Details
-
Bug
-
Status: Open
-
Major
-
Resolution: Unresolved
-
4.3
-
None
-
None
-
New
Description
I think this bug is some extreme corner case...
This test currently only uses up to 9 terms. By increasing it to 26 and blasting the test, I was able to uncover a bug.
Here's the seed: ant test -Dtestcase=TestMinShouldMatch2 -Dtests.method=testNextAllTerms -Dtests.seed=E0334C37E6E190D8 -Dtests.slow=true -Dtests.locale=pl_PL -Dtests.timezone=Asia/Thimphu -Dtests.file.encoding=US-ASCII
Here's the patch to make the test use 26 terms.
Index: lucene/core/src/test/org/apache/lucene/search/TestMinShouldMatch2.java =================================================================== --- lucene/core/src/test/org/apache/lucene/search/TestMinShouldMatch2.java (revision 1459937) +++ lucene/core/src/test/org/apache/lucene/search/TestMinShouldMatch2.java (working copy) @@ -56,7 +56,7 @@ static final String alwaysTerms[] = { "a" }; static final String commonTerms[] = { "b", "c", "d" }; static final String mediumTerms[] = { "e", "f", "g" }; - static final String rareTerms[] = { "h", "i", "j" }; + static final String rareTerms[] = { "h", "i", "j", "k", "l", "m", "n", "o", "p", "q", "r", "s", "t", "u", "v", "w", "x", "y", "z" }; @Override public void setUp() throws Exception {