Improve parse perf using bitwise flags

[blakeembrey]

Before:

 ✓ src/index.bench.ts > parse 17104ms
     name                                                       hz     min     max    mean     p75     p99    p995    p999     rme  samples
   · basic                                           16,113,888.97  0.0000  1.0510  0.0001  0.0001  0.0001  0.0002  0.0002  ±0.70%  8056945
   · simple parameters                                2,838,935.69  0.0002  0.3220  0.0004  0.0004  0.0004  0.0005  0.0009  ±0.40%  1419468
   · single parameter                                 5,456,431.11  0.0001  0.0475  0.0002  0.0002  0.0003  0.0003  0.0004  ±0.11%  2728216
   · uppercase single parameter                       5,453,110.96  0.0001  0.3120  0.0002  0.0002  0.0002  0.0003  0.0005  ±0.16%  2726556
   · accept header (comma = true)                     6,396,519.98  0.0000  0.0564  0.0002  0.0002  0.0002  0.0002  0.0003  ±0.09%  3198261
   · accept header (comma = false)                    3,673,513.63  0.0002  0.3119  0.0003  0.0003  0.0003  0.0004  0.0008  ±0.36%  1836757
   · simple quoted parameter                          5,080,187.10  0.0001  0.2890  0.0002  0.0002  0.0002  0.0003  0.0006  ±0.29%  2540094
   · simple parameters (options.parameters = false)  11,690,130.60  0.0000  0.3273  0.0001  0.0001  0.0001  0.0001  0.0002  ±0.18%  5845066
   · quoted and escaped parameters                    2,313,647.64  0.0003  0.2310  0.0004  0.0004  0.0005  0.0006  0.0035  ±0.31%  1156824
   · OWS-heavy parameters                             2,609,728.36  0.0002  0.2555  0.0004  0.0004  0.0005  0.0005  0.0009  ±0.29%  1304865
   · many parameters                                  1,805,848.65  0.0004  0.2545  0.0006  0.0005  0.0007  0.0007  0.0011  ±0.27%   902925
   · escape-heavy quoted parameters                   1,458,234.16  0.0005  0.2559  0.0007  0.0007  0.0008  0.0011  0.0014  ±0.36%   729118

After:

 ✓ src/index.bench.ts > parse 19407ms
     name                                                       hz     min      max    mean     p75     p99    p995    p999      rme   samples
   · basic                                           20,056,451.24  0.0000   0.0883  0.0000  0.0000  0.0001  0.0001  0.0002   ±0.10%  10028226
   · simple parameters                                3,340,430.35  0.0002   4.5476  0.0003  0.0003  0.0004  0.0004  0.0009   ±1.85%   1670216
   · single parameter                                 5,666,319.18  0.0000  12.1667  0.0002  0.0002  0.0003  0.0004  0.0005   ±6.64%   2833160
   · uppercase single parameter                       5,323,281.24  0.0001   0.4540  0.0002  0.0002  0.0003  0.0003  0.0009   ±0.35%   2661641
   · accept header (comma = true)                     7,068,431.77  0.0000   0.0619  0.0001  0.0002  0.0002  0.0002  0.0003   ±0.14%   3534216
   · accept header (comma = false)                    4,006,771.94  0.0001   0.4999  0.0002  0.0003  0.0003  0.0003  0.0007   ±0.59%   2003386
   · simple quoted parameter                          7,277,198.79  0.0000   0.0597  0.0001  0.0002  0.0002  0.0002  0.0003   ±0.13%   3638600
   · simple parameters (options.parameters = false)  16,085,461.42  0.0000  34.7025  0.0001  0.0001  0.0001  0.0001  0.0002  ±13.60%   8042731
   · quoted and escaped parameters                    2,676,170.14  0.0002   0.3357  0.0004  0.0004  0.0004  0.0005  0.0009   ±0.38%   1338086
   · OWS-heavy parameters                             3,052,145.49  0.0002   0.3206  0.0003  0.0003  0.0004  0.0004  0.0008   ±0.19%   1526073
   · many parameters                                  2,167,994.21  0.0003   0.3490  0.0005  0.0005  0.0006  0.0006  0.0017   ±0.20%   1083998
   · escape-heavy quoted parameters                   1,489,250.93  0.0005   0.4189  0.0007  0.0007  0.0008  0.0010  0.0017   ±0.46%    744626

I think I might be able to combine these flags with the validation logic to also improve that performance.

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 100.00%. Comparing base (04b9688) to head (d7850d3).

Additional details and impacted files

@@            Coverage Diff            @@
##            master       #76   +/-   ##
=========================================
  Coverage   100.00%   100.00%           
=========================================
  Files            1         1           
  Lines           95       155   +60     
  Branches        20        40   +20     
=========================================
+ Hits            95       155   +60     

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

🚀 New features to boost your workflow:

• ❄️ Test Analytics: Detect flaky tests, report on failures, and find test suite problems.
• 📦 JS Bundle Analysis: Save yourself from yourself by tracking and limiting bundle sizes in JS merges.

[blakeembrey]

Inlined skipValue, skipOWS and removing trailingOWS in favor of tracking the last whitespace showed marginal improvements to performance (a few % each).

[blakeembrey]

Unquoting seemed like a large perf improvement with the right conditions (most won't have escapes, turns into a single slice).

[blakeembrey]

Skipping toLowerCase was the largest perf win when the value is already lower cased.