Expose parse_format_string to external users to allow for attempting to typecast raw data

[DaBs]

Heya, very neat to have this module!

We don't have the correct types initially for e.g. char types, as we're using this for compressed logging, so all of our "arguments" start out as u8 types, and we don't have the original type available for each argument. By exposing the format parts, we get to use them to attempt to create / typecast the different arguments to the correct types to then finally cast as &dyn Printf.

[tjol]

I suppose I don't see the harm in making this public.

If the function is public it should probably have a doc comment

[DaBs]

Sorry for the late reply - I've updated the PR with a doc comment now.
LMK if there's anything else to change.

[tjol]

No need to apologize. Clearly, I'm worse.

(Sorry for leaving this PR hanging and not even approving the CI run, my private life has actually been quite hectic recently)

It looks like the doctest doesn't compile right now

[DaBs]

All good - looks like I mis-formatted the doc comment. Should run correctly now.
On the subject of tests, wondering if it makes sense to add some for this as well now, seeing as it's a public facing API. But it's indirectly tested via the existing tests I suppose.

[tjol]

So I finally got around to releasing the change. (Hopeless, I know)

After reviewing the docs, I shifted things around a bit and decided to make the parser module public instead of pulling everything into the crate namespace. So be aware that there's a breaking API change between the recent git HEAD and releae 0.2.0.