Alternate API ordering suggestion
MavenRain opened this issue · comments
Perhaps it might be more ergonomic to be able to write something like haystack.memchr(b'z')
. It's a simple change to make if there is an appetite for it, but would like to know what the community's thoughts are on it.
I think that's what bstr
is for. It has things like <[u8]>::find_byte
.
I would be open to an API refresh of memchr
, but I think I would want to do that as part of a memchr 3
release, if ever. I'm not a huge fan of just adding a bunch of things to the API because it can create questions like, "wait, which one of these should I be using?" While I acknowledge the current API is perhaps not the most ergonomic, I think it's better than having two distinct APIs.
I'm going to close this because I don't think it's imminently actionable. I think it's very unlikely that I add a bunch of alternate-but-equivalent APIs to memchr 2
, and this would instead need to be considered more holistically in a memchr 3
API refresh. However, I don't have any plans of doing that any time soon unfortunately.