[DRAFT] Rust interpreter #158
Conversation
| @@ -47,6 +47,7 @@ path = "src/lib.rs" | |||
|
|
|||
| [features] | |||
| external-secp = [] | |||
| rust-interpreter = [] | |||
There was a problem hiding this comment.
| rust-interpreter = [] |
Since features can be enabled transitively and/or accidentally, it would be better to use an explicit compilation flag for now.
| @@ -70,6 +70,9 @@ mod orchard_ffi; | |||
| mod streams_ffi; | |||
| mod transaction_ffi; | |||
|
|
|||
| #[cfg(feature = "rust-interpreter")] | |||
There was a problem hiding this comment.
| #[cfg(feature = "rust-interpreter")] | |
| #[cfg(zcash_unstable = "rust-interpreter")] |
This follows the convention that we use in librustzcash, e.g. https://github.com/zcash/librustzcash/blob/main/zcash_primitives/src/lib.rs#L29-L30
|
|
||
| Ok(match leading_byte { | ||
| OP_PUSHDATA1 | OP_PUSHDATA2 | OP_PUSHDATA4 => { | ||
| let read_le = |script: &mut &[u8], needed_bytes: usize| { |
There was a problem hiding this comment.
Unlike in the C++ code, require_minimal_pushes is not checked as part of parsing.
|
|
||
| pub fn parse_opcode( | ||
| script: &mut &[u8], | ||
| mut buffer: Option<&mut Vec<u8>>, |
There was a problem hiding this comment.
We shouldn't need to pass a mutable buffer here, since we clear it out; let's just allocate a new one? This may be a microoptimization, but it also introduces the potential for error.
| size <<= 8; | ||
| size |= script[i] as usize; | ||
| } | ||
| *script = &script[needed_bytes..]; |
There was a problem hiding this comment.
Instead of mutating *script, we could return the slice of the underlying buffer that remains after parsing the first opcode.
| buffer.map(|buffer| { | ||
| buffer.extend(&script[0..size]); | ||
| *script = &script[size..]; | ||
| }); |
There was a problem hiding this comment.
With sufficiently recent Rust, we'll want to use https://doc.rust-lang.org/std/primitive.slice.html#method.split_first_chunk here.
| return Err(ScriptError::ScriptSize); | ||
| } | ||
|
|
||
| let mut script = (*self).clone(); |
There was a problem hiding this comment.
Why not just self.0.clone()?
No description provided.