introduce hasKnownResult to optimize several parsers.

[johnynek]

Often, we can statically see what the result value will be if a parser succeeds.

If we know this, can optimize map, select and flatMap.

In particular, we can avoid reallocating a parser inside each flatMap.

Most users should avoid using flatMap anyway, but in a library function, you might need flatMap but want to get the optimization when the user passes you a constant parser.

[codecov-commenter]

Codecov Report

Merging #348 (faa8f8f) into main (3acc967) will increase coverage by 0.34%.
The diff coverage is 98.75%.

@@            Coverage Diff             @@
##             main     #348      +/-   ##
==========================================
+ Coverage   96.56%   96.91%   +0.34%     
==========================================
  Files           9        9              
  Lines        1049     1103      +54     
  Branches       94       98       +4     
==========================================
+ Hits         1013     1069      +56     
+ Misses         36       34       -2     

Impacted Files	Coverage Δ
core/shared/src/main/scala/cats/parse/Parser.scala	96.73% <98.75%> (+0.31%)	⬆️
...shared/src/main/scala/cats/parse/Accumulator.scala	100.00% <0.00%> (+2.43%)	⬆️

---

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 3acc967...faa8f8f. Read the comment docs.

[regadas]

It will be interesting to see new benchmark numbers.

[regadas]

@johnynek could you rebase this?

[johnynek]

Yes! I will do. Thanks for the review.

Note, this optimization will mostly apply to flatMap, and most parser don't need flatMap (e.g. our json benchmarks don't use flatMap). But still, I thought it was worth trying to limit the worst case.

[johnynek]

@regadas let me know if this looks good.

[regadas]

Looks great!