by SOLR-2649
SOLR-8812: edismax: turn off mm processing if no explicit mm spec is provided and there are explicit operators (except for AND) - addresses problems caused by SOLR-2649
Project: http://git-wip-us.apache.org/repos/asf/lucene-solr/repo
Commit: http://git-wip-us.apache.org/repos/asf/lucene-solr/commit/38714399
Tree: http://git-wip-us.apache.org/repos/asf/lucene-solr/tree/38714399
Diff: http://git-wip-us.apache.org/repos/asf/lucene-solr/diff/38714399
Branch: refs/heads/apiv2
Commit: 38714399760889d2d7b188a87341aade6139ffef
Parents: 9632069
Author: Steve Rowe <sa...@apache.org>
Authored: Fri Jun 10 02:29:26 2016 -0400
Committer: Steve Rowe <sa...@apache.org>
Committed: Fri Jun 10 02:29:26 2016 -0400
----------------------------------------------------------------------
solr/CHANGES.txt | 4 +
.../solr/search/ExtendedDismaxQParser.java | 37 ++++-
.../solr/search/TestExtendedDismaxParser.java | 155 +++++++++++++++++++
3 files changed, 191 insertions(+), 5 deletions(-)
----------------------------------------------------------------------
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/38714399/solr/CHANGES.txt
----------------------------------------------------------------------
diff --git a/solr/CHANGES.txt b/solr/CHANGES.txt
index ceb7a4a..73a60f2 100644
--- a/solr/CHANGES.txt
+++ b/solr/CHANGES.txt
@@ -203,6 +203,10 @@ Bug Fixes
* SOLR-9198: config APIs unable to add multiple values with same name (noble)
+* SOLR-8812: edismax: turn off mm processing if no explicit mm spec is provided
+ and there are explicit operators (except for AND) - addresses problems caused by SOLR-2649.
+ (Greg Pendlebury, Jan H�ydahl, Erick Erickson, Steve Rowe)
+
Optimizations
----------------------
* SOLR-8722: Don't force a full ZkStateReader refresh on every Overseer operation.
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/38714399/solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java
----------------------------------------------------------------------
diff --git a/solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java b/solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java
index 8401f3e..c6e5116 100644
--- a/solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java
+++ b/solr/core/src/java/org/apache/solr/search/ExtendedDismaxQParser.java
@@ -337,11 +337,17 @@ public class ExtendedDismaxQParser extends QParser {
if(query == null) {
return null;
}
-
- // For correct lucene queries, turn off mm processing if there
- // were explicit operators (except for AND).
+ // For correct lucene queries, turn off mm processing if no explicit mm spec was provided
+ // and there were explicit operators (except for AND).
if (query instanceof BooleanQuery) {
- query = SolrPluginUtils.setMinShouldMatch((BooleanQuery)query, config.minShouldMatch, config.mmAutoRelax);
+ // config.minShouldMatch holds the value of mm which MIGHT have come from the user,
+ // but could also have been derived from q.op.
+ String mmSpec = config.minShouldMatch;
+
+ if (foundOperators(clauses, config.lowercaseOperators)) {
+ mmSpec = params.get(DisMaxParams.MM, "0%"); // Use provided mm spec if present, otherwise turn off mm processing
+ }
+ query = SolrPluginUtils.setMinShouldMatch((BooleanQuery)query, mmSpec, config.mmAutoRelax);
}
return query;
}
@@ -391,7 +397,28 @@ public class ExtendedDismaxQParser extends QParser {
}
return sb.toString();
}
-
+
+ /**
+ * Returns true if at least one of the clauses is/has an explicit operator (except for AND)
+ */
+ private boolean foundOperators(List<Clause> clauses, boolean lowercaseOperators) {
+ for (Clause clause : clauses) {
+ if (clause.must == '+') return true;
+ if (clause.must == '-') return true;
+ if (clause.isBareWord()) {
+ String s = clause.val;
+ if ("OR".equals(s)) {
+ return true;
+ } else if ("NOT".equals(s)) {
+ return true;
+ } else if (lowercaseOperators && "or".equals(s)) {
+ return true;
+ }
+ }
+ }
+ return false;
+ }
+
/**
* Generates a query string from the raw clauses, uppercasing
* 'and' and 'or' as needed.
