Apache/lucene
1
0
Fork 0
mirror of https://github.com/apache/lucene.git synced 2025-02-07 10:38:40 +00:00
Code Issues Packages Projects Releases Wiki Activity

LUCENE-10272: cross-check norms with postings in checkindex (#493)

Browse Source 
Previously, CheckIndex would iterate norms and validate each one. But if norms that should be there were missing, nothing would fail. Now it computes an expected count of norms and ensures it saw them all.
This commit is contained in:
Robert Muir 2021-11-30 14:21:40 -05:00
parent 51e023bf7a
commit c89c78cee0
No known key found for this signature in database
GPG Key ID: 817AE1DD322D7ECA
1 changed files with 30 additions and 6 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

36
lucene/core/src/java/org/apache/lucene/index/CheckIndex.java
View File

@ -2115,6 +2115,8 @@ public final class CheckIndex implements Closeable {
if (fieldInfo.hasNorms() && isVectors == false) {
final NumericDocValues norms = normsProducer.getNorms(fieldInfo);
// count of valid norm values found for the field
int actualCount = 0;
// Cross-check terms with norms
for (int doc = norms.nextDoc();
doc != DocIdSetIterator.NO_MORE_DOCS;
@ -2126,12 +2128,15 @@ public final class CheckIndex implements Closeable {
continue;
}
final long norm = norms.longValue();
if (norm != 0 && visitedDocs.get(doc) == false) {
throw new CheckIndexException(
"Document "
+ doc
+ " doesn't have terms according to postings but has a norm value that is not zero: "
+ Long.toUnsignedString(norm));
if (norm != 0) {
actualCount++;
if (visitedDocs.get(doc) == false) {
throw new CheckIndexException(
"Document "
+ doc
+ " doesn't have terms according to postings but has a norm value that is not zero: "
+ Long.toUnsignedString(norm));
}
} else if (norm == 0 && visitedDocs.get(doc)) {
throw new CheckIndexException(
"Document "
@ -2139,6 +2144,25 @@ public final class CheckIndex implements Closeable {
+ " has terms according to postings but its norm value is 0, which may only be used on documents that have no terms");
}
}
int expectedCount = 0;
for (int doc = visitedDocs.nextSetBit(0);
doc != DocIdSetIterator.NO_MORE_DOCS;
doc =
doc + 1 >= visitedDocs.length()
? DocIdSetIterator.NO_MORE_DOCS
: visitedDocs.nextSetBit(doc + 1)) {
if (liveDocs != null && liveDocs.get(doc) == false) {
// Norms may only be out of sync with terms on deleted documents.
// This happens when a document fails indexing and in that case it
// should be immediately marked as deleted by the IndexWriter.
continue;
}
expectedCount++;
}
if (expectedCount != actualCount) {
throw new CheckIndexException(
"actual norm count: " + actualCount + " but expected: " + expectedCount);
}
}
// Test seek to last term: