-
Notifications
You must be signed in to change notification settings - Fork 29.7k
New issue
Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.
By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.
Already on GitHub? Sign in to your account
fs: refactoring declaratively with the Array.fromAsync function #54644
Conversation
4dd3a9e
to
07d075a
Compare
Codecov ReportAll modified and coverable lines are covered by tests ✅
Additional details and impacted files@@ Coverage Diff @@
## main #54644 +/- ##
==========================================
+ Coverage 87.34% 88.24% +0.90%
==========================================
Files 649 651 +2
Lines 182524 183875 +1351
Branches 35026 35862 +836
==========================================
+ Hits 159420 162264 +2844
+ Misses 16373 14896 -1477
+ Partials 6731 6715 -16
|
for await (const entry of new Glob(pattern, options).glob()) { | ||
ArrayPrototypePush(res, entry); | ||
} | ||
const res = await ArrayFromAsync(new Glob(pattern, options).glob()); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I'm not sure if there is a benchmark for this, but this should probably be benchmarked.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev Thanks for the review!
There are currently no benchmarks, so I'll add them!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev
Sorry for being late!
I added the fs glob benchmark f4025e5.
The benchmark results are as follows:
node-benchmark-compare compare-glob-result.csv
confidence improvement accuracy (*) (**) (***)
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='benchmark' n=1 0.68 % ±1.50% ±2.01% ±2.62%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='benchmark' n=10 0.47 % ±1.59% ±2.11% ±2.75%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='benchmark' n=100 * 0.68 % ±0.63% ±0.84% ±1.10%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='lib' n=1 -0.40 % ±1.61% ±2.15% ±2.82%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='lib' n=10 0.23 % ±1.31% ±1.74% ±2.27%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='lib' n=100 1.67 % ±1.67% ±2.24% ±2.96%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='test/parallel' n=1 0.77 % ±0.83% ±1.11% ±1.46%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='test/parallel' n=10 *** 3.75 % ±1.90% ±2.55% ±3.37%
fs/bench-glob.js recursive='false' mode='async' pattern='*.js' dir='test/parallel' n=100 * 1.51 % ±1.16% ±1.54% ±2.01%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='benchmark' n=1 0.13 % ±1.53% ±2.04% ±2.67%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='benchmark' n=10 1.42 % ±1.64% ±2.20% ±2.91%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='benchmark' n=100 0.27 % ±0.53% ±0.71% ±0.93%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='lib' n=1 *** 1.39 % ±0.75% ±1.00% ±1.31%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='lib' n=10 -3.22 % ±5.11% ±6.88% ±9.13%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='lib' n=100 0.05 % ±0.78% ±1.04% ±1.35%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='test/parallel' n=1 1.09 % ±1.37% ±1.82% ±2.38%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='test/parallel' n=10 *** 5.67 % ±1.26% ±1.69% ±2.23%
fs/bench-glob.js recursive='false' mode='async' pattern='**/*' dir='test/parallel' n=100 1.22 % ±1.55% ±2.07% ±2.72%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='benchmark' n=1 0.39 % ±3.01% ±4.03% ±5.27%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='benchmark' n=10 0.72 % ±1.31% ±1.74% ±2.27%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='benchmark' n=100 0.35 % ±1.11% ±1.49% ±1.97%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='lib' n=1 1.28 % ±1.48% ±1.99% ±2.62%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='lib' n=10 0.33 % ±1.02% ±1.35% ±1.76%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='lib' n=100 0.03 % ±0.71% ±0.94% ±1.22%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='test/parallel' n=1 * 1.49 % ±1.44% ±1.92% ±2.51%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='test/parallel' n=10 *** 7.41 % ±2.81% ±3.79% ±5.03%
fs/bench-glob.js recursive='false' mode='async' pattern='**/**.js' dir='test/parallel' n=100 *** 1.56 % ±0.75% ±1.00% ±1.31%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='benchmark' n=1 -0.43 % ±1.26% ±1.68% ±2.19%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='benchmark' n=10 -0.12 % ±1.12% ±1.49% ±1.94%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='benchmark' n=100 -0.10 % ±0.91% ±1.21% ±1.58%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='lib' n=1 2.79 % ±4.15% ±5.59% ±7.39%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='lib' n=10 0.12 % ±2.12% ±2.83% ±3.70%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='lib' n=100 0.07 % ±3.03% ±4.04% ±5.26%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='test/parallel' n=1 0.41 % ±1.25% ±1.66% ±2.17%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='test/parallel' n=10 0.79 % ±2.11% ±2.81% ±3.66%
