parser: optimize performance

[lauhg]

What problem does this PR solve?

Optimizing parser performance

What is changed and how it works?

Here is the benchmark, 8%~12% faster than before:

goos: linux
goarch: amd64

Before optimization:

BenchmarkSysbenchSelect       200000          6121 ns/op        1824 B/op         20 allocs/op
BenchmarkParseComplex          10000        223586 ns/op        61772 B/op        642 allocs/op
BenchmarkParseSimple          100000         19047 ns/op        5144 B/op         71 allocs/op

After optimization:

BenchmarkSysbenchSelect       200000          5586 ns/op        1824 B/op         20 allocs/op
BenchmarkParseComplex          10000        206988 ns/op        61772 B/op        642 allocs/op
BenchmarkParseSimple          100000         16873 ns/op        5144 B/op         71 allocs/op

Check List

Tests

Unit test

[codecov]

Codecov Report

Merging #562 into master will increase coverage by <.01%.
The diff coverage is 40%.

@@            Coverage Diff             @@
##           master     #562      +/-   ##
==========================================
+ Coverage   71.76%   71.77%   +<.01%     
==========================================
  Files          32       32              
  Lines        7813     7811       -2     
==========================================
- Hits         5607     5606       -1     
+ Misses       1676     1675       -1     
  Partials      530      530

Impacted Files	Coverage Δ
yy_parser.go	83.46% <ø> (ø)	⬆️
goyacc/main.go	0% <0%> (ø)	⬆️
parser.go	70.37% <80%> (-0.22%)	⬇️
ast/functions.go	76.74% <0%> (ø)	⬆️

---

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 33636bc...6a0bfe2. Read the comment docs.

[tiancaiamao]

LGTM
I've tested this branch with TiDB and the CI passed.

[tiancaiamao]

PTAL @winkyao @zz-jason

[zz-jason]

LGTM