Fix uhighlight fragmentation algorithm breaking on large fragment_siz…

…e values
elastic · Jul 11, 2022 · cafdece · cafdece
1 parent 7f7822b
commit cafdece
Show file tree

Hide file tree

Showing 2 changed files with 15 additions and 3 deletions.
diff --git a/...src/main/java/org/elasticsearch/lucene/search/uhighlight/BoundedBreakIteratorScanner.java b/...src/main/java/org/elasticsearch/lucene/search/uhighlight/BoundedBreakIteratorScanner.java
@@ -91,20 +91,20 @@ public int preceding(int offset) {
         } else {
             innerStart = Math.max(mainBreak.preceding(offset), 0);
 
-            final int targetEndOffset = offset + Math.max(0, maxLen - (offset - innerStart));
+            final long targetEndOffset = (long)offset + Math.max(0, maxLen - (offset - innerStart));
             final int textEndIndex = getText().getEndIndex();
 
             if (targetEndOffset + 1 > textEndIndex) {
                 innerEnd = textEndIndex;
             } else {
-                innerEnd = mainBreak.preceding(targetEndOffset + 1);
+                innerEnd = mainBreak.preceding((int)targetEndOffset + 1);
             }
 
             assert innerEnd != DONE && innerEnd >= innerStart;
 
             // in case no break was found up to maxLen, find one afterwards.
             if (innerStart == innerEnd) {
-                innerEnd = mainBreak.following(targetEndOffset);
+                innerEnd = mainBreak.following((int)targetEndOffset);
                 assert innerEnd - innerStart > maxLen;
             } else {
                 assert innerEnd - innerStart <= maxLen;

diff --git a/...est/java/org/elasticsearch/lucene/search/uhighlight/BoundedBreakIteratorScannerTests.java b/...est/java/org/elasticsearch/lucene/search/uhighlight/BoundedBreakIteratorScannerTests.java
@@ -133,4 +133,16 @@ public void testTextThatEndsBeforeMaxLen() {
         assertEquals(33, bi.preceding(offset));
         assertEquals(text.length(), bi.following(offset - 1));
     }
+
+    public void testFragmentSizeThatIsTooBig() {
+        final int fragmentSize = Integer.MAX_VALUE;
+        BreakIterator bi = BoundedBreakIteratorScanner.getSentence(Locale.ROOT, fragmentSize);
+
+        final String text = "Any sentence";
+        final int offset = 0; // find at beggining of text
+
+        bi.setText(text);
+        assertEquals(0, bi.preceding(offset));
+        assertEquals(text.length(), bi.following(offset - 1));
+    }
 }