od: preserve non-UTF-8 file operands

src/uu/od/src/multifile_reader.rs

@@ -5,6 +5,7 @@
 
 // spell-checker:ignore (ToDO) multifile curr fnames fname xfrd fillloop mockstream
 
+use std::ffi::OsStr;
 use std::fs::File;
 use std::io;
 #[cfg(unix)]
@@ -16,7 +17,7 @@ use uucore::{display::Quotable, error::strip_errno, show_error, translate};
 const SKIP_BUFFER_SIZE: usize = 16 * 1024;
 
 pub enum InputSource<'a> {
-    FileName(&'a str),
+    FileName(&'a OsStr),
     Stdin,
 }
 
@@ -55,7 +56,7 @@ pub struct MultifileReader<'a> {
     ni: Vec<InputSource<'a>>,
     curr_file: Option<CurrentReader>,
     any_err: bool,
-    file_name: Option<&'a str>,
+    file_name: Option<&'a OsStr>,
 }
 
 pub trait HasError {
@@ -238,7 +239,10 @@ impl io::Read for MultifileReader<'_> {
                             Err(e) => {
                                 show_error!(
                                     "{}: {}",
-                                    self.file_name.unwrap_or("I/O"),
+                                    self.file_name
+                                        .unwrap_or(OsStr::new("I/O"))
+                                        .maybe_quote()
+                                        .external(true),
                                     strip_errno(&e)
                                 );
                                 self.any_err = true;
@@ -275,11 +279,11 @@ mod tests {
     fn test_multi_file_reader_one_read() {
         let mut file1 = NamedTempFile::new().unwrap();
         file1.write_all(b"abcd").unwrap();
-        let path1 = file1.path().to_str().unwrap();
+        let path1 = file1.path().as_os_str();
 
         let mut file2 = NamedTempFile::new().unwrap();
         file2.write_all(b"ABCD").unwrap();
-        let path2 = file2.path().to_str().unwrap();
+        let path2 = file2.path().as_os_str();
 
         let inputs = vec![InputSource::FileName(path1), InputSource::FileName(path2)];
         let mut v = [0; 10];
@@ -295,11 +299,11 @@ mod tests {
     fn test_multi_file_reader_two_reads() {
         let mut file1 = NamedTempFile::new().unwrap();
         file1.write_all(b"abcd").unwrap();
-        let path1 = file1.path().to_str().unwrap();
+        let path1 = file1.path().as_os_str();
 
         let mut file2 = NamedTempFile::new().unwrap();
         file2.write_all(b"ABCD").unwrap();
-        let path2 = file2.path().to_str().unwrap();
+        let path2 = file2.path().as_os_str();
 
         let inputs = vec![InputSource::FileName(path1), InputSource::FileName(path2)];
         let mut v = [0; 5];

src/uu/od/src/od.rs

@@ -70,7 +70,7 @@ struct OdOptions {
     skip_bytes: u64,
     read_bytes: Option<u64>,
     label: Option<u64>,
-    input_strings: Vec<String>,
+    input_strings: Vec<OsString>,
     formats: Vec<ParsedFormatterItemInfo>,
     line_bytes: usize,
     output_duplicates: bool,
@@ -253,12 +253,16 @@ impl OdOptions {
 /// opens the input and calls `odfunc` to process the input.
 #[uucore::main]
 pub fn uumain(args: impl uucore::Args) -> UResult<()> {
-    let args = args.collect_ignore();
-    let raw_args: Vec<OsString> = args.iter().map(OsString::from).collect();
+    let raw_args: Vec<OsString> = args.collect();
+    // Only used to look up how options were spelled; file operands keep their raw form.
+    let args: Vec<String> = raw_args
+        .iter()
+        .map(|arg| arg.to_string_lossy().into_owned())
+        .collect();
 
     let clap_opts = uu_app();
 
-    let clap_matches = uucore::clap_localization::handle_clap_result(clap_opts, &args)?;
+    let clap_matches = uucore::clap_localization::handle_clap_result(clap_opts, &raw_args)?;
 
     // Kept for the caret in SIZE diagnostics, which echoes the command line.
     let diag_args = uucore::diagnostics::capture(&raw_args);
@@ -511,6 +515,7 @@ pub fn uu_app() -> Command {
             Arg::new(options::FILENAME)
                 .hide(true)
                 .action(ArgAction::Append)
+                .value_parser(clap::value_parser!(OsString))
                 .value_hint(clap::ValueHint::FilePath),
         )
 }
@@ -597,7 +602,7 @@ where
 
 /// Extract and display printable strings from input (od -S option)
 fn extract_strings_from_input(
-    input_strings: &[String],
+    input_strings: &[OsString],
     skip_bytes: u64,
     read_bytes: Option<u64>,
     min_length: usize,
@@ -788,12 +793,15 @@ fn write_bytes(
 }
 
 /// Helper function to convert input strings to InputSource
-fn map_input_strings(input_strings: &[String]) -> Vec<InputSource<'_>> {
+fn map_input_strings(input_strings: &[OsString]) -> Vec<InputSource<'_>> {
     input_strings
         .iter()
-        .map(|w| match w as &str {
-            "-" => InputSource::Stdin,
-            x => InputSource::FileName(x),
+        .map(|w| {
+            if w == "-" {
+                InputSource::Stdin
+            } else {
+                InputSource::FileName(w)
+            }
         })
         .collect()
 }
@@ -803,7 +811,7 @@ fn map_input_strings(input_strings: &[String]) -> Vec<InputSource<'_>> {
 /// `skip_bytes` is the number of bytes skipped from the input
 /// `read_bytes` is an optional limit to the number of bytes to read
 fn open_input_peek_reader(
-    input_strings: &[String],
+    input_strings: &[OsString],
     skip_bytes: u64,
     read_bytes: Option<u64>,
 ) -> UResult<PeekReader<BufReader<PartialReader<MultifileReader<'_>>>>> {