fs/bench-glob.js recursive='false' mode='sync' pattern='*.js' dir='test/parallel' n=100 0.08 % ±0.92% ±1.22% ±1.59%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='benchmark' n=1 -1.47 % ±4.43% ±5.96% ±7.90%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='benchmark' n=10 0.49 % ±0.51% ±0.68% ±0.89%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='benchmark' n=100 1.15 % ±1.23% ±1.66% ±2.19%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='lib' n=1 0.57 % ±0.91% ±1.21% ±1.58%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='lib' n=10 -1.87 % ±4.44% ±5.97% ±7.90%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='lib' n=100 2.01 % ±3.92% ±5.29% ±7.01%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='test/parallel' n=1 -0.10 % ±0.88% ±1.18% ±1.54%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='test/parallel' n=10 -0.12 % ±0.83% ±1.11% ±1.45%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/*' dir='test/parallel' n=100 -0.17 % ±0.79% ±1.06% ±1.38%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='benchmark' n=1 * 1.12 % ±1.08% ±1.44% ±1.88%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='benchmark' n=10 1.85 % ±2.73% ±3.65% ±4.80%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='benchmark' n=100 0.16 % ±1.07% ±1.42% ±1.85%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='lib' n=1 0.54 % ±1.06% ±1.42% ±1.84%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='lib' n=10 -0.62 % ±1.21% ±1.62% ±2.14%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='lib' n=100 0.66 % ±1.14% ±1.52% ±2.00%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='test/parallel' n=1 -2.67 % ±5.08% ±6.85% ±9.09%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='test/parallel' n=10 -0.12 % ±0.73% ±0.97% ±1.26%
fs/bench-glob.js recursive='false' mode='sync' pattern='**/**.js' dir='test/parallel' n=100 0.55 % ±0.60% ±0.80% ±1.05%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='benchmark' n=1 0.22 % ±1.42% ±1.90% ±2.49%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='benchmark' n=10 0.08 % ±2.60% ±3.50% ±4.64%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='benchmark' n=100 0.08 % ±0.78% ±1.03% ±1.35%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='lib' n=1 -0.03 % ±1.26% ±1.68% ±2.20%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='lib' n=10 0.38 % ±2.30% ±3.07% ±4.03%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='lib' n=100 2.81 % ±3.46% ±4.65% ±6.16%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='test/parallel' n=1 *** 1.77 % ±0.50% ±0.66% ±0.86%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='test/parallel' n=10 *** 4.75 % ±1.45% ±1.93% ±2.52%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='test/parallel' n=100 0.26 % ±1.28% ±1.72% ±2.27%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='benchmark' n=1 0.45 % ±1.22% ±1.63% ±2.13%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='benchmark' n=10 0.31 % ±0.81% ±1.09% ±1.42%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='benchmark' n=100 -0.18 % ±0.84% ±1.13% ±1.47%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='lib' n=1 0.21 % ±1.48% ±1.97% ±2.56%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='lib' n=10 -0.23 % ±0.57% ±0.76% ±1.00%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='lib' n=100 0.05 % ±1.43% ±1.92% ±2.52%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='test/parallel' n=1 ** 2.56 % ±1.63% ±2.18% ±2.88%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='test/parallel' n=10 *** 4.31 % ±1.28% ±1.71% ±2.23%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='test/parallel' n=100 * 1.00 % ±0.76% ±1.02% ±1.34%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='benchmark' n=1 2.96 % ±4.42% ±5.95% ±7.89%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='benchmark' n=10 0.75 % ±1.01% ±1.35% ±1.78%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='benchmark' n=100 0.04 % ±0.75% ±1.00% ±1.32%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='lib' n=1 1.02 % ±1.31% ±1.74% ±2.27%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='lib' n=10 -0.29 % ±1.13% ±1.52% ±1.99%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='lib' n=100 0.26 % ±0.70% ±0.93% ±1.21%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='test/parallel' n=1 *** 2.64 % ±1.26% ±1.69% ±2.23%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='test/parallel' n=10 *** 5.59 % ±1.19% ±1.59% ±2.10%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='test/parallel' n=100 ** 1.59 % ±1.12% ±1.49% ±1.96%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='benchmark' n=1 1.62 % ±2.70% ±3.61% ±4.72%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='benchmark' n=10 -0.09 % ±1.25% ±1.67% ±2.20%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='benchmark' n=100 0.03 % ±0.60% ±0.80% ±1.04%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='lib' n=1 0.87 % ±1.34% ±1.78% ±2.32%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='lib' n=10 0.81 % ±1.74% ±2.32% ±3.03%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='lib' n=100 0.71 % ±1.92% ±2.56% ±3.35%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='test/parallel' n=1 -0.14 % ±0.75% ±1.01% ±1.33%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='test/parallel' n=10 0.16 % ±0.61% ±0.81% ±1.06%
fs/bench-glob.js recursive='true' mode='sync' pattern='*.js' dir='test/parallel' n=100 0.09 % ±1.20% ±1.60% ±2.08%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='benchmark' n=1 0.35 % ±1.17% ±1.57% ±2.06%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='benchmark' n=10 0.77 % ±0.79% ±1.05% ±1.38%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='benchmark' n=100 1.09 % ±1.19% ±1.60% ±2.12%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='lib' n=1 0.57 % ±1.04% ±1.38% ±1.80%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='lib' n=10 -0.57 % ±2.63% ±3.52% ±4.65%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='lib' n=100 2.33 % ±3.84% ±5.17% ±6.86%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='test/parallel' n=1 -0.26 % ±0.60% ±0.80% ±1.04%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='test/parallel' n=10 0.28 % ±0.66% ±0.88% ±1.15%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/*' dir='test/parallel' n=100 0.05 % ±0.93% ±1.24% ±1.61%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='benchmark' n=1 * 1.61 % ±1.44% ±1.94% ±2.55%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='benchmark' n=10 0.77 % ±3.60% ±4.82% ±6.33%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='benchmark' n=100 1.17 % ±1.52% ±2.05% ±2.71%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='lib' n=1 0.88 % ±0.93% ±1.23% ±1.60%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='lib' n=10 -0.21 % ±0.92% ±1.22% ±1.59%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='lib' n=100 0.86 % ±1.28% ±1.72% ±2.27%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='test/parallel' n=1 0.28 % ±0.50% ±0.66% ±0.87%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='test/parallel' n=10 1.55 % ±2.77% ±3.71% ±4.90%
fs/bench-glob.js recursive='true' mode='sync' pattern='**/**.js' dir='test/parallel' n=100 * 0.86 % ±0.86% ±1.15% ±1.52%
Be aware that when doing many comparisons the risk of a false-positive result increases.
In this case, there are 108 comparisons, you can thus expect the following amount of false-positive results:
5.40 false positives, when considering a 5% risk acceptance (*, **, ***),
1.08 false positives, when considering a 1% risk acceptance (**, ***),
0.11 false positives, when considering a 0.1% risk acceptance (***)
Performance is significantly improved when asynchronous compared to synchronous.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev
Please confirm when you have time. Many Thanks!
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think that benchmark is accurate. See my review comments.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev
Many Thanks, we were able to improve the benchmark code.
Here are the results of the improved benchmark!
The results of recursive=true useAsync=true
are:
confidence improvement accuracy (*) (**) (***)
fs/bench-glob.js recursive='true' useAsync='true' pattern='*.js' dir='benchmark' n=1000 1.03 % ±2.95% ±3.95% ±5.21%
fs/bench-glob.js recursive='true' useAsync='true' pattern='*.js' dir='lib' n=1000 1.14 % ±1.67% ±2.23% ±2.90%
fs/bench-glob.js recursive='true' useAsync='true' pattern='*.js' dir='test/parallel' n=1000 * 1.53 % ±1.32% ±1.75% ±2.28%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/*' dir='benchmark' n=1000 0.70 % ±1.27% ±1.70% ±2.23%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/*' dir='lib' n=1000 0.97 % ±2.06% ±2.75% ±3.58%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/*' dir='test/parallel' n=1000 0.73 % ±1.55% ±2.07% ±2.69%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/**.js' dir='benchmark' n=1000 -0.88 % ±1.31% ±1.74% ±2.27%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/**.js' dir='lib' n=1000 1.23 % ±2.51% ±3.35% ±4.36%
fs/bench-glob.js recursive='true' useAsync='true' pattern='**/**.js' dir='test/parallel' n=1000 0.66 % ±1.42% ±1.89% ±2.47%
Be aware that when doing many comparisons the risk of a false-positive result increases.