http://git-wip-us.apache.org/repos/asf/lucene-solr/blob/38714399/solr/core/src/test/org/apache/solr/search/TestExtendedDismaxParser.java
----------------------------------------------------------------------
diff --git a/solr/core/src/test/org/apache/solr/search/TestExtendedDismaxParser.java b/solr/core/src/test/org/apache/solr/search/TestExtendedDismaxParser.java
index 019c54d..9b305f2 100644
--- a/solr/core/src/test/org/apache/solr/search/TestExtendedDismaxParser.java
+++ b/solr/core/src/test/org/apache/solr/search/TestExtendedDismaxParser.java
@@ -81,6 +81,13 @@ public class TestExtendedDismaxParser extends SolrTestCaseJ4 {
assertU(adoc("id", "63", "text_sw", "gold stocks"));
assertU(adoc("id", "64", "text_sw", "stocks gold stockade"));
assertU(adoc("id", "65", "text_sw", "snake oil"));
+ // SOLR-8812 user query example
+ assertU(adoc("id", "66", "text_sw", "hair ties barbie"));
+ assertU(adoc("id", "67", "text_sw", "hair ties"));
+ assertU(adoc("id", "68", "text_sw", "hair barbie"));
+ assertU(adoc("id", "69", "text_sw", "ties barbie"));
+ assertU(adoc("id", "70", "text_sw", "hair"));
+ assertU(adoc("id", "71", "text_sw", "ties"));
assertU(commit());
}
@@ -1145,6 +1152,154 @@ public class TestExtendedDismaxParser extends SolrTestCaseJ4 {
"mm", "100%",
"defType", "edismax")
, "*[count(//doc)=0]");
+
+ // SOLR-9174
+ assertQ("test minShouldMatch=1<-1 with explicit OR, one impossible clause, and no explicit q.op",
+ req("q", "barbie OR (hair AND nonexistentword)",
+ "qf", "text_sw",
+ "mm", "1<-1",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+ }
+
+ /* SOLR-8812 */
+ @Test
+ public void testDefaultMM() throws Exception {
+ // Ensure MM is off when explicit operators (+/-/OR/NOT) are used and no explicit mm spec is specified.
+ assertQ("Explicit OR in query with no explicit mm and q.op=AND => mm = 0%",
+ req("q", "oil OR stocks",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=4]");
+ assertQ("Explicit 'or' in query with lowercaseOperators=true, no explicit mm and q.op=AND => mm = 0%",
+ req("q", "oil or stocks",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "lowercaseOperators", "true",
+ "defType", "edismax")
+ , "*[count(//doc)=4]");
+ assertQ("Explicit OR in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "oil OR stocks",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=4]");
+ assertQ("No operator in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "oil stocks",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=4]");
+ assertQ("No operator in query with no explicit mm and q.op=AND => mm = 100%",
+ req("q", "oil stocks",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+ assertQ("No operator in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "oil stocks",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=4]");
+
+ assertQ("Explicit '-' operator in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "hair ties -barbie",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+ assertQ("Explicit NOT in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "hair ties NOT barbie",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+
+ assertQ("Explicit '-' operator in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair ties -barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+ assertQ("Explicit NOT in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair ties NOT barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+
+ assertQ("Explicit '-' operator in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair AND ties -barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+ assertQ("Explicit NOT in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair AND ties -barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+
+ assertQ("No explicit non-AND operator in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair AND ties barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=2]");
+ assertQ("No explicit non-AND operator in query with no explicit mm and q.op=AND => mm = 100%",
+ req("q", "hair AND ties barbie",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+ assertQ("No explicit non-AND operator in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "hair AND ties barbie",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=2]");
+ assertQ("No explicit non-AND operator in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "hair and ties barbie",
+ "qf", "text_sw",
+ "lowercaseOperators", "true",
+ "defType", "edismax")
+ , "*[count(//doc)=2]");
+
+ assertQ("Explicit '-' operator in query with no explicit mm and q.op=AND => mm = 100%",
+ req("q", "hair ties -barbie",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+ assertQ("Explicit NOT in query with no explicit mm and q.op=AND => mm = 100%",
+ req("q", "hair ties NOT barbie",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
+
+ assertQ("Explicit OR in query with no explicit mm and q.op=AND => mm = 0%",
+ req("q", "hair OR ties barbie",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=3]");
+ assertQ("Explicit OR in query with no explicit mm and q.op=OR => mm = 0%",
+ req("q", "hair OR ties barbie",
+ "qf", "text_sw",
+ "q.op", "OR",
+ "defType", "edismax")
+ , "*[count(//doc)=6]");
+ assertQ("Explicit OR in query with no explicit mm and no explicit q.op => mm = 0%",
+ req("q", "hair OR ties barbie",
+ "qf", "text_sw",
+ "defType", "edismax")
+ , "*[count(//doc)=6]");
+
+ assertQ("Explicit '+' operator in query with no explicit mm and q.op=AND => mm = 0%",
+ req("q", "hair ties +barbie",
+ "qf", "text_sw",
+ "q.op", "AND",
+ "defType", "edismax")
+ , "*[count(//doc)=1]");
}
public void testEdismaxSimpleExtension() throws SyntaxError {