Avi Weinstock

It seems like the exception handler's catch block gets duplicated in `AstBuilder.convertToAst`. I printed the following internal structures: ``` --- a/Procyon.CompilerTools/src/main/java/com/strobel/decompiler/ast/AstBuilder.java +++ b/Procyon.CompilerTools/src/main/java/com/strobel/decompiler/ast/AstBuilder.java @@ -106,18 +106,26 @@ public final class...

It seems like the call to `ensureDesiredProtectedRanges` in `AstBuilder.pruneExceptionHandlers` is enlarging the range of the second try block in `NestedCatchA` from `[9, 10)` to `[9, 19)`. Commenting that out causes...

I'd still like to fix this (this is a minimized reproducer for a larger program that fails to decompile some methods, with the same exception as `NestedCatchB`). Do you think...

Yes, MIT license works.

We've developed a workaround that discovers the runtime class name through `.__proto__.constructor.name`: ``` use wasm_bindgen::convert::FromWasmAbi; pub fn generic_of_jsval(js: JsValue, classname: &str) -> Result { use js_sys::{Object, Reflect}; let ctor_name =...

Partially addresses #1146.

I don't see a reason not to merge both. The `cat!` macro is terser (e.g. `cat![a, b; c, d]` is shorter than `vstack((&hstack((&a, &b)), &hstack((&c, &d))))`), as well as being...

Since you refer to them as "`vstack!` and `hstack!`", and that "`nalgebra` should only have one way of doing stacking/concatenation with macros", I think it's worth clarifying that `vstack` and...

Update: after having added the necessary lifetime parameters to the `HStack`/`VStack` structs, I'm more confident that the `Block` impl for tuples of arbitrary size would look like: ```rust impl Block...

I've implemented the lazy versions of `hstack` and `vstack`, and verified that they optimize properly. ``` #[inline(never)] #[no_mangle] fn codegen_hvstack( a: &Matrix4, b: &Matrix4, c: &Matrix4, d: &Matrix4, ) ->...