Apache
/
lucene
mirror of https://github.com/apache/lucene.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity

LUCENE-6734: sidestep auto-prefix terms test bug 

git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1704660 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Michael McCandless 2015-09-22 15:51:40 +00:00
parent 4c27e2b374
commit 4cbf0da8c6
2 changed files with 23 additions and 3 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

24
lucene/codecs/src/test/org/apache/lucene/codecs/autoprefix/TestAutoPrefixTerms.java
View File

@ -60,6 +60,7 @@ import org.apache.lucene.util.AttributeSource;
import org.apache.lucene.util.BytesRef;
import org.apache.lucene.util.FixedBitSet;
import org.apache.lucene.util.LuceneTestCase;
import org.apache.lucene.util.MathUtil;
import org.apache.lucene.util.StringHelper;
import org.apache.lucene.util.TestUtil;
import org.apache.lucene.util.automaton.Automata;
@ -547,7 +548,7 @@ public class TestAutoPrefixTerms extends LuceneTestCase {
public void finish(int expectedNumHits, int maxPrefixCount) {
if (maxPrefixCount != -1) {
// Auto-terms were used in this test
// Auto-prefix terms were used in this test
long allowedMaxTerms;
if (bounds.length == 1) {
@ -572,6 +573,25 @@ public class TestAutoPrefixTerms extends LuceneTestCase {
}
allowedMaxTerms = maxPrefixCount * (long) ((minTerm.length-commonPrefix) + (maxTerm.length-commonPrefix));
if (commonPrefix == 0) {
int min;
if (minTerm.length == 0) {
min = 0;
} else {
min = minTerm.bytes[minTerm.offset] & 0xff;
}
int max;
if (maxTerm.length == 0) {
max = 0;
} else {
max = maxTerm.bytes[maxTerm.offset] & 0xff;
}
if (max > min) {
// When maxPrefixCount is small (< 16), each byte of the term can require more than one "level" of auto-prefixing:
// NOTE: this is still only approximate ... it's tricky to get a closed form max bound that's "tight"
allowedMaxTerms += MathUtil.log(max-min, maxPrefixCount);
}
}
}
assertTrue("totTermCount=" + totTermCount + " is > allowedMaxTerms=" + allowedMaxTerms, totTermCount <= allowedMaxTerms);
@ -587,7 +607,7 @@ public class TestAutoPrefixTerms extends LuceneTestCase {
}
if (maxPrefixCount != -1) {
// Auto-terms were used in this test
// Auto-prefix terms were used in this test
int sumLeftoverSuffix = 0;
for(BytesRef bound : bounds) {

2
lucene/test-framework/src/java/org/apache/lucene/util/TestUtil.java
View File

@ -1256,7 +1256,7 @@ public final class TestUtil {
} else {
try {
return br.utf8ToString() + " " + br.toString();
} catch (IllegalArgumentException t) {
} catch (AssertionError | IllegalArgumentException t) {
// If BytesRef isn't actually UTF8, or it's eg a
// prefix of UTF8 that ends mid-unicode-char, we
// fallback to hex: