[
https://issues.apache.org/jira/browse/GROOVY-12055?page=com.atlassian.jira.plugin.system.issuetabpanels:comment-tabpanel&focusedCommentId=18084722#comment-18084722
]
ASF GitHub Bot commented on GROOVY-12055:
-----------------------------------------
Copilot commented on code in PR #2579:
URL: https://github.com/apache/groovy/pull/2579#discussion_r3330077342
##########
subprojects/groovy-sql/src/main/java/groovy/sql/Sql.java:
##########
@@ -4688,7 +4689,7 @@ protected String nullify(String sql) {
* @return the index of the found keyword or -1 if not found
*/
protected int findWhereKeyword(String sql) {
- char[] chars = sql.toLowerCase().toCharArray();
+ char[] chars = sql.toLowerCase(Locale.ROOT).toCharArray();
Review Comment:
Lowercasing the whole SQL string can change its length for some Unicode
characters (for example U+0130 lowercases to two code points under
Locale.ROOT). Because this method returns an index that is later applied to the
original SQL string, a non-ASCII identifier before `WHERE` can produce an
incorrect offset and break `nullify`'s replacements. Compare characters
case-insensitively without creating a lowercased copy so indices stay aligned
with the original SQL.
> Assess toLower/UpperCase calls which should use Locale.ROOT
> -----------------------------------------------------------
>
> Key: GROOVY-12055
> URL: https://issues.apache.org/jira/browse/GROOVY-12055
> Project: Groovy
> Issue Type: Improvement
> Reporter: Paul King
> Priority: Major
>
--
This message was sent by Atlassian Jira
(v8.20.10#820010)
