evalengine: Implement PERIOD_ADD

[beingnoble03]

Description

This PR adds implementation of PERIOD_ADD func in evalengine.

Related Issue(s)

• evalengine: Support all built-in MySQL functions #9647

Checklist

• "Backport to:" labels have been added if this change should be back-ported to release branches
• If this change is to be back-ported to previous releases, a justification is included in the PR description
• Tests were added or are not required
• Did the new or modified tests pass consistently locally and on CI?
• Documentation was added or is not required

Deployment Notes

[vitess-bot]

Review Checklist

Hello reviewers! 👋 Please follow this checklist when reviewing this Pull Request.

General

• Ensure that the Pull Request has a descriptive title.
• Ensure there is a link to an issue (except for internal cleanup and flaky test fixes), new features should have an RFC that documents use cases and test cases.

Tests

• Bug fixes should have at least one unit or end-to-end test, enhancement and new features should have a sufficient number of tests.

Documentation

• Apply the release notes (needs details) label if users need to know about this change.
• New features should be documented.
• There should be some code comments as to why things are implemented the way they are.
• There should be a comment at the top of each new or modified test to explain what the test does.

New flags

• Is this flag really necessary?
• Flag names must be clear and intuitive, use dashes (-), and have a clear help text.

If a workflow is added or modified:

• Each item in Jobs should be named in order to mark it as required.
• If the workflow needs to be marked as required, the maintainer team must be notified.

Backward compatibility

• Protobuf changes should be wire-compatible.
• Changes to _vt tables and RPCs need to be backward compatible.
• RPC changes should be compatible with vitess-operator
• If a flag is removed, then it should also be removed from vitess-operator and arewefastyet, if used there.
• vtctl command output order should be stable and awk-able.

[codecov]

Codecov Report

Attention: Patch coverage is 90.10989% with 9 lines in your changes missing coverage. Please review.

Project coverage is 68.75%. Comparing base (eb29999) to head (06d18ad).
Report is 71 commits behind head on main.

Files	Patch %	Lines
go/vt/vtgate/evalengine/fn_time.go	90.62%	3 Missing ⚠️
go/mysql/datetime/mydate.go	92.59%	2 Missing ⚠️
go/vt/vtgate/evalengine/compiler_asm.go	86.66%	2 Missing ⚠️
go/vt/vtgate/evalengine/translate_builtin.go	66.66%	2 Missing ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #16492      +/-   ##
==========================================
+ Coverage   68.69%   68.75%   +0.06%     
==========================================
  Files        1547     1556       +9     
  Lines      198297   199798    +1501     
==========================================
+ Hits       136228   137380    +1152     
- Misses      62069    62418     +349     

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

[systay]

this seems to be the only use of MonthsToPeriod and PeriodToMonths. Wouldn't it make sense to have it accept int64 in, and return int64 values? that way we wouldn't have to cast them here

[dbussink]

Yeah, not a huge issue but let's make it consistent so we don't cast here. @beingnoble03 Any reason it must be using an unsigned type?

[beingnoble03]

Done. If we want to avoid type-casting here, we would have to type-cast it in MonthsToPeriod and PeriodToMonths. Also, we'll using these functions for PERIOD_DIFF, so this will avoid type-casting it there too.

@dbussink mysql-server uses uint64 for these functions. Results would differ if we don't type-cast it into uint64. (https://github.com/mysql/mysql-server/blob/trunk/mysys/my_time.cc#L2254)

[dbussink]

@dbussink mysql-server uses uint64 for these functions. Results would differ if we don't type-cast it into uint64. (https://github.com/mysql/mysql-server/blob/trunk/mysys/my_time.cc#L2254)

I don't think that's very relevant here since those are values way out of bounds anyway?

[beingnoble03]

But months can be negative here. So, we'll get different results for some negative cases e.g. PERIOD_ADD(1206, -1231232332)

[beingnoble03]

@dbussink We already run FnPeriodAdd in cases.go through some of these overflow cases (values in inputBitwise in inputs.go), which fail when we remove type-casting. Should we include more of these cases specifically in FnPeriodAdd?

[dbussink]

@beingnoble03 I guess I'm confused how CI here passes then after the latest changes? Or are you talking about additional cleanups here?

[beingnoble03]

@dbussink I type casted it inside the function in the latest changes. That's why it's passing.

func PeriodToMonths(period int64) int64 {
	p := uint64(period)

func MonthsToPeriod(months int64) int64 {
	m := uint64(months)

The tests will fail, if this type casting is removed. I think I may have misunderstood. Did you mean something else?

[dbussink]

Ah yeah, that makes sense. I think we should cast it there where it makes most sense. If we also add PERIOD_DIFF, were should we put it you think?

[beingnoble03]

@dbussink I think it will be best to type cast inside the function (the way it's done now). That would eliminate the need to type cast separately. I can create a PR for PERIOD_DIFF once this is merged.