Add full UTF-8 support in RON incl. unicode identifiers

[juntyr]

Fixes #321

Superseded #444, which it rebases + adds 1aee993 to ensure that ron is always valid UTF-8

• I've included my change in CHANGELOG.md

[codecov-commenter]

Codecov Report

Patch coverage: 100.00% and project coverage change: +2.01% 🎉

Comparison is base (f693d7e) 96.12% compared to head (b14096a) 98.14%.

❗ Your organization is not using the GitHub App Integration. As a result you may experience degraded service beginning May 15th. Please install the GitHub App Integration for your organization. Read more.

Additional details and impacted files

@@            Coverage Diff             @@
##           master     #488      +/-   ##
==========================================
+ Coverage   96.12%   98.14%   +2.01%     
==========================================
  Files          77       78       +1     
  Lines        9890    10308     +418     
==========================================
+ Hits         9507    10117     +610     
+ Misses        383      191     -192     

Files Changed	Coverage Δ
src/value/mod.rs	94.87% <ø> (-0.10%)	⬇️
tests/152_bitflags.rs	100.00% <ø> (+33.33%)	⬆️
tests/struct_integers.rs	100.00% <ø> (ø)
src/de/id.rs	100.00% <100.00%> (ø)
src/de/mod.rs	98.63% <100.00%> (+1.74%)	⬆️
src/de/tag.rs	100.00% <100.00%> (+17.15%)	⬆️
src/de/tests.rs	100.00% <100.00%> (ø)
src/de/value.rs	97.96% <100.00%> (+0.20%)	⬆️
src/error.rs	53.62% <100.00%> (+7.52%)	⬆️
src/extensions.rs	100.00% <100.00%> (+18.42%)	⬆️
... and 22 more

☔ View full report in Codecov by Sentry.
📢 Have feedback on the report? Share it here.

[juntyr]

@torkleyy I've finally gotten to rebasing #444. I would propose merging it after #438 lands - while I will try to find some performance improvements earlier, I think parsing in UTF8 cleans up the code so much that landing early and optimising in later commits could also be warranted.

The last commit, 93f367c, adds an API-breaking change to better encode that ron only produces valid UTF-8 during serialising (by switching from io::Write to fmt::Write). If we don't want the API breakage, we could also add our own trait that is implemented for String and anything that implements io::Write.

[juntyr]

@ModProg @torkleyy This PR is now (almost) ready to be merged. I'll review the changes again tomorrow and add a CHANGELOG entry. Apart from that, I think this is a very good new implementation (thank you so much @ModProg for doing most of the work towards it), which can be improved performance-wise in follow-up PRs if necessary

[juntyr]

@torkleyy We are also getting closer and closer to 100% code coverage ... maybe before releasing v0.9 that's the last thing after this PR that I take care of.

[juntyr]

Some small cleanup is needed