update readme

bril-rs/rs2bril/README.md

@@ -12,8 +12,8 @@ View the interface with `cargo doc --open` or install with `make install` using
 - Automatic static memory management <https://www.cs.cornell.edu/courses/cs6120/2020fa/blog/asmm/> has not been implemented so arrays must be explicitly dropped where `drop` is specialized to translate to a call to free: `drop([0]);`
 - For loops, `continue`, `break`, and ranges have not been implemented(but could be).
 - Memory is implemented using Rust arrays. These are statically sized values unlike how calls to Bril `alloc` can be dynamically sized. One solution is to just allocate a large enough array and then treat the dynamic size like the length. (A subset of vectors could also be specialized in the future).
+- Arrays must be initialized with repetition syntax, which is awkward for nested arrays.
 - In normal Rust, `if` can also be used as an expression that evaluates a value to be put in a variable. This is not implemented and it is assumed that there will only be if statements.
-- The Bril code that it produces is super inefficient and it is left to other tools to optimize it. Array initialization is unrolled is not an optimal solution.
 - `!=` and automatic promotions of integer literals to floats are not implemented.
-- to support indexing into arrays, you can cast to usize in the Rust code. This will be ignored when generating Bril. `arr[i as usize];`
-- The parts of Rust which make it valid like lifetimes, references, mutability, and function visibility are ignored and compiled away in Bril.
+- To support indexing into arrays, you can cast to usize in the Rust code. This will be ignored when generating Bril. `arr[i as usize];`
+- The parts of Rust which make it valid like lifetimes, references, mutability, and function visibility are ignored and compiled away in Bril.