Merge branch 'main' into feature/env_string_args_try_with_shell_words

.github/workflows/CICD.yml

@@ -615,6 +615,8 @@ jobs:
           arm-unknown-linux-gnueabihf) sudo apt-get -y update ; sudo apt-get -y install gcc-arm-linux-gnueabihf ;;
           aarch64-unknown-linux-*) sudo apt-get -y update ; sudo apt-get -y install gcc-aarch64-linux-gnu ;;
           *-redox*) sudo apt-get -y update ; sudo apt-get -y install fuse3 libfuse-dev ;;
+          # Update binutils if MinGW due to https://github.com/rust-lang/rust/issues/112368
+          x86_64-pc-windows-gnu) C:/msys64/usr/bin/pacman.exe -Syu --needed mingw-w64-x86_64-gcc --noconfirm ; echo "C:\msys64\mingw64\bin" >> $GITHUB_PATH ;;
         esac
         case '${{ matrix.job.os }}' in
           macos-latest) brew install coreutils ;; # needed for testing
@@ -984,6 +986,10 @@ jobs:
             echo "foo" > /home/runner/.plan
             ;;
         esac
+        case '${{ matrix.job.os }}' in
+          # Update binutils if MinGW due to https://github.com/rust-lang/rust/issues/112368
+          windows-latest) C:/msys64/usr/bin/pacman.exe -Syu --needed mingw-w64-x86_64-gcc --noconfirm ; echo "C:\msys64\mingw64\bin" >> $GITHUB_PATH ;;
+        esac
     - name: Initialize toolchain-dependent workflow variables
       id: dep_vars
       shell: bash

Cargo.toml

@@ -319,7 +319,7 @@ self_cell = "1.0.3"
 selinux = "0.4"
 signal-hook = "0.3.17"
 smallvec = { version = "1.11", features = ["union"] }
-tempfile = "3.8.1"
+tempfile = "3.9.0"
 uutils_term_grid = "0.3"
 terminal_size = "0.3.0"
 textwrap = { version = "0.16.0", features = ["terminal_size"] }

deny.toml

@@ -62,7 +62,7 @@ skip = [
   { name = "linux-raw-sys", version = "0.3.8" },
   # terminal_size
   { name = "rustix", version = "0.37.26" },
-  # various crates
+  # notify
   { name = "windows-sys", version = "0.45.0" },
   # various crates
   { name = "windows-sys", version = "0.48.0" },

fuzz/fuzz_targets/fuzz_common.rs

@@ -232,7 +232,10 @@ pub fn run_gnu_cmd(
 
     let output = if let Some(input_str) = pipe_input {
         // We have an pipe input
-        command.stdin(Stdio::piped()).stdout(Stdio::piped());
+        command
+            .stdin(Stdio::piped())
+            .stdout(Stdio::piped())
+            .stderr(Stdio::piped());
 
         let mut child = command.spawn().expect("Failed to execute command");
         let child_stdin = child.stdin.as_mut().unwrap();

src/uu/cksum/src/cksum.rs

@@ -15,8 +15,8 @@ use std::io::{self, stdin, stdout, BufReader, Read, Write};
 use std::iter;
 use std::path::Path;
 use uucore::{
-    error::{FromIo, UError, UResult},
-    format_usage, help_about, help_section, help_usage,
+    error::{FromIo, UError, UResult, USimpleError},
+    format_usage, help_about, help_section, help_usage, show,
     sum::{
         div_ceil, Blake2b, Digest, DigestWriter, Md5, Sha1, Sha224, Sha256, Sha384, Sha512, Sm3,
         BSD, CRC, SYSV,
@@ -174,7 +174,13 @@ where
         });
         let (sum, sz) = digest_read(&mut options.digest, &mut file, options.output_bits)
             .map_err_context(|| "failed to read input".to_string())?;
-
+        if filename.is_dir() {
+            show!(USimpleError::new(
+                1,
+                format!("{}: Is a directory", filename.display())
+            ));
+            continue;
+        }
         if options.raw {
             let bytes = match options.algo_name {
                 ALGORITHM_OPTIONS_CRC => sum.parse::<u32>().unwrap().to_be_bytes().to_vec(),
@@ -214,13 +220,6 @@ where
             (ALGORITHM_OPTIONS_CRC, true) => println!("{sum} {sz}"),
             (ALGORITHM_OPTIONS_CRC, false) => println!("{sum} {sz} {}", filename.display()),
             (ALGORITHM_OPTIONS_BLAKE2B, _) if !options.untagged => {
-                if filename.is_dir() {
-                    return Err(io::Error::new(
-                        io::ErrorKind::InvalidInput,
-                        format!("{}: Is a directory", filename.display()),
-                    )
-                    .into());
-                }
                 if let Some(length) = options.length {
                     // Multiply by 8 here, as we want to print the length in bits.
                     println!("BLAKE2b-{} ({}) = {sum}", length * 8, filename.display());

src/uu/cut/src/cut.rs

@@ -510,6 +510,7 @@ pub fn uu_app() -> Command {
         .about(ABOUT)
         .after_help(AFTER_HELP)
         .infer_long_args(true)
+        .args_override_self(true)
         .arg(
             Arg::new(options::BYTES)
                 .short('b')

src/uu/printf/src/printf.rs

@@ -11,7 +11,7 @@ use std::ops::ControlFlow;
 
 use clap::{crate_version, Arg, ArgAction, Command};
 use uucore::error::{UResult, UUsageError};
-use uucore::format::{parse_spec_and_escape, FormatArgument};
+use uucore::format::{parse_spec_and_escape, FormatArgument, FormatItem};
 use uucore::{format_usage, help_about, help_section, help_usage};
 
 const VERSION: &str = "version";
@@ -38,14 +38,24 @@ pub fn uumain(args: impl uucore::Args) -> UResult<()> {
         None => vec![],
     };
 
