Improve performance of BoundedBreakIteratorScanner

jusbrasil · May 26, 2021 · 85cb090 · 85cb090
1 parent a479a2a
commit 85cb090
Showing 1 changed file with 20 additions and 8 deletions.
diff --git a/server/src/main/java/org/apache/lucene/search/uhighlight/BoundedBreakIteratorScanner.java b/server/src/main/java/org/apache/lucene/search/uhighlight/BoundedBreakIteratorScanner.java
@@ -100,16 +100,28 @@ public int preceding(int offset) {
             innerStart = innerEnd;
             innerEnd = windowEnd;
         } else {
-            windowStart = innerStart = mainBreak.preceding(offset);
-            windowEnd = innerEnd = mainBreak.following(offset - 1);
-            // expand to next break until we reach maxLen
-            while (innerEnd - innerStart < maxLen) {
-                int newEnd = mainBreak.following(innerEnd);
-                if (newEnd == DONE || (newEnd - innerStart) > maxLen) {
-                    break;
+            innerStart = Math.max(mainBreak.preceding(offset), 0);
+
+            final int targetEndOffset = offset + Math.max(0, maxLen - (offset - innerStart));
+            final int textEndIndex = getText().getEndIndex();
+
+            innerEnd = mainBreak.preceding(Math.min(targetEndOffset + 1, textEndIndex));
+            assert innerEnd != DONE && innerEnd >= innerStart;
+
+            // in case no break was found up to maxLen, find one afterwards.
+            if (innerStart == innerEnd) {
+                if (targetEndOffset >= textEndIndex) {
+                    innerEnd = textEndIndex;
+                } else {
+                    innerEnd = mainBreak.following(targetEndOffset);
+                    assert innerEnd - innerStart > maxLen;
                 }
-                windowEnd = innerEnd = newEnd;
+            } else {
+                assert innerEnd - innerStart <= maxLen;
             }
+
+            windowStart = innerStart;
+            windowEnd = innerEnd;
         }
 
         if (innerEnd - innerStart > maxLen) {