In this case, there are 9 comparisons, you can thus expect the following amount of false-positive results:
0.45 false positives, when considering a 5% risk acceptance (*, **, ***),
0.09 false positives, when considering a 1% risk acceptance (**, ***),
0.01 false positives, when considering a 0.1% risk acceptance (***)
The results of recursive=true useAsync=false
are:
confidence improvement accuracy (*) (**) (***)
fs/bench-glob.js recursive='true' useAsync='false' pattern='*.js' dir='benchmark' n=1000 0.33 % ±2.25% ±3.00% ±3.92%
fs/bench-glob.js recursive='true' useAsync='false' pattern='*.js' dir='lib' n=1000 2.97 % ±4.11% ±5.52% ±7.27%
fs/bench-glob.js recursive='true' useAsync='false' pattern='*.js' dir='test/parallel' n=1000 1.08 % ±1.16% ±1.54% ±2.01%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/*' dir='benchmark' n=1000 0.59 % ±1.82% ±2.43% ±3.16%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/*' dir='lib' n=1000 0.40 % ±1.85% ±2.47% ±3.22%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/*' dir='test/parallel' n=1000 1.25 % ±1.52% ±2.02% ±2.63%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/**.js' dir='benchmark' n=1000 0.33 % ±1.56% ±2.08% ±2.71%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/**.js' dir='lib' n=1000 0.42 % ±1.68% ±2.24% ±2.91%
fs/bench-glob.js recursive='true' useAsync='false' pattern='**/**.js' dir='test/parallel' n=1000 1.03 % ±1.15% ±1.53% ±1.99%
Be aware that when doing many comparisons the risk of a false-positive result increases.
In this case, there are 9 comparisons, you can thus expect the following amount of false-positive results:
0.45 false positives, when considering a 5% risk acceptance (*, **, ***),
0.09 false positives, when considering a 1% risk acceptance (**, ***),
0.01 false positives, when considering a 0.1% risk acceptance (***)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Nice! Looks like a bit of improvement
@nodejs/fs @nodejs/benchmarking |
dir: ['lib', 'test/parallel', 'benchmark'], | ||
pattern: ['**/*', '*.js', '**/**.js'], | ||
mode: ['async', 'sync'], | ||
recursive: ['true', 'false'], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
recursive: ['true', 'false'], | |
recursive: [true, false], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I didn't seem to use boolean in benchmark (error on CI), so I fixed that in this commit!
Many Thanks!
benchmark/fs/bench-glob.js
Outdated
const benchmarkDirectory = path.resolve(__dirname, '..', '..'); | ||
|
||
const configs = { | ||
n: [1, 10, 100], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
IMO n
should be a single (higher) value. Something like 3e4
(but that might be too high)?
100 is way to low.
benchmark/fs/bench-glob.js
Outdated
|
||
bench.start(); | ||
|
||
let counter = 0; |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
let counter = 0; |
No need to count the output
benchmark/fs/bench-glob.js
Outdated
} | ||
} | ||
|
||
bench.end(counter); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
bench.end(counter); | |
bench.end(config.n); |
for await (const entry of new Glob(pattern, options).glob()) { | ||
ArrayPrototypePush(res, entry); | ||
} | ||
const res = await ArrayFromAsync(new Glob(pattern, options).glob()); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
I don't think that benchmark is accurate. See my review comments.
