Improve benchmark suite

[dralley]

No description provided.

[codecov-commenter]

Codecov Report

Merging #407 (f2afd0b) into master (f6d5e99) will decrease coverage by 8.48%.
The diff coverage is 97.09%.

@@            Coverage Diff             @@
##           master     #407      +/-   ##
==========================================
- Coverage   63.84%   55.35%   -8.49%     
==========================================
  Files          20       22       +2     
  Lines       10870    13863    +2993     
==========================================
+ Hits         6940     7674     +734     
- Misses       3930     6189    +2259     

Flag	Coverage Δ
unittests	55.35% <97.09%> (-8.49%)	⬇️

Flags with carried forward coverage won't be shown. Click here to find out more.

Impacted Files	Coverage Δ
benches/microbenches.rs	96.26% <96.26%> (ø)
benches/macrobenches.rs	100.00% <100.00%> (ø)
src/writer.rs	49.01% <0.00%> (-41.36%)	⬇️
src/reader.rs	63.46% <0.00%> (-25.54%)	⬇️
src/utils.rs	66.95% <0.00%> (-14.96%)	⬇️
src/errors.rs	1.73% <0.00%> (-7.80%)	⬇️
src/escapei.rs	13.96% <0.00%> (+0.05%)	⬆️
src/lib.rs	22.63% <0.00%> (+1.24%)	⬆️

---

Continue to review full report at Codecov.

Legend - Click here to learn more
Δ = absolute <relative> (impact), ø = not affected, ? = missing data
Powered by Codecov. Last update f6d5e99...f2afd0b. Read the comment docs.

[Mingun]

I spot some minor nits, but it is good enough. Feel free to merge it as is

[dralley]

Changed:

• Added a changelog entry
• Split the cargo fmt off into a separate lint CI job that runs once
• Reverted the bench compare stuff and the change to that one benchmark that wasn't needed
• Removed code coverage tracking from cargo test --benches since it would be a bit misleading
• Added a README describing (some of) the documents in /tests/documents/