fix mutf-8 codec

[hayanesuru]

Summary by CodeRabbit

• Bug Fixes
  • Stricter validation and error reporting for invalid modified UTF-8 continuation and overlong/null-encoding sequences.
  • Improved handling and correct conversion of UTF-16 surrogate pairs to UTF-8, ensuring buffered data is emitted in the right order.

[coderabbitai]

📝 Walkthrough

Walkthrough

Adds explicit handling and validation for 0xC0 and 0xC2..=0xDF MUTF-8 sequences, updates continuation-byte checks, and extends 0xED surrogate handling to compute code points and emit corresponding 4-byte UTF-8 sequences while adjusting buffer/index management.

Changes

Cohort / File(s)	Summary
MUTF-8 decoding logic haya_mutf8/src/lib.rs	Introduced a distinct 0xC0 branch that enforces 0xC0 0x80 as NUL, added a 0xC2..=0xDF branch with explicit continuation-byte validation and length updates in decode_mutf8_len and decode_mutf8. Invalid continuations now produce errors or are skipped as implemented.
Surrogate pair handling (0xED) haya_mutf8/src/lib.rs	Extended the 0xED branch to compute the Unicode code point from surrogate pairs, flush buffered bytes before writing the derived 4-byte UTF-8 sequence, advance indices by 6, and reset start position accordingly.

Estimated code review effort

🎯 3 (Moderate) | ⏱️ ~20 minutes

Poem

🐰
I nibble bytes where rules were blurred,
0xC0 now hops in single word,
Surrogates leap, four bytes sing true,
A mittened patch, decoded new. 🥕

✨ Finishing touches

• 📝 Generate docstrings

🧪 Generate unit tests (beta)

• Create PR with unit tests
• Post copyable unit tests in a comment
• Commit unit tests in branch fix-mutf8

---

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

[coderabbitai]

Actionable comments posted: 2

🤖 Fix all issues with AI agents

In `@haya_mutf8/src/lib.rs`:
- Around line 139-148: In the 0xC0 match arm that inspects let [sec, ref rest @
..] = bytes[..], add an explicit error branch when sec != 0x80 so the function
returns Err(Error) instead of silently accepting an invalid continuation byte;
update the logic around variables bytes, sec, rest and len in that arm to only
increment len and advance bytes when sec == 0x80, otherwise return Err(Error) to
enforce the MUTF-8 rule for the 0xC0 0x80 sequence.
- Around line 209-222: The 0xC0 match arm is advancing index before writing
which causes the 0xC0 0x80 sequence to be copied; in the match arm handling
MUTF-8 null (symbols: bytes, index, start, w.write, w.write_byte) move the index
+= 2 to after the w.write call so you first write
bytes.get_unchecked(start..index) (which excludes the 0xC0), then increment
index by 2, call w.write_byte(b'\0'), and set start = index; keep the same error
path when sec != 0x80.

[coderabbitai]

Actionable comments posted: 1

🤖 Fix all issues with AI agents

In `@haya_mutf8/src/lib.rs`:
- Around line 225-234: The 0xC2..=0xDF branch currently validates the
continuation byte using `sec as i8 >= -64`; replace that check with the same
bitmask style used elsewhere (e.g., `byte & 0xC0 == 0x80`) to match the
`0xE0..=0xEF` branch for consistent readability — locate the `0xC2..=0xDF` match
arm (uses `sec` from `bytes.get(index + 1)`) and update its validation condition
while keeping the same early `Err(Error)` return and `index += 2` behavior.