fix: keep non-ascii text intact through the windows clipboard

rust/src/clipboard.rs

@@ -9,6 +9,19 @@ struct ClipboardCommand {
     args: &'static [&'static str],
 }
 
+// Windows PowerShell talks to a redirected pipe in the console's OEM code page (437 or 850), not
+// UTF-8, so a selection like "Grão moído" comes back as "Grao mo\u{fffd}do" and text we write
+// reaches the clipboard as mojibake. Reading pins stdout to UTF-8 without a BOM. Writing decodes
+// the raw stdin bytes as UTF-8 itself, because setting [Console]::InputEncoding does not change
+// how `$input` is decoded. Like `$input | Set-Clipboard` did, it drops one trailing line break and
+// joins the lines with CRLF.
+//
+// Not `#[cfg(windows)]`: the `if cfg!(windows)` branches below are still compiled elsewhere, so
+// the names must exist on every platform.
+const WINDOWS_READ_SCRIPT: &str =
+    "[Console]::OutputEncoding = New-Object System.Text.UTF8Encoding $false; Get-Clipboard -Raw";
+const WINDOWS_WRITE_SCRIPT: &str = "$m = New-Object System.IO.MemoryStream; [Console]::OpenStandardInput().CopyTo($m); $t = [System.Text.Encoding]::UTF8.GetString($m.ToArray()) -replace '\\r?\\n\\z', ''; if ($t) { Set-Clipboard -Value ($t -replace '(?<!\\r)\\n', [Environment]::NewLine) } else { Set-Clipboard -Value $null }";
+
 fn read_commands() -> Vec<ClipboardCommand> {
     if cfg!(target_os = "macos") {
         return vec![ClipboardCommand {
@@ -23,7 +36,7 @@ fn read_commands() -> Vec<ClipboardCommand> {
                 "-NoProfile",
                 "-NonInteractive",
                 "-Command",
-                "Get-Clipboard -Raw",
+                WINDOWS_READ_SCRIPT,
             ],
         }];
     }
@@ -57,7 +70,7 @@ fn write_commands() -> Vec<ClipboardCommand> {
                 "-NoProfile",
                 "-NonInteractive",
                 "-Command",
-                "$input | Set-Clipboard",
+                WINDOWS_WRITE_SCRIPT,
             ],
         }];
     }
@@ -138,4 +151,63 @@ mod tests {
         assert_ne!(read_commands().len(), 0);
         assert_ne!(write_commands().len(), 0);
     }
+
+    // These two drive the real system clipboard, so they are `#[ignore]`d. Run them on Windows
+    // with `cargo test clipboard -- --ignored`. The other side of each check is a separate
+    // PowerShell that moves UTF-16 code units as plain ASCII numbers, so the result does not
+    // depend on any code page. Round-tripping `write_clipboard` into `read_clipboard` would not
+    // catch this bug: the two code-page errors cancel out.
+    //
+    // The read test fails without the fix. The write test only fails when powershell.exe starts
+    // with a console of its own, as it does when Herdr runs the plugin with no console attached;
+    // from a terminal the child shares that terminal's console and can pass without the fix.
+    #[cfg(windows)]
+    mod windows_utf8 {
+        #![allow(clippy::expect_used, reason = "tests assert by panicking")]
+        use super::super::*;
+
+        const TEXT: &str = "Grão moído, açúcar — ✓";
+
+        fn powershell(script: &str) -> String {
+            let output = process("powershell.exe")
+                .args(["-NoProfile", "-NonInteractive", "-Command", script])
+                .stdin(Stdio::null())
+                .output()
+                .expect("run powershell.exe");
+            assert!(
+                output.status.success(),
+                "powershell failed: {}",
+                String::from_utf8_lossy(&output.stderr)
+            );
+            String::from_utf8_lossy(&output.stdout).into_owned()
+        }
+
+        fn units() -> Vec<u16> {
+            TEXT.encode_utf16().collect()
+        }
+
+        #[test]
+        #[ignore = "overwrites the system clipboard"]
+        fn reads_non_ascii_text_copied_by_another_program() {
+            let list: Vec<String> = units().iter().map(u16::to_string).collect();
+            powershell(&format!(
+                "Set-Clipboard -Value (-join ([char[]]({})))",
+                list.join(",")
+            ));
+            let read = read_clipboard().expect("read the clipboard");
+            assert_eq!(read.trim_end(), TEXT);
+        }
+
+        #[test]
+        #[ignore = "overwrites the system clipboard"]
+        fn writes_non_ascii_text_other_programs_can_read() {
+            write_clipboard(TEXT).expect("write the clipboard");
+            let seen: Vec<u16> =
+                powershell("(Get-Clipboard -Raw).ToCharArray() | ForEach-Object { [int]$_ }")
+                    .lines()
+                    .map(|line| line.trim().parse().expect("a UTF-16 code unit"))
+                    .collect();
+            assert_eq!(seen, units());
+        }
+    }
 }