- Add utility to do substring extraction by codepoints, instead of java char index.

- Test and use it in SubstringExpression in indeing language.

1 files changed, 3 insertions, 9 deletions

diff --git a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/expressions/SubstringExpression.java b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/expressions/SubstringExpression.java
index 6f96a215edb..1f9a341519e 100644
--- a/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/expressions/SubstringExpression.java
+++ b/indexinglanguage/src/main/java/com/yahoo/vespa/indexinglanguage/expressions/SubstringExpression.java
@@ -3,6 +3,7 @@ package com.yahoo.vespa.indexinglanguage.expressions;
 
 import com.yahoo.document.DataType;
 import com.yahoo.document.datatypes.StringFieldValue;
+import com.yahoo.text.Text;
 
 /**
  * @author Simon Thoresen Hult
@@ -32,15 +33,8 @@ public final class SubstringExpression extends Expression {
     @Override
     protected void doExecute(ExecutionContext context) {
         String input = String.valueOf(context.getValue());
-        int len = input.length();
-        if (from >= len) {
-            input = "";
-        } else if (to >= len) {
-            input = input.substring(from);
-        } else {
-            input = input.substring(from, to);
-        }
-        context.setValue(new StringFieldValue(input));
+        String substring = Text.substringByCodepoints(input, from, to);
+        context.setValue(new StringFieldValue(substring));
     }
 
     @Override