You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@jena.apache.org by an...@apache.org on 2026/04/21 10:35:28 UTC
(jena) 04/04: GH-3868: Output illegal surrogates as \uFFFD
This is an automated email from the ASF dual-hosted git repository.
afs pushed a commit to branch main
in repository https://gitbox.apache.org/repos/asf/jena.git
commit 0715a32b28fd3828804da95d5d9bbc1663fc8004
Author: Andy Seaborne <an...@apache.org>
AuthorDate: Sun Apr 19 13:09:46 2026 +0100
GH-3868: Output illegal surrogates as \uFFFD
---
.../java/org/apache/jena/atlas/lib/EscapeStr.java | 32 +++++++++++++++++++++-
.../org/apache/jena/atlas/lib/TestEscapeStr.java | 21 +++++++++++++-
.../apache/jena/shex/runner/RunnerPrintShex.java | 2 +-
3 files changed, 52 insertions(+), 3 deletions(-)
diff --git a/jena-base/src/main/java/org/apache/jena/atlas/lib/EscapeStr.java b/jena-base/src/main/java/org/apache/jena/atlas/lib/EscapeStr.java
index 298b7a9e70..5c2dde137d 100644
--- a/jena-base/src/main/java/org/apache/jena/atlas/lib/EscapeStr.java
+++ b/jena-base/src/main/java/org/apache/jena/atlas/lib/EscapeStr.java
@@ -135,10 +135,40 @@ public class EscapeStr
}
// Normal case!
- out.print(c);
+ if ( ! Character.isSurrogate(c) ) {
+ out.print(c);
+ continue;
+ }
+
+ // Surrogate. Check if high-low (legal).
+
+ if ( i < len-1 ) {
+ // Peek
+ char c2 = s.charAt(i+1);
+ if ( Character.isSurrogatePair(c, c2)) {
+ // Accept c2
+ i++;
+ // Valid surrogate pair!
+ // Print both surrogates raw, and let the character encoder deal with it.
+ out.print(c);
+ out.print(c2);
+ continue;
+ }
+ // The next character could be a low surrogate making this a lone-low surrogate then legal
+ }
+ // Bad surrogate. low-high, or end of string.
+ // c2 is read again.
+ outputReplacement(out, c);
+ continue;
}
}
+ private static void outputReplacement(AWriter out, char c) {
+ // This is our policy instead of Java's default of a single '?' (done deep inside UTF_8.Encoder)
+ //out.printf("\\u%04X", (int)c);
+ out.print("\\uFFFD");
+ }
+
/** Write a string with Unicode to ASCII conversion using \-u escapes */
public static void writeASCII(AWriter out, String s) {
int len = s.length();
diff --git a/jena-base/src/test/java/org/apache/jena/atlas/lib/TestEscapeStr.java b/jena-base/src/test/java/org/apache/jena/atlas/lib/TestEscapeStr.java
index 7a7a4b9c4a..bac02e4dd5 100644
--- a/jena-base/src/test/java/org/apache/jena/atlas/lib/TestEscapeStr.java
+++ b/jena-base/src/test/java/org/apache/jena/atlas/lib/TestEscapeStr.java
@@ -141,6 +141,10 @@ public class TestEscapeStr {
assertEquals(expected, output);
}
+ private void test_escape(String input, String expected) {
+ String output = EscapeStr.stringEsc(input);
+ assertEquals(expected, output);
+ }
@Test public void unescape_unicode_1() { test_unesc_unicode("", "") ; }
@Test public void unescape_unicode_2() { test_unesc_unicode("abc\\u0020def", "abc def") ; }
@Test public void unescape_unicode_3() { test_unesc_unicode("\\u0020", " ") ; }
@@ -153,7 +157,6 @@ public class TestEscapeStr {
@Test public void unescape_unicode_12() { test_unesc_unicode("\\(\\)", "\\(\\)") ; }
@Test public void unescape_unicode_13() { test_unesc_unicode("\\\\", "\\\\") ; }
- // See also TestEscapeStr
// \-u{...} style Unicode escapes
@Test public void unescape_unicode_20() { test_unesc_unicode("\\u{41}", "A") ; }
@Test public void unescape_unicode_21() { test_unesc_unicode("\\u{000000}", "\u0000") ; }
@@ -168,6 +171,22 @@ public class TestEscapeStr {
@Test public void unescape_unicode_33() { assertThrows(AtlasException.class, ()->test_unesc_unicode("\\u{1234567}", "")) ; }
@Test public void unescape_unicode_34() { assertThrows(AtlasException.class, ()->test_unesc_unicode("\\u{0000000}", "")) ; }
+ // Escaped surrogates, good and bad.
+ // Use java character escapes to put the surrogates into the java string.
+ // 🂡 is U+D83C U+DCA1
+ @Test public void escape_unicode_50() { test_escape("\uD83C\uDCA1", "🂡"); }
+ @Test public void escape_unicode_51() { test_escape("abc\uD83C\uDCA1xyz", "abc🂡xyz"); }
+ // low, then high -> illegal
+ @Test public void escape_unicode_55() { test_escape("\uDCA1\uD83C", "\\uFFFD\\uFFFD"); }
+ @Test public void escape_unicode_56() { test_escape("\uDCA1\uD83C@", "\\uFFFD\\uFFFD@"); }
+ // Lone surrogate
+ @Test public void escape_unicode_60() { test_escape("\uD83C", "\\uFFFD"); }
+ @Test public void escape_unicode_61() { test_escape("abc\uD83Cxyz", "abc\\uFFFDxyz"); }
+ @Test public void escape_unicode_62() { test_escape("\uDCA1", "\\uFFFD"); }
+ @Test public void escape_unicode_63() { test_escape("abc\uDCA1xyz", "abc\\uFFFDxyz"); }
+
+ // low, then high/low -> one illegal, encode legal pair.
+ @Test public void escape_unicode_59() { test_escape("\uDCA1\uD83C\uDCA1", "\\uFFFD🂡"); }
private void test_unesc_unicode(String input, String expected) {
String output = EscapeStr.unescapeUnicode(input) ;
diff --git a/jena-shex/src/test/java/org/apache/jena/shex/runner/RunnerPrintShex.java b/jena-shex/src/test/java/org/apache/jena/shex/runner/RunnerPrintShex.java
index b9aa9f7504..7148a348bc 100644
--- a/jena-shex/src/test/java/org/apache/jena/shex/runner/RunnerPrintShex.java
+++ b/jena-shex/src/test/java/org/apache/jena/shex/runner/RunnerPrintShex.java
@@ -109,7 +109,7 @@ public class RunnerPrintShex extends org.apache.jena.shex.runner.AbstractRunnerF
System.out.println("-- --");
Shex.printSchema(System.out, schema2);
System.out.println("== ==");
- fail("ShEx schames not equivalent");
+ fail("ShEx schemas not equivalent");
}
}