+    let mut format_seen = false;
     let mut args = values.iter().peekable();
     for item in parse_spec_and_escape(format_string.as_ref()) {
+        if let Ok(FormatItem::Spec(_)) = item {
+            format_seen = true;
+        }
         match item?.write(stdout(), &mut args)? {
             ControlFlow::Continue(()) => {}
             ControlFlow::Break(()) => return Ok(()),
         };
     }
 
+    // Without format specs in the string, the iter would not consume any args,
+    // leading to an infinite loop. Thus, we exit early.
+    if !format_seen {
+        return Ok(());
+    }
+
     while args.peek().is_some() {
         for item in parse_spec_and_escape(format_string.as_ref()) {
             match item?.write(stdout(), &mut args)? {

src/uucore/src/lib/features/format/argument.rs

@@ -45,7 +45,7 @@ impl<'a, T: Iterator<Item = &'a FormatArgument>> ArgumentIter<'a> for T {
         };
         match next {
             FormatArgument::Char(c) => *c,
-            FormatArgument::Unparsed(s) => s.chars().next().unwrap_or('\0'),
+            FormatArgument::Unparsed(s) => s.bytes().next().map_or('\0', char::from),
             _ => '\0',
         }
     }

src/uucore/src/lib/features/format/escape.rs

@@ -108,6 +108,7 @@ pub fn parse_escape_code(rest: &mut &[u8]) -> EscapedChar {
         *rest = new_rest;
         match c {
             b'\\' => EscapedChar::Byte(b'\\'),
+            b'"' => EscapedChar::Byte(b'"'),
             b'a' => EscapedChar::Byte(b'\x07'),
             b'b' => EscapedChar::Byte(b'\x08'),
             b'c' => EscapedChar::End,

tests/by-util/test_cksum.rs

@@ -80,18 +80,6 @@ fn test_nonexisting_file() {
         .stderr_contains(format!("cksum: {file_name}: No such file or directory"));
 }
 
-#[test]
-fn test_folder() {
-    let (at, mut ucmd) = at_and_ucmd!();
-
-    let folder_name = "a_folder";
-    at.mkdir(folder_name);
-
-    ucmd.arg(folder_name)
-        .succeeds()
-        .stdout_only(format!("4294967295 0 {folder_name}\n"));
-}
-
 // Make sure crc is correct for files larger than 32 bytes
 // but <128 bytes (1 fold pclmul) // spell-checker:disable-line
 #[test]
@@ -312,15 +300,52 @@ fn test_raw_multiple_files() {
 }
 
 #[test]
-fn test_blake2b_fail_on_directory() {
+fn test_fail_on_folder() {
     let (at, mut ucmd) = at_and_ucmd!();
 
     let folder_name = "a_folder";
     at.mkdir(folder_name);
 
-    ucmd.arg("--algorithm=blake2b")
-        .arg(folder_name)
+    ucmd.arg(folder_name)
         .fails()
         .no_stdout()
         .stderr_contains(format!("cksum: {folder_name}: Is a directory"));
 }
+
+#[test]
+fn test_all_algorithms_fail_on_folder() {
+    let scene = TestScenario::new(util_name!());
+
+    let at = &scene.fixtures;
+
+    let folder_name = "a_folder";
+    at.mkdir(folder_name);
+
+    for algo in ALGOS {
+        scene
+            .ucmd()
+            .arg(format!("--algorithm={algo}"))
+            .arg(folder_name)
+            .fails()
+            .no_stdout()
+            .stderr_contains(format!("cksum: {folder_name}: Is a directory"));
+    }
+}
+
+#[test]
+fn test_folder_and_file() {
+    let scene = TestScenario::new(util_name!());
+
+    let at = &scene.fixtures;
+
+    let folder_name = "a_folder";
+    at.mkdir(folder_name);
+
+    scene
+        .ucmd()
+        .arg(folder_name)
+        .arg("lorem_ipsum.txt")
+        .fails()
+        .stderr_contains(format!("cksum: {folder_name}: Is a directory"))
+        .stdout_is_fixture("crc_single_file.expected");
+}

tests/by-util/test_cut.rs

@@ -255,3 +255,13 @@ fn test_equal_as_delimiter3() {
         .succeeds()
         .stdout_only_bytes("abZcd\n");
 }
+
+#[test]
+fn test_multiple() {
+    let result = new_ucmd!()
+        .args(&["-f2", "-d:", "-d="])
+        .pipe_in("a=b\n")
+        .succeeds();
+    assert_eq!(result.stdout_str(), "b\n");
+    assert_eq!(result.stderr_str(), "");
+}

tests/by-util/test_printf.rs

@@ -36,6 +36,11 @@ fn escaped_slash() {
         .stdout_only("hello\\ world");
 }
 
+#[test]
+fn unescaped_double_quote() {
+    new_ucmd!().args(&["\\\""]).succeeds().stdout_only("\"");
+}
+
 #[test]
 fn escaped_hex() {
     new_ucmd!().args(&["\\x41"]).succeeds().stdout_only("A");
@@ -639,3 +644,13 @@ fn partial_char() {
             "printf: warning: bc: character(s) following character constant have been ignored\n",
         );
 }
+
+#[test]
+fn char_as_byte() {
+    new_ucmd!().args(&["%c", "🙃"]).succeeds().stdout_only("ð");
+}
+
+#[test]
+fn no_infinite_loop() {
+    new_ucmd!().args(&["a", "b"]).succeeds().stdout_only("a");
+}