fix(import/zsh): zsh use a special format to escape some characters (#1490)

* fix(import/zsh): zsh use a special format to escape some characters

unescape those correctly rather than throw them away.

zsh side code:
https://github.com/zsh-users/zsh/blob/9f57ca4ac8ae071727b1d77cbb8c4c0d893b9099/Src/utils.c#L4889-L4900

* fix code style

1 files changed, 42 insertions, 4 deletions

diff --git a/atuin-client/src/import/zsh.rs b/atuin-client/src/import/zsh.rs
index a4b03f2d..5bc8fc16 100644
--- a/atuin-client/src/import/zsh.rs
+++ b/atuin-client/src/import/zsh.rs
@@ -1,6 +1,7 @@
 // import old shell history!
 // automatically hoover up all that we can find
 
+use std::borrow::Cow;
 use std::path::PathBuf;
 
 use async_trait::async_trait;
@@ -61,16 +62,16 @@ impl Importer for Zsh {
 
         let mut counter = 0;
         for b in unix_byte_lines(&self.bytes) {
-            let s = match std::str::from_utf8(b) {
-                Ok(s) => s,
-                Err(_) => continue, // we can skip past things like invalid utf8
+            let s = match unmetafy(b) {
+                Some(s) => s,
+                _ => continue, // we can skip past things like invalid utf8
             };
 
             if let Some(s) = s.strip_suffix('\\') {
                 line.push_str(s);
                 line.push_str("\\\n");
             } else {
-                line.push_str(s);
+                line.push_str(&s);
                 let command = std::mem::take(&mut line);
 
                 if let Some(command) = command.strip_prefix(": ") {
@@ -116,6 +117,26 @@ fn parse_extended(line: &str, counter: i64) -> History {
     imported.build().into()
 }
 
+fn unmetafy(line: &[u8]) -> Option<Cow<str>> {
+    if line.contains(&0x83) {
+        let mut s = Vec::with_capacity(line.len());
+        let mut is_meta = false;
+        for ch in line {
+            if *ch == 0x83 {
+                is_meta = true;
+            } else if is_meta {
+                is_meta = false;
+                s.push(*ch ^ 32);
+            } else {
+                s.push(*ch)
+            }
+        }
+        String::from_utf8(s).ok().map(Cow::Owned)
+    } else {
+        std::str::from_utf8(line).ok().map(Cow::Borrowed)
+    }
+}
+
 #[cfg(test)]
 mod test {
     use itertools::assert_equal;
@@ -188,4 +209,21 @@ cargo update
             ],
         );
     }
+
+    #[tokio::test]
+    async fn test_parse_metafied() {
+        let bytes =
+            b"echo \xe4\xbd\x83\x80\xe5\xa5\xbd\nls ~/\xe9\x83\xbf\xb3\xe4\xb9\x83\xb0\n".to_vec();
+
+        let mut zsh = Zsh { bytes };
+        assert_eq!(zsh.entries().await.unwrap(), 2);
+
+        let mut loader = TestLoader::default();
+        zsh.load(&mut loader).await.unwrap();
+
+        assert_equal(
+            loader.buf.iter().map(|h| h.command.as_str()),
+            ["echo 你好", "ls ~/音乐"],
+        );
+    }
 }