printenv skips environment variables with invalid UTF-8

closes: uutils#9701

Author: sylvestre

src/uu/printenv/src/printenv.rs

@@ -5,9 +5,14 @@
 
 use clap::{Arg, ArgAction, Command};
 use std::env;
+#[cfg(unix)]
+use std::io::{self, Write};
 use uucore::translate;
 use uucore::{error::UResult, format_usage};
 
+#[cfg(unix)]
+use std::os::unix::ffi::OsStrExt;
+
 static OPT_NULL: &str = "null";
 
 static ARG_VARIABLES: &str = "variables";
@@ -21,15 +26,34 @@ pub fn uumain(args: impl uucore::Args) -> UResult<()> {
         .map(|v| v.map(ToString::to_string).collect())
         .unwrap_or_default();
 
-    let separator = if matches.get_flag(OPT_NULL) {
-        "\x00"
+    let separator: &[u8] = if matches.get_flag(OPT_NULL) {
+        b"\x00"
     } else {
-        "\n"
+        b"\n"
     };
 
+    #[cfg(unix)]
+    let mut stdout = io::stdout().lock();
+
     if variables.is_empty() {
-        for (env_var, value) in env::vars() {
-            print!("{env_var}={value}{separator}");
+        for (env_var, value) in env::vars_os() {
+            #[cfg(unix)]
+            {
+                stdout.write_all(env_var.as_bytes())?;
+                stdout.write_all(b"=")?;
+                stdout.write_all(value.as_bytes())?;
+                stdout.write_all(separator)?;
+            }
+            #[cfg(not(unix))]
+            {
+                // On non-Unix, use lossy conversion as OsStrExt is not available
+                print!(
+                    "{}={}{}",
+                    env_var.to_string_lossy(),
+                    value.to_string_lossy(),
+                    String::from_utf8_lossy(separator)
+                );
+            }
         }
         return Ok(());
     }
@@ -41,8 +65,20 @@ pub fn uumain(args: impl uucore::Args) -> UResult<()> {
             error_found = true;
             continue;
         }
-        if let Ok(var) = env::var(env_var) {
-            print!("{var}{separator}");
+        if let Some(var) = env::var_os(&env_var) {
+            #[cfg(unix)]
+            {
+                stdout.write_all(var.as_bytes())?;
+                stdout.write_all(separator)?;
+            }
+            #[cfg(not(unix))]
+            {
+                print!(
+                    "{}{}",
+                    var.to_string_lossy(),
+                    String::from_utf8_lossy(separator)
+                );
+            }
         } else {
             error_found = true;
         }

tests/by-util/test_printenv.rs

@@ -2,6 +2,10 @@
 //
 // For the full copyright and license information, please view the LICENSE
 // file that was distributed with this source code.
+#[cfg(unix)]
+use std::ffi::OsString;
+#[cfg(unix)]
+use std::os::unix::ffi::OsStringExt;
 use uutests::new_ucmd;
 
 #[test]
@@ -90,3 +94,28 @@ fn test_null_separator() {
             .stdout_is("FOO\x00VALUE\x00");
     }
 }
+
+#[test]
+#[cfg(unix)]
+#[cfg(not(any(target_os = "freebsd", target_os = "android", target_os = "openbsd")))]
+fn test_non_utf8_value() {
+    // Environment variable values can contain non-UTF-8 bytes on Unix.
+    // printenv should output them correctly, matching GNU behavior.
+    // Reproduces: LD_PRELOAD=$'/tmp/lib.so\xff' printenv LD_PRELOAD
+    let value_with_invalid_utf8 = OsString::from_vec(b"/tmp/lib.so\xff".to_vec());
+
+    let result = new_ucmd!()
+        .env("LD_PRELOAD", &value_with_invalid_utf8)
+        .arg("LD_PRELOAD")
+        .run();
+
+    // Use byte-based assertions to avoid UTF-8 conversion issues
+    // when the test framework tries to format error messages
+    assert!(
+        result.succeeded(),
+        "Command failed with exit code: {:?}, stderr: {:?}",
+        result.code(),
+        String::from_utf8_lossy(result.stderr())
+    );
+    result.stdout_is_bytes(b"/tmp/lib.so\xff\n");
+}