assert: improve partialDeepStrictEqual performance and add benchmark

[puskin]

now that #54630 has landed, I took the liberty to review the code I wrote to look for improvements that would squeeze some more performance out of the comparison mechanism

• simplified some logic
• reviewed loops and introduced more breaking points to make them run for less cycles
• added a benchmark file to test future possible regressions when it comes to partialDeepStrictEqual

Refs: #54630

[nodejs-github-bot]

Review requested:

• @nodejs/performance

[codecov]

Codecov Report

Attention: Patch coverage is 57.74648% with 30 lines in your changes missing coverage. Please review.

Project coverage is 90.23%. Comparing base (6b3937a) to head (4502b76).
Report is 515 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/assert.js	57.74%	30 Missing ⚠️

Additional details and impacted files

@@ Coverage Diff @@## main #56555 +/- ##
==========================================
+ Coverage 89.17% 90.23% +1.05% 
==========================================
Files 662 630 -32 Lines 191672 184968 -6704 Branches 36884 36192 -692 ==========================================
- Hits 170922 166904 -4018 + Misses 13614 11084 -2530 + Partials 7136 6980 -156 

Files with missing lines	Coverage Δ
lib/assert.js	96.50% <57.74%> (-2.54%)	⬇️

... and 236 files with indirect coverage changes

🚀 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.

[puskin]

Is anything else needed here?

[lemire]

@puskin94 Can you share the results of your benchmarking?

[puskin]

@puskin94 Can you share the results of your benchmarking?

of course 😄

➜ node git:(partial-deep-strict-equal-perf) ./node --no-warnings benchmark/assert/partial-deep-strict-equal.js assert/partial-deep-strict-equal.js datasetName="objects" size=1000 n=50: 169.87881572367604
assert/partial-deep-strict-equal.js datasetName="sets" size=1000 n=50: 141.24912248982653
assert/partial-deep-strict-equal.js datasetName="maps" size=1000 n=50: 151.36189759383777
assert/partial-deep-strict-equal.js datasetName="arrayBuffers" size=1000 n=50: 14.16071061277997
assert/partial-deep-strict-equal.js datasetName="dataViewArrayBuffers" size=1000 n=50: 16.398050206238278

[aduh95]

If we're looking for performance, we should probably be using a classic for(;;) loop. I'd recommend rolling back this change (i.e. put back the safeExpected assignment, because it has nothing to do with performance) for this PR, and make a new one switching to a for(;;) loop on which we can run benchmarks

[puskin]

if you do this tho you would have to call the ArrayFrom primordial. The performance enhancement was to adjust the "low hanging fruit" stuff I could get, "structure" wise, without influencing readability. IMHO the current implementation is better, but happy to change if you don't agree!

reverted the assignment tho

[puskin]

Is anything else needed here?

[puskin]

is any change required here?

[aduh95]

@BridgeAR can you clarify whether you're still blocking?

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/65363/

[mcollina]

lgtm

[BridgeAR]

Sorry for the late reply. I guess this improves a couple of cases, so that's great! :)
I believe there is more room for improvement by reusing the existing implementation. That could be used for a follow-up PR.
Before landing: could we run the benchmark to see the improvement as percentage? :)

Just did a new review

[aduh95]

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1664/

Results

 confidence improvement accuracy (*) (**) (***)
assert/partial-deep-strict-equal.js datasetName='arrayBuffers' size=1000 n=50 *** -0.79 % ±0.38% ±0.51% ±0.67%
assert/partial-deep-strict-equal.js datasetName='dataViewArrayBuffers' size=1000 n=50 ** 0.64 % ±0.45% ±0.60% ±0.78%
assert/partial-deep-strict-equal.js datasetName='maps' size=1000 n=50 *** 0.56 % ±0.26% ±0.35% ±0.46%
assert/partial-deep-strict-equal.js datasetName='objects' size=1000 n=50 *** -0.49 % ±0.28% ±0.37% ±0.48%
assert/partial-deep-strict-equal.js datasetName='sets' size=1000 n=50 * 0.34 % ±0.26% ±0.35% ±0.45%
Be aware that when doing many comparisons the risk of a false-positive
result increases. In this case, there are 5 comparisons, you can thus
expect the following amount of false-positive results:
0.25 false positives, when considering a 5% risk acceptance (*, **, ***),
0.05 false positives, when considering a 1% risk acceptance (**, ***),
0.01 false positives, when considering a 0.1% risk acceptance (***)

[puskin]

I must have messed up something with my local codebase. I just pushed again with new performance improvements, this should make some difference

[aduh95]

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1665/

[aduh95]

Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1666/

[BridgeAR]

New benchmark

https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1667/

[BridgeAR]

If I am not mistaken, this mainly improves the performance for sets and maps that have mixed entries (primitives + objects). The current benchmark has no such entry.

I also guess that the primordial usage of the iterator is costly. Could you check if that's the case?

[BridgeAR]

Latest benchmark https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1671

(*) (**) (***)
00:25:02 assert/partial-deep-strict-equal.js datasetName='arrayBuffers' size=1000 n=10 *** 0.67 % ±0.37% ±0.49% ±0.64%
00:25:02 assert/partial-deep-strict-equal.js datasetName='arrayBuffers' size=1000 n=200 *** 0.87 % ±0.33% ±0.45% ±0.58%
00:25:02 assert/partial-deep-strict-equal.js datasetName='arrayBuffers' size=1000 n=50 *** 1.01 % ±0.40% ±0.54% ±0.70%
00:25:02 assert/partial-deep-strict-equal.js datasetName='circularRefs' size=1000 n=200 *** -0.75 % ±0.38% ±0.50% ±0.65%
00:25:02 assert/partial-deep-strict-equal.js datasetName='maps' size=1000 n=50 *** -0.48 % ±0.27% ±0.36% ±0.46%
00:25:02 assert/partial-deep-strict-equal.js datasetName='objects' size=1000 n=200 *** 0.48 % ±0.23% ±0.30% ±0.40%
00:25:02 assert/partial-deep-strict-equal.js datasetName='objects' size=1000 n=50 *** 0.73 % ±0.28% ±0.38% ±0.49%
00:25:02 assert/partial-deep-strict-equal.js datasetName='typedArrays' size=1000 n=50 *** 0.55 % ±0.31% ±0.41% ±0.53%

Only *** result listed. Looking at the changes and the benchmark outcome, I guess that most work is spend elsewhere. I suggest to land this as is nevertheless.

[aduh95]

This needs a rebase

[BridgeAR]

This was superseded by the other PR

[puskin]

superseeded by #57370