This is an automated email from the ASF dual-hosted git repository.
garydgregory pushed a commit to branch master
in repository https://gitbox.apache.org/repos/asf/commons-lang.git
The following commit(s) were added to refs/heads/master by this push:
new ae9c787a6 Keep WordUtils.wrap from splitting a surrogate pair (#1731)
ae9c787a6 is described below
commit ae9c787a68a7a97aead7d2d7e02e78aed56922c9
Author: alhuda <[email protected]>
AuthorDate: Fri Jun 26 00:48:57 2026 +0530
Keep WordUtils.wrap from splitting a surrogate pair (#1731)
---
src/main/java/org/apache/commons/lang3/text/WordUtils.java | 10 +++++++---
src/test/java/org/apache/commons/lang3/text/WordUtilsTest.java | 8 ++++++++
2 files changed, 15 insertions(+), 3 deletions(-)
diff --git a/src/main/java/org/apache/commons/lang3/text/WordUtils.java
b/src/main/java/org/apache/commons/lang3/text/WordUtils.java
index 81039bf69..d711db3f5 100644
--- a/src/main/java/org/apache/commons/lang3/text/WordUtils.java
+++ b/src/main/java/org/apache/commons/lang3/text/WordUtils.java
@@ -704,10 +704,14 @@ public static String wrap(final String str, int
wrapLength, String newLineStr, f
offset = endOfWrapAt;
} else // really long word or URL
if (wrapLongWords) {
- // wrap really long word one line at a time
- wrappedLine.append(str, offset, wrapLength + offset);
+ // wrap really long word one line at a time, but keep a
surrogate pair whole
+ int wrapAt = wrapLength + offset;
+ if (Character.isHighSurrogate(str.charAt(wrapAt - 1)) &&
Character.isLowSurrogate(str.charAt(wrapAt))) {
+ wrapAt++;
+ }
+ wrappedLine.append(str, offset, wrapAt);
wrappedLine.append(newLineStr);
- offset += wrapLength;
+ offset = wrapAt;
} else {
// do not wrap really long word, just extend beyond limit
matcher = patternToWrapOn.matcher(str.substring(offset +
wrapLength));
diff --git a/src/test/java/org/apache/commons/lang3/text/WordUtilsTest.java
b/src/test/java/org/apache/commons/lang3/text/WordUtilsTest.java
index ab34e7ea8..eed53b8ca 100644
--- a/src/test/java/org/apache/commons/lang3/text/WordUtilsTest.java
+++ b/src/test/java/org/apache/commons/lang3/text/WordUtilsTest.java
@@ -479,6 +479,14 @@ void testWrap_StringIntStringBoolean() {
assertEquals(expected, WordUtils.wrap(input, 20, "\n", false));
expected = "Click here,\nhttps://commons.apac\nhe.org, to jump to\nthe
commons website";
assertEquals(expected, WordUtils.wrap(input, 20, "\n", true));
+
+ // a hard break for a long word must not split a surrogate pair across
the new line
+ input = "a\uD83D\uDE00\uD83D\uDE00\uD83D\uDE00\uD83D\uDE00";
+ expected = "a\uD83D\uDE00\uD83D\uDE00\n\uD83D\uDE00\uD83D\uDE00";
+ assertEquals(expected, WordUtils.wrap(input, 4, "\n", true));
+ input = "\uD83D\uDE00\uD83D\uDE00\uD83D\uDE00";
+ expected = "\uD83D\uDE00\uD83D\uDE00\n\uD83D\uDE00";
+ assertEquals(expected, WordUtils.wrap(input, 3, "\n", true));
}
@ParameterizedTest