const fullPath = path.resolve(benchmarkDirectory, config.dir); | ||
const { pattern, recursive, useAsync } = config; | ||
|
||
bench.start(); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
bench.start(); | |
let noDead; | |
bench.start(); |
benchmark/fs/bench-glob.js
Outdated
|
||
for (let i = 0; i < config.n; i++) { | ||
if (useAsync) { | ||
await new Promise((resolve, reject) => { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
await new Promise((resolve, reject) => { | |
noDead = await new Promise((resolve, reject) => { |
benchmark/fs/bench-glob.js
Outdated
}); | ||
}); | ||
} else { | ||
fs.globSync(pattern, { cwd: fullPath, recursive }); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
fs.globSync(pattern, { cwd: fullPath, recursive }); | |
noDead = fs.globSync(pattern, { cwd: fullPath, recursive }); |
} | ||
} | ||
|
||
bench.end(config.n); |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
bench.end(config.n); | |
bench.end(config.n); | |
assert.ok(noDead); |
@RedYetiDev |
modify config property because boolean cannot be used
ed1d2a8
to
c976446
Compare
bench.start(); | ||
|
||
for (let i = 0; i < config.n; i++) { | ||
if (mode === 'async') { |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Does this statement affect benchmarking speeds?
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev
Thank you for the review!
I think wrapping a Promise with each iteration affects benchmarking, So I fixed it.
After benchmarking again, there seems to be a decrease in performance..🥲
So, I'm trying to close the current PR and upload a PR that only adds benchmarking code. Is that okay?
mode: async, recursive: true
confidence improvement accuracy (*) (**) (***)
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='benchmark' n=1000 -0.55 % ±1.61% ±2.14% ±2.79%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='lib' n=1000 -0.16 % ±1.88% ±2.51% ±3.27%
fs/bench-glob.js recursive='true' mode='async' pattern='*.js' dir='test/parallel' n=1000 1.76 % ±3.88% ±5.19% ±6.83%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='benchmark' n=1000 ** -3.39 % ±2.20% ±2.93% ±3.83%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='lib' n=1000 -0.69 % ±2.17% ±2.89% ±3.77%
fs/bench-glob.js recursive='true' mode='async' pattern='**/*' dir='test/parallel' n=1000 -0.20 % ±1.24% ±1.65% ±2.14%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='benchmark' n=1000 -2.56 % ±2.61% ±3.49% ±4.57%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='lib' n=1000 -0.49 % ±2.24% ±2.98% ±3.87%
fs/bench-glob.js recursive='true' mode='async' pattern='**/**.js' dir='test/parallel' n=1000 -1.24 % ±1.48% ±1.97% ±2.56%
Be aware that when doing many comparisons the risk of a false-positive result increases.
In this case, there are 9 comparisons, you can thus expect the following amount of false-positive results:
0.45 false positives, when considering a 5% risk acceptance (*, **, ***),
0.09 false positives, when considering a 1% risk acceptance (**, ***),
0.01 false positives, when considering a 0.1% risk acceptance (***)
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
@RedYetiDev, Sorry for the frequent comments.
I look forward to hearing from you.
Many thanks!
Benchmark CI: https://ci.nodejs.org/view/Node.js%20benchmark/job/benchmark-node-micro-benchmarks/1642/
|
benchmark/fs/bench-glob.js
Outdated
dir: ['lib', 'test/parallel', 'benchmark'], | ||
pattern: ['**/*', '*.js', '**/**.js'], |
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
nit: There seem to be too many cases. Suggest reducing a few options if possible. Please consider if it's possible to obtain sufficient results with fewer iterations and cases.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you so much for the review!
If it's the review you mentioned, can I ask if it would be a good idea to try reducing the number of dir
and pattern
types?
I thought about it this way:
dir: ['lib'],
pattern: ['**/*', '*.js', '**/**.js'],
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Yes, I think the configs would be enough.
There was a problem hiding this comment.
Choose a reason for hiding this comment
The reason will be displayed to describe this comment to others. Learn more.
Thank you! There were definitely a variety of cases..! I fixed from defe492!
Landed in 090add7 |
Refs: nodejs#51912 PR-URL: nodejs#54644 Reviewed-By: Daeyeon Jeong <daeyeon.dev@gmail.com>
The following applies to this PR:
for await...of
loop withArrayFromAsync()
.ArrayPrototypePush
.ArrayFromAsync
type toprimordials.d.ts
fileArray.fromAsync
was defined relatively recently, but I used it because I knew it was officially adopted in Node.js v22. (MDN Document)