Apache/lucene
1
0
Fork 0
mirror of https://github.com/apache/lucene.git synced 2025-03-02 14:29:23 +00:00
Code Issues Packages Projects Releases Wiki Activity

sanity check offsets in checkindex

Browse Source 
git-svn-id: https://svn.apache.org/repos/asf/lucene/dev/trunk@1349349 13f79535-47bb-0310-9956-ffa450edef68
This commit is contained in:
Robert Muir 2012-06-12 13:42:50 +00:00
parent 2397fb9d65
commit 4b212a7c4f
1 changed files with 50 additions and 2 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

52
lucene/core/src/java/org/apache/lucene/index/CheckIndex.java
View File

@ -681,6 +681,7 @@ public class CheckIndex {
DocsEnum docs = null; DocsEnum docs = null;
DocsEnum docsAndFreqs = null; DocsEnum docsAndFreqs = null;
DocsAndPositionsEnum postings = null; DocsAndPositionsEnum postings = null;
DocsAndPositionsEnum offsets = null;
String lastField = null; String lastField = null;
final FieldsEnum fieldsEnum = fields.iterator(); final FieldsEnum fieldsEnum = fields.iterator();
@ -756,6 +757,7 @@ public class CheckIndex {
docs = termsEnum.docs(liveDocs, docs, false); docs = termsEnum.docs(liveDocs, docs, false);
docsAndFreqs = termsEnum.docs(liveDocs, docsAndFreqs, true); docsAndFreqs = termsEnum.docs(liveDocs, docsAndFreqs, true);
postings = termsEnum.docsAndPositions(liveDocs, postings, false); postings = termsEnum.docsAndPositions(liveDocs, postings, false);
offsets = termsEnum.docsAndPositions(liveDocs, offsets, true);
if (hasOrd) { if (hasOrd) {
long ord = -1; long ord = -1;
@ -779,19 +781,29 @@ public class CheckIndex {
final DocsEnum docsAndFreqs2; final DocsEnum docsAndFreqs2;
final boolean hasPositions; final boolean hasPositions;
final boolean hasFreqs; final boolean hasFreqs;
if (postings != null) { final boolean hasOffsets;
if (offsets != null) {
docs2 = postings = offsets;
docsAndFreqs2 = postings = offsets;
hasOffsets = true;
hasPositions = true;
hasFreqs = true;
} else if (postings != null) {
docs2 = postings; docs2 = postings;
docsAndFreqs2 = postings; docsAndFreqs2 = postings;
hasOffsets = false;
hasPositions = true; hasPositions = true;
hasFreqs = true; hasFreqs = true;
} else if (docsAndFreqs != null) { } else if (docsAndFreqs != null) {
docs2 = docsAndFreqs; docs2 = docsAndFreqs;
docsAndFreqs2 = docsAndFreqs; docsAndFreqs2 = docsAndFreqs;
hasOffsets = false;
hasPositions = false; hasPositions = false;
hasFreqs = true; hasFreqs = true;
} else { } else {
docs2 = docs; docs2 = docs;
docsAndFreqs2 = null; docsAndFreqs2 = null;
hasOffsets = false;
hasPositions = false; hasPositions = false;
hasFreqs = false; hasFreqs = false;
} }
@ -826,6 +838,7 @@ public class CheckIndex {
lastDoc = doc; lastDoc = doc;
int lastPos = -1; int lastPos = -1;
int lastOffset = 0;
if (hasPositions) { if (hasPositions) {
for(int j=0;j<freq;j++) { for(int j=0;j<freq;j++) {
final int pos = postings.nextPosition(); final int pos = postings.nextPosition();
@ -846,6 +859,23 @@ public class CheckIndex {
if (postings.hasPayload()) { if (postings.hasPayload()) {
postings.getPayload(); postings.getPayload();
} }
if (hasOffsets) {
int startOffset = postings.startOffset();
int endOffset = postings.endOffset();
if (startOffset < 0) {
throw new RuntimeException("term " + term + ": doc " + doc + ": pos " + pos + ": startOffset " + startOffset + " is out of bounds");
}
if (startOffset < lastOffset) {
throw new RuntimeException("term " + term + ": doc " + doc + ": pos " + pos + ": startOffset " + startOffset + " < lastStartOffset " + lastOffset);
}
if (endOffset < 0) {
throw new RuntimeException("term " + term + ": doc " + doc + ": pos " + pos + ": endOffset " + endOffset + " is out of bounds");
}
if (endOffset < startOffset) {
throw new RuntimeException("term " + term + ": doc " + doc + ": pos " + pos + ": endOffset " + endOffset + " < startOffset " + startOffset);
}
lastOffset = startOffset;
}
} }
} }
} }
@ -892,7 +922,7 @@ public class CheckIndex {
if (hasPositions) { if (hasPositions) {
for(int idx=0;idx<7;idx++) { for(int idx=0;idx<7;idx++) {
final int skipDocID = (int) (((idx+1)*(long) maxDoc)/8); final int skipDocID = (int) (((idx+1)*(long) maxDoc)/8);
postings = termsEnum.docsAndPositions(liveDocs, postings, false); postings = termsEnum.docsAndPositions(liveDocs, postings, hasOffsets);
final int docID = postings.advance(skipDocID); final int docID = postings.advance(skipDocID);
if (docID == DocIdSetIterator.NO_MORE_DOCS) { if (docID == DocIdSetIterator.NO_MORE_DOCS) {
break; break;
@ -905,6 +935,7 @@ public class CheckIndex {
throw new RuntimeException("termFreq " + freq + " is out of bounds"); throw new RuntimeException("termFreq " + freq + " is out of bounds");
} }
int lastPosition = -1; int lastPosition = -1;
int lastOffset = 0;
for(int posUpto=0;posUpto<freq;posUpto++) { for(int posUpto=0;posUpto<freq;posUpto++) {
final int pos = postings.nextPosition(); final int pos = postings.nextPosition();
// NOTE: pos=-1 is allowed because of ancient bug // NOTE: pos=-1 is allowed because of ancient bug
@ -921,6 +952,23 @@ public class CheckIndex {
throw new RuntimeException("position " + pos + " is < lastPosition " + lastPosition); throw new RuntimeException("position " + pos + " is < lastPosition " + lastPosition);
} }
lastPosition = pos; lastPosition = pos;
if (hasOffsets) {
int startOffset = postings.startOffset();
int endOffset = postings.endOffset();
if (startOffset < 0) {
throw new RuntimeException("term " + term + ": doc " + docID + ": pos " + pos + ": startOffset " + startOffset + " is out of bounds");
}
if (startOffset < lastOffset) {
throw new RuntimeException("term " + term + ": doc " + docID + ": pos " + pos + ": startOffset " + startOffset + " < lastStartOffset " + lastOffset);
}
if (endOffset < 0) {
throw new RuntimeException("term " + term + ": doc " + docID + ": pos " + pos + ": endOffset " + endOffset + " is out of bounds");
}
if (endOffset < startOffset) {
throw new RuntimeException("term " + term + ": doc " + docID + ": pos " + pos + ": endOffset " + endOffset + " < startOffset " + startOffset);
}
lastOffset = startOffset;
}
} }
final int nextDocID = postings.nextDoc(); final int nextDocID = postings.nextDoc();