I followed the same testing approach when writing a Wasm binary parser (technically, a decoder)[0].
It was pretty helpful having the official spec suite available and a major boost of confidence that your parser is compliant.
Nevertheless, it was my own tests that found a regression in the latest published version of the spec[1], which shows how important it is to have a variety of implementations.
This is a really interesting project. The ability to inspect and manipulate Wasm modules at runtime opens up a lot of possibilities for debugging and tooling. Curious whether you've benchmarked the overhead of the toolkit itself on larger Go-compiled Wasm binaries?
4 comments
[ 2.9 ms ] story [ 23.1 ms ] threadBtw, if Eli reads this: thanks for the WAT samples, they were very helpful working on wasm2go!
It was pretty helpful having the official spec suite available and a major boost of confidence that your parser is compliant.
Nevertheless, it was my own tests that found a regression in the latest published version of the spec[1], which shows how important it is to have a variety of implementations.
[0] https://github.com/agis/wadec
[1] https://github.com/WebAssembly/spec/issues/2066