numfmt: buffer output in non-abort invalid modes and re-enable GNU tests

Buffer formatted output so that --invalid=fail does not duplicate lines
on error, and --invalid=abort streams directly for partial output.

Fixes GNU numfmt.pl test: field-3. Re-enables the full numfmt.pl test
suite by removing the skip list added earlier in the stack.

Author: sylvestre

src/uu/numfmt/src/format.rs

@@ -599,7 +599,7 @@ fn split_bytes<'a>(input: &'a [u8], delim: &'a [u8]) -> impl Iterator<Item = &'a
     })
 }
 
-pub fn write_formatted_with_delimiter<W: std::io::Write>(
+pub fn write_formatted_with_delimiter<W: std::io::Write + ?Sized>(
     writer: &mut W,
     input: &[u8],
     options: &NumfmtOptions,
@@ -635,7 +635,7 @@ pub fn write_formatted_with_delimiter<W: std::io::Write>(
     Ok(())
 }
 
-pub fn write_formatted_with_whitespace<W: std::io::Write>(
+pub fn write_formatted_with_whitespace<W: std::io::Write + ?Sized>(
     writer: &mut W,
     s: &str,
     options: &NumfmtOptions,

src/uu/numfmt/src/numfmt.rs

@@ -20,6 +20,7 @@ use std::str::FromStr;
 use units::{IEC_BASES, SI_BASES};
 use uucore::display::Quotable;
 use uucore::error::UResult;
+use uucore::util_name;
 
 use uucore::i18n::decimal::locale_grouping_separator;
 use uucore::parser::shortcut_value_parser::ShortcutValueParser;
@@ -94,15 +95,25 @@ fn write_line<W: std::io::Write>(
         Some(i) => &input_line[..i],
         None => input_line,
     };
+    let buffer_output = !matches!(options.invalid, InvalidModes::Abort);
     let mut formatted_line = Vec::new();
-    let handled_line = if options.delimiter.is_some() {
-        write_formatted_with_delimiter(&mut formatted_line, line, options, eol)
-    } else {
-        // Whitespace mode requires valid UTF-8
-        match std::str::from_utf8(line) {
-            Ok(s) => write_formatted_with_whitespace(&mut formatted_line, s, options, eol),
-            Err(_) => {
-                Err(translate!("numfmt-error-invalid-number", "input" => escape_line(line).quote()))
+    let handled_line = {
+        let output: &mut dyn std::io::Write = if buffer_output {
+            &mut formatted_line
+        } else {
+            writer
+        };
+
+        if options.delimiter.is_some() {
+            write_formatted_with_delimiter(output, line, options, eol)
+        } else {
+            // Whitespace mode requires valid UTF-8
+            match std::str::from_utf8(line) {
+                Ok(s) => write_formatted_with_whitespace(output, s, options, eol),
+                Err(_) => Err(translate!(
+                    "numfmt-error-invalid-number",
+                    "input" => escape_line(line).quote()
+                )),
             }
         }
     };
@@ -128,7 +139,9 @@ fn write_line<W: std::io::Write>(
         return Ok(true);
     }
 
-    writer.write_all(&formatted_line)?;
+    if buffer_output {
+        writer.write_all(&formatted_line)?;
+    }
     Ok(false)
 }
 

tests/by-util/test_numfmt.rs

@@ -1261,6 +1261,15 @@ fn test_invalid_fail_with_fields_does_not_duplicate_output() {
         .stderr_is("numfmt: invalid number: 'Foo'\n");
 }
 
+#[test]
+fn test_abort_with_fields_preserves_partial_output() {
+    new_ucmd!()
+        .args(&["--field=3", "--from=auto", "Hello 40M World 90G"])
+        .fails_with_code(2)
+        .stdout_is("Hello 40M ")
+        .stderr_is("numfmt: invalid number: 'World'\n");
+}
+
 #[test]
 fn test_rejects_malformed_number_forms() {
     new_ucmd!()

util/build-gnu.sh

@@ -192,24 +192,6 @@ sed -i "s|--coreutils-prog=||g" tests/misc/coreutils.sh
 # Use the system coreutils where the test fails due to error in a util that is not the one being tested
 sed -i "s|grep '^#define HAVE_CAP 1' \$CONFIG_HEADER > /dev/null|true|"  tests/ls/capability.sh
 
-# Keep GNU numfmt coverage focused on the behavior we currently support.
-sed -i '/# uutils-numfmt-adjustments start/,/# uutils-numfmt-adjustments end/d' tests/numfmt/numfmt.pl
-sed -i '/push @Tests, @Locale_Tests if $locale ne '\''C'\'';/a\
-# uutils-numfmt-adjustments start\
-my %skip_uutils_numfmt_tests = map { $_ => 1 } qw(\
-  unit-sep-22-fail grp-1 grp-2 delim-7 field-1\
-  field-range-err-1 field-range-err-2 field-range-err-3 field-range-err-4\
-  field-range-err-6 field-range-err-7 field-range-err-8 field-range-err-9\
-  field-range-err-10 field-range-err-11 field-range-err-12 field-range-err-13\
-  strtod-6 strtod-6.1 leading-4 debug-2\
-  devdebug-1 devdebug-2 devdebug-3 devdebug-4 devdebug-5 devdebug-6\
-  devdebug-7 devdebug-9 devdebug-10 devdebug-11\
-  help-1 fmt-err-9 fmt-err-11 fmt-15\
-  ign-err-5 ign-err-m2.2 ign-err-m3.1\
-);\
-@Tests = grep { !$skip_uutils_numfmt_tests{$_->[0]} } @Tests;\
-# uutils-numfmt-adjustments end' tests/numfmt/numfmt.pl
-
 # our messages are better
 sed -i "s|cannot stat 'symlink': Permission denied|not writing through dangling symlink 'symlink'|" tests/cp/fail-perm.sh
 sed -i "s|cp: target directory 'symlink': Permission denied|cp: 'symlink' is not a directory|" tests/cp/fail-perm.sh