Closed h33p closed 10 months ago
@h33p hey there, thanks for the PR! Gave it a quick skim and it looks fine to me. Have no issue with merging this in to support no_std. Letting tests run, and once they pass I'll merge this in.
@h33p hey there, thanks for the PR! Gave it a quick skim and it looks fine to me. Have no issue with merging this in to support no_std. Letting tests run, and once they pass I'll merge this in.
Wonderful! I noticed there was a typo in the workflow making lints not fully work. Now it's fixed.
@h33p oh, and would you mind updating the changelog to reflect your changes?
@h33p oh, and would you mind updating the changelog to reflect your changes?
Done
@h33p your feature is now incorporated into release 0.7.0
. I didn't update the readme to reflect support for it (forgot), so we can do that in a later change.
Hey there, wonderful crate!
I'm trying to build filesystem abstractions that do not depend on the standard library (i.e. could be embedded into no_std kernel and alike), and
std::path::Path(Buf)
strictly depend on existence of standard library. Funnily enough, there don't seem to be any no_std path implementations. However, this crate seems like a good candidate, because it does not merely wrap path objects, but reimplements them!This PR divorces the codebase from
std
, and only makes it depend onalloc
, with optional dependency tostd
. I supposealloc
could be void too if one were to feature gatePathBuf
, but I don't think it's worth to do it all in one go (or at all).Please let me know what you think, whether this seems like a reasonable change, and if there are any changes you'd like to see.