Faster dependency copying

[systay]

Description

This PR introduces a small but valuable optimization to how dependencies are copied, resulting in performance improvements.

1. During tree walking, we clone parts of the AST and copy dependency/type info for each node. For example:

   output := sqlparser.CopyOnRewrite(in, pre, post, ctx.SemTable.CopySemanticInfo)

2. Here, CopySemanticInfo is called for each cloned node and internally invokes CopyDependencies.
3. Currently, CopyDependencies calculates dependencies by walking the entire subtree and then stores the results in the semantic table. This extra work is unnecessary—it can be done on demand.
4. This PR changes the logic to copy only the existing info for the specific node, rather than re-walking and recalculating for its entire subtree.

Stats

                                 │ ../main.bench │     ../smart-deps-copy.bench      │
                                 │    sec/op     │   sec/op     vs base              │
OLTP-8                               229.2µ ± 3%   222.5µ ± 1%  -2.92% (p=0.002 n=6)
TPCC-8                               1.643m ± 1%   1.591m ± 1%  -3.16% (p=0.002 n=6)
TPCH-8                              10.114m ± 2%   9.581m ± 2%  -5.27% (p=0.002 n=6)
Planner/from_cases.json-gen4-8       7.213m ± 2%   6.924m ± 1%  -4.00% (p=0.002 n=6)
Planner/filter_cases.json-gen4-8     152.2m ± 1%   148.0m ± 0%  -2.77% (p=0.002 n=6)
Planner/large_cases.json-gen4-8      468.5µ ± 2%   450.0µ ± 1%  -3.95% (p=0.002 n=6)
Planner/aggr_cases.json-gen4-8       12.51m ± 0%   12.28m ± 1%  -1.84% (p=0.002 n=6)
Planner/select_cases.json-gen4-8    10.016m ± 1%   9.828m ± 1%  -1.88% (p=0.002 n=6)
Planner/union_cases.json-gen4-8      3.552m ± 0%   3.560m ± 1%       ~ (p=0.485 n=6)
SemAnalysis-8                        27.73m ± 0%   27.86m ± 0%  +0.48% (p=0.002 n=6)
BaselineVsMirrored/Baseline-8        113.7µ ± 3%   114.3µ ± 2%       ~ (p=0.589 n=6)
BaselineVsMirrored/Mirrored-8        319.4µ ± 3%   313.0µ ± 0%  -2.00% (p=0.002 n=6)
geomean                              3.128m        3.058m       -2.23%

                                 │ ../main.bench │      ../smart-deps-copy.bench      │
                                 │     B/op      │     B/op      vs base              │
OLTP-8                              151.2Ki ± 0%   151.2Ki ± 0%  -0.01% (p=0.002 n=6)
TPCC-8                              993.6Ki ± 0%   993.2Ki ± 0%  -0.04% (p=0.002 n=6)
TPCH-8                              5.226Mi ± 0%   5.034Mi ± 0%  -3.67% (p=0.002 n=6)
Planner/from_cases.json-gen4-8      4.419Mi ± 0%   4.396Mi ± 0%  -0.52% (p=0.002 n=6)
Planner/filter_cases.json-gen4-8    19.36Mi ± 0%   19.28Mi ± 0%  -0.42% (p=0.002 n=6)
Planner/large_cases.json-gen4-8     273.7Ki ± 0%   273.7Ki ± 0%  -0.02% (p=0.004 n=6)
Planner/aggr_cases.json-gen4-8      7.128Mi ± 0%   6.964Mi ± 0%  -2.29% (p=0.002 n=6)
Planner/select_cases.json-gen4-8    6.142Mi ± 0%   6.003Mi ± 0%  -2.27% (p=0.002 n=6)
Planner/union_cases.json-gen4-8     2.188Mi ± 0%   2.187Mi ± 0%  -0.07% (p=0.002 n=6)
SemAnalysis-8                       18.98Mi ± 0%   18.98Mi ± 0%       ~ (p=0.310 n=6)
BaselineVsMirrored/Baseline-8       81.69Ki ± 0%   81.68Ki ± 0%  -0.02% (p=0.002 n=6)
BaselineVsMirrored/Mirrored-8       201.5Ki ± 0%   201.5Ki ± 0%  -0.03% (p=0.002 n=6)
geomean                             1.675Mi        1.662Mi       -0.79%

                                 │ ../main.bench │      ../smart-deps-copy.bench       │
                                 │   allocs/op   │  allocs/op   vs base                │
OLTP-8                               3.705k ± 0%   3.705k ± 0%       ~ (p=1.000 n=6) ¹
TPCC-8                               23.71k ± 0%   23.71k ± 0%       ~ (p=0.455 n=6)
TPCH-8                               131.8k ± 0%   131.6k ± 0%  -0.22% (p=0.002 n=6)
Planner/from_cases.json-gen4-8       102.5k ± 0%   102.5k ± 0%  -0.04% (p=0.002 n=6)
Planner/filter_cases.json-gen4-8     508.9k ± 0%   508.7k ± 0%  -0.03% (p=0.002 n=6)
Planner/large_cases.json-gen4-8      10.61k ± 0%   10.61k ± 0%       ~ (p=1.000 n=6) ¹
Planner/aggr_cases.json-gen4-8       160.4k ± 0%   159.9k ± 0%  -0.32% (p=0.002 n=6)
Planner/select_cases.json-gen4-8     141.0k ± 0%   140.6k ± 0%  -0.26% (p=0.002 n=6)
Planner/union_cases.json-gen4-8      50.66k ± 0%   50.65k ± 0%  -0.00% (p=0.002 n=6)
SemAnalysis-8                        231.3k ± 0%   231.3k ± 0%       ~ (p=0.323 n=6)
BaselineVsMirrored/Baseline-8        2.017k ± 0%   2.017k ± 0%       ~ (p=1.000 n=6) ¹
BaselineVsMirrored/Mirrored-8        5.578k ± 0%   5.578k ± 0%       ~ (p=1.000 n=6) ¹
geomean                              40.44k        40.41k       -0.07%

Related Issue(s)

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

All modified and coverable lines are covered by tests ✅

Project coverage is 67.96%. Comparing base (747c308) to head (007a189).
Report is 5 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #17708      +/-   ##
==========================================
+ Coverage   67.79%   67.96%   +0.17%     
==========================================
  Files        1587     1586       -1     
  Lines      255829   255217     -612     
==========================================
+ Hits       173427   173455      +28     
+ Misses      82402    81762     -640     

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