sort: support l10n sorting

Author: sylvestre

src/uu/sort/Cargo.toml

@@ -34,7 +34,7 @@ self_cell = { workspace = true }
 tempfile = { workspace = true }
 thiserror = { workspace = true }
 unicode-width = { workspace = true }
-uucore = { workspace = true, features = ["fs", "parser", "version-cmp"] }
+uucore = { workspace = true, features = ["fs", "parser", "version-cmp", "i18n-collator"] }
 fluent = { workspace = true }
 
 [target.'cfg(target_os = "linux")'.dependencies]

src/uu/sort/src/custom_str_cmp.rs

@@ -8,6 +8,7 @@
 //! The goal is to compare strings without transforming them first (i.e. not allocating new strings)
 
 use std::cmp::Ordering;
+use uucore::i18n::collator::locale_cmp;
 
 fn filter_char(c: u8, ignore_non_printing: bool, ignore_non_dictionary: bool) -> bool {
     if ignore_non_dictionary && !(c.is_ascii_alphanumeric() || c.is_ascii_whitespace()) {
@@ -35,8 +36,8 @@ pub fn custom_str_cmp(
     ignore_case: bool,
 ) -> Ordering {
     if !(ignore_case || ignore_non_dictionary || ignore_non_printing) {
-        // There are no custom settings. Fall back to the default strcmp, which is faster.
-        return a.cmp(b);
+        // There are no custom settings. Fall back to locale-aware comparison.
+        return locale_cmp(a, b);
     }
     let mut a_chars = a
         .iter()

src/uu/sort/src/sort.rs

@@ -46,6 +46,7 @@ use uucore::error::{FromIo, strip_errno};
 use uucore::error::{UError, UResult, USimpleError, UUsageError};
 use uucore::extendedbigdecimal::ExtendedBigDecimal;
 use uucore::format_usage;
+use uucore::i18n::collator::CollatorOptions;
 use uucore::line_ending::LineEnding;
 use uucore::parser::num_parser::{ExtendedParser, ExtendedParserError};
 use uucore::parser::parse_size::{ParseSizeError, Parser};
@@ -1040,6 +1041,9 @@ const STDIN_FILE: &str = "-";
 #[uucore::main]
 #[allow(clippy::cognitive_complexity)]
 pub fn uumain(args: impl uucore::Args) -> UResult<()> {
+    // Initialize locale-aware collator for string comparisons
+    uucore::i18n::collator::try_init_collator(CollatorOptions::default());
+
     let mut settings = GlobalSettings::default();
 
     let matches = uucore::clap_localization::handle_clap_result_with_exit_code(uu_app(), args, 2)?;

tests/by-util/test_sort.rs

@@ -1906,4 +1906,24 @@ fn test_color_environment_variables() {
     }
 }
 
+#[test]
+fn test_locale_sorting_c() {
+    // Test LC_ALL=C locale sorting (should sort by byte values)
+    new_ucmd!()
+        .env("LC_ALL", "C")
+        .pipe_in("a\no\nu\nä\nö\nü\n")
+        .succeeds()
+        .stdout_is("a\no\nu\nä\nö\nü\n");
+}
+
+#[test]
+fn test_locale_sorting_german() {
+    // Test LC_ALL=de_DE.utf-8 locale sorting (should respect German collation)
+    new_ucmd!()
+        .env("LC_ALL", "de_DE.utf-8")
+        .pipe_in("a\no\nu\nä\nö\nü\n")
+        .succeeds()
+        .stdout_is("a\nä\no\nö\nu\nü\n");
+}
+
 /* spell-checker: enable */