Make Unicode listing GNU-sed compatible

README.md

@@ -120,8 +120,13 @@ cargo test
 ### New extensions
 * Unicode characters can be specified in regular expression pattern, replacement
   and transliteration sequences using `\uXXXX` or `\UXXXXXXXX` sequences.
+
+### Incompatible extensions
+The `-U` or `--uutil-extensions` option enables useful extensions or bug fixes
+that aren't compatible with GNU sed or POSIX.
+
 * The `l` command lists Unicode characters using the `\uXXXX` and `\UXXXXXXXX`
-  sequences.
+  escapes rather than as octal UTF-8 byte sequences.
 
 ### Incompatibilities
 * Similarly to GNU _sed_, input is processed as raw bytes or as valid UTF-8

src/sed/command.rs

@@ -38,6 +38,7 @@ pub struct ProcessingContext {
     pub sandbox: bool,
     pub unbuffered: bool,
     pub null_data: bool,
+    pub uutil_extensions: bool,
 
     // Other context
     /// Currently processed input file name (not script)

src/sed/mod.rs

@@ -114,6 +114,11 @@ pub fn uu_app() -> Command {
             arg!(-s --separate "Consider files as separate rather than as a long stream."),
             arg!(--sandbox "Operate in a sandbox by disabling e/r/w commands."),
             arg!(-u --unbuffered "Load minimal input data and flush output buffers regularly."),
+            Arg::new("uutil-extensions")
+                .short('U')
+                .long("uutil-extensions")
+                .help("Enable incompatible extensions.")
+                .action(clap::ArgAction::SetTrue),
             Arg::new("null-data")
                 .short('z')
                 .long("null-data")
@@ -243,6 +248,7 @@ fn build_context(matches: &ArgMatches) -> UResult<ProcessingContext> {
         sandbox: matches.get_flag("sandbox"),
         unbuffered: matches.get_flag("unbuffered"),
         null_data: matches.get_flag("null-data"),
+        uutil_extensions: matches.get_flag("uutil-extensions"),
 
         // Environment
         character_mode: character_mode_for_locale(&locale)?,
@@ -389,6 +395,7 @@ mod tests {
         assert!(!ctx.sandbox);
         assert!(!ctx.unbuffered);
         assert!(!ctx.null_data);
+        assert!(!ctx.uutil_extensions);
     }
 
     #[test]
@@ -406,6 +413,7 @@ mod tests {
             "-s",
             "--sandbox",
             "-u",
+            "-U",
             "-z",
         ]);
 
@@ -424,6 +432,7 @@ mod tests {
         assert!(ctx.sandbox);
         assert!(ctx.unbuffered);
         assert!(ctx.null_data);
+        assert!(ctx.uutil_extensions);
     }
 
     #[test]

src/sed/processor.rs

@@ -573,7 +573,8 @@ fn list(
 
     let mut list_line = ListLine::new(max_width);
 
-    if context.character_mode == CharacterMode::Byte {
+    if !context.uutil_extensions || context.character_mode == CharacterMode::Byte {
+        // List non-ASCII bytes in octal.
         for &byte in line.as_bytes() {
             if byte == b'\n' {
                 list_line.write_embedded_newline(output)?;
@@ -583,6 +584,7 @@ fn list(
             list_line.write_item(output, &out_str)?;
         }
     } else {
+        // List non-ASCII 8-bit characters in octal; Unicode in hex \u or \U.
         let line = line.as_str().map_err(|e| {
             input_runtime_error::<()>(
                 location,

tests/by-util/test_sed.rs

@@ -1776,16 +1776,31 @@ fn filename_non_posix() {
         .stderr_contains("invalid command code");
 }
 
-/// List Unicode input under an explicit UTF-8 locale.
+/// List Unicode input under an explicit UTF-8 locale
+/// with uutil extensions enabled.
 #[test]
 fn list_unicode() {
     new_ucmd!()
         .env("LC_ALL", "C.UTF-8")
-        .args(&["l 60", "input/unicode"])
+        .args(&["--uutil-extensions", "l 60", "input/unicode"])
         .succeeds()
         .stdout_is_fixture_bytes("output/list_unicode");
 }
 
+// List Unicode input without uutil extensions should generate octal bytes.
+check_output!(list_unicode_octal, ["l 60", "input/unicode"]);
+
+/// List Unicode input without uutil extensions should generate octal bytes,
+/// even under an explicit UTF-8 locale.
+#[test]
+fn list_unicode_octal_env() {
+    new_ucmd!()
+        .env("LC_ALL", "C.UTF-8")
+        .args(&["l 60", "input/unicode"])
+        .succeeds()
+        .stdout_is_fixture_bytes("output/list_unicode_octal");
+}
+
 /// List invalid UTF-8 bytes without decoding in byte mode.
 #[test]
 fn list_invalid_utf8_byte_locale() {

tests/fixtures/sed/output/list_unicode_octal

@@ -0,0 +1,5 @@
+Hello World or \316\232\316\261\316\273\316\267\316\274\316\
+\255\317\201\316\261 \316\272\317\214\317\203\316\274\316\
+\265 or \343\201\223\343\202\223\343\201\253\343\201\241\
+\343\201\257 \344\270\226\347\225\214 \360\237\230\200$
+Hello World or Καλημέρα κόσμε or こんにちは 世界 😀