* add err define of mysql to moerr

* remove struct of mysql error

Co-authored-by: mergify[bot] <37929162+mergify[bot]@users.noreply.github.com>
Co-authored-by: fengttt <fengttt@gmail.com>

issue: 4867

Approved by: @fengttt

Declaration of supplementary C function

Approved by: @nnsgmsone, @fengttt

Convert the result of decimal64 sum aggregation to decimal128

Approved by: @nnsgmsone, @aunjgr, @fengttt

* aggrate sum decimal

* sum with decimal

* modify aggrate sum code

Rework logical operations with cgo.

Approved by: @daviszhen, @fengttt, @ouyuanning, @nnsgmsone, @aunjgr

Using CGO to reconstruct the comparison operation

Approved by: @daviszhen, @ouyuanning, @fengttt, @nnsgmsone, @aunjgr

Rework arithmetic operations and add overflow detection

Approved by: @fengttt, @daviszhen

Approved by: @fengttt

* simplify Makefile in CGO

* remove gcc -M option

* extract decNumber Makefile

* Rewrite decimal function macro.

* Move int/uint/float add to cgo.  Benchmarked.

For 8K int/uint/float operations, go perf is roughly 10us.   Cgo
is roughly 2us.

For 8K dec64/dec128, right now it is 100 - 200 us. This may argue
for our own implementation of decimal.

Plus first.  If it looks ok, I will work out the rest.

intermediate libxxx.a files are no longer kept

- add license header checker
- fix  license header for files
- enable full static check

* Adding decNumber library.

* Work on decimal and rebase.

I will be damned if it works at all.

* Pass ut.

* Will never pass bvt.

* Fixes.

* Missed a few UT cases.

I cheated on a parser case.   Decimal will output 100 as 1E+2.
I think we can accept this.

* Use PRId64 and PRIu64 in printf.

* Minor clean up.

* CI

added -ldecnumber, removed pdf file.

* Fix a static check error.

* CGO fun.

* CGO fun. (#6)

Co-authored-by: Feng Tian <fengttt@gmail.com>

* Update Makefile

enable cgo for static-check

* Update run_ut.sh

enable cgo for ut

Co-authored-by: maomao <77312370+sukki37@users.noreply.github.com>