Module reader 

Little-endian cursor: u1..u8 reads, LEB128, MUTF-8.

Structs

Cursor

Functions

mutf8_decode

Format a MUTF-8 string like the DEX spec’s uleb-terminated layout: utf16_len then bytes then 0x00.

mutf8_decode_lossy

Decode MUTF-8 bytes that are already delimited (no uleb length): stops at the slice end or a bare NUL. Corrupt sequences degrade to U+FFFD. Used by the metadata query path, where the string table entry has already been sliced.

mutf8_encode

Encode a Rust string as MUTF-8 (CESU-8): BMP characters use their normal UTF-8 bytes, supplementary characters become surrogate pairs, and U+0000 becomes 0xC0 0x80. Byte-level string matching (filters, findrefs queries) needs this form to reach names/literals above the BMP — a UTF-8 needle can never match their dex encoding.