Unexpand: use byte count for multibyte characters for column width when using -a flag (uutils#9949)

* Remove Unicode width calculation for characters

* Add test for unexpand with multibyte UTF-8 input

Author: JaneIllario

src/uu/unexpand/src/unexpand.rs

@@ -13,7 +13,6 @@ use std::num::IntErrorKind;
 use std::path::Path;
 use std::str::from_utf8;
 use thiserror::Error;
-use unicode_width::UnicodeWidthChar;
 use uucore::display::Quotable;
 use uucore::error::{FromIo, UError, UResult, USimpleError};
 use uucore::translate;
@@ -279,11 +278,7 @@ fn next_char_info(uflag: bool, buf: &[u8], byte: usize) -> (CharType, usize, usi
                 Some(' ') => (CharType::Space, 0, 1),
                 Some('\t') => (CharType::Tab, 0, 1),
                 Some('\x08') => (CharType::Backspace, 0, 1),
-                Some(c) => (
-                    CharType::Other,
-                    UnicodeWidthChar::width(c).unwrap_or(0),
-                    nbytes,
-                ),
+                Some(_) => (CharType::Other, nbytes, nbytes),
                 None => {
                     // invalid char snuck past the utf8_validation_iterator somehow???
                     (CharType::Other, 1, 1)

tests/by-util/test_unexpand.rs

@@ -295,3 +295,15 @@ fn test_non_utf8_filename() {
 
     ucmd.arg(&filename).succeeds().stdout_is("\ta\n");
 }
+
+#[test]
+fn unexpand_multibyte_utf8_gnu_compat() {
+    // Verifies GNU-compatible behavior: column position uses byte count, not display width
+    // "1ΔΔΔ5" is 8 bytes (1 + 2*3 + 1), already at tab stop 8
+    // So 3 spaces should NOT convert to tab (would need 8 more to reach tab stop 16)
+    new_ucmd!()
+        .args(&["-a"])
+        .pipe_in("1ΔΔΔ5   99999\n")
+        .succeeds()
+        .stdout_is("1ΔΔΔ5   99999\n");
+}