Add rustfilt-like implementation - #95
Conversation
This is a much simpler implementation (no argument handling etc) but suffices for ~all of my use cases and is ~2x faster (due to the demangle_stream usage, which has landed upstream but hasn't been released).
I would be happy to hand it over! It has only ever been a thin wrapper around |
|
Please let me know what I'd need to do to transfer ownership. I'm happy to transfer the repo into the |
|
Can you add rust-lang-owner as an owner on the crates.io? If you add me as a collaborator to the GitHub repo I can move it into rust-lang (we might archive it in favor of just moving the code into this repository). Thanks! |
This is essentially trivial (no argument handling etc) but it would be enough for ~all of my use cases and is ~2x faster than published
rustfiltdue to the demangle_stream usage, which has landed upstream but hasn't been released.@luser, would you be up for moving ownership of rustfilt on crates.io into my / T-compiler hands? I think if so, then it would make more sense to have a clean-room implementation of it (or re-license under MIT OR Apache), and then publish under that name.
If we don't hear back my suggestion is to move forward with this --
rustfilthas a few more features, e.g., support for file paths,--hashes, and demangling arguments rather than only stdin/stdout, but in my experience the implementation here is very often sufficient. This is basically zero effort to maintain, and we can (IMO) drop it if rustfilt is more actively maintained.