docs update & bump to 0.12

[kylebarron]

• I agree to follow the project's code of conduct.
• I added an entry to CHANGES.md if knowledge of this change could be valuable to users.
• I ran cargo fmt

---

• Update top-level crate documentation with new information about geo-traits
• Add simple docstrings to all public functions
• Remove panic

[michaelkirk]

Should we make the switch to std::fmt::Write before releasing?

[kylebarron]

Should we make the switch to std::fmt::Write before releasing?

We already have std::fmt::Write. I was thinking more whether we should switch to std::io::Write. But it sounds like no?

[michaelkirk]

Originally I meant "we should go all in on fmt::Write and change the existing ToWkt::write_wkt to accept a fmt::Write".

But now I'm starting to think the opposite and that maybe we should change all the new API's to use io::Write.

The main sticking point for me: It seems like while I can stream WKT to a File (and friends) with io::Write, I can't do that with fmt::Write. I have to first build up the entire string in memory with the Fmt API and then write that entire string in one go.

For big geometries, being able to buffer the writes out as you process seems like a better way to go.

WDYT @kylebarron ?

[kylebarron]

The main sticking point for me: It seems like while I can stream WKT to a File (and friends) with io::Write, I can't do that with fmt::Write.

I think you can still stream WKT to io::Write with a small adapter: https://github.com/georust/wkt/pull/126/files#diff-afc819097d10bcb89581940115867e201017cd2bc9413c58f01b3dd610170f83R39-R49

Every call to write_str gets translated into the underlying io::Write::write, right? So I don't think there's any buffering still.

It's mostly a little ugly that you lost the content of any io::Error because that gets mapped to a fmt::Error.

[michaelkirk]

I think you can still stream WKT to io::Write with a small adapter: https://github.com/georust/wkt/pull/126/files#diff-afc819097d10bcb89581940115867e201017cd2bc9413c58f01b3dd610170f83R39-R49

You're right again!

It's mostly a little ugly that you lost the content of any io::Error because that gets mapped to a fmt::Error.

I might see a way around this... after I wrap up some georust/geo stuff, I'll take a swing at it.

[michaelkirk]

LGTM

Anything else you want to include before releasing @kylebarron or anyone else?

Otherwise, I'll merge and release this tomorrow to give anybody else a final chance to weigh in.

[kylebarron]

It looks good to me! #125 would be nice to get to in the medium term, but I don't think it should block a release.

[michaelkirk]

released: https://crates.io/crates/wkt/0.12.0
tagged: https://github.com/georust/wkt/releases/tag/0.12.0