perf: use swc to transform TS in jest

[deodad]

Motivation

Closes #314. Speed up yarn test by ~2x.

Change Summary

Replace ts-jest with @swc/jest to transform our TS files in Jest.

Merge Checklist

• The title of this PR adheres to the conventional commits standard

[socket-security]

Socket Security Pull Request Report

👍 No new dependency issues detected in pull request

Pull request report summary

Issue	Status
Install scripts	✅ 0 issues
Native code	✅ 0 issues
Bin script confusion	✅ 0 issues
Bin script shell injection	✅ 0 issues
Unresolved require	✅ 0 issues
Invalid package.json	✅ 0 issues
HTTP dependency	✅ 0 issues
Git dependency	✅ 0 issues
Potential typo squat	✅ 0 issues
Known Malware	✅ 0 issues
Telemetry	✅ 0 issues
Protestware/Troll package	✅ 0 issues

Bot Commands

To ignore an alert, reply with a comment starting with @SocketSecurity ignore followed by a space separated list of package-name@version specifiers. e.g. @SocketSecurity ignore foo@1.0.0 bar@2.4.2

Ignoring: @swc/core@1.3.24

Powered by socket.dev

[codecov]

Codecov Report

Base: 80.92% // Head: 80.61% // Decreases project coverage by -0.30% ⚠️

Coverage data is based on head (2a7acce) compared to base (c7bc251).
Patch has no changes to coverable lines.

Additional details and impacted files

@@            Coverage Diff             @@
##             main     #330      +/-   ##
==========================================
- Coverage   80.92%   80.61%   -0.31%     
==========================================
  Files          65       65              
  Lines        5777     5691      -86     
  Branches     1014     1041      +27     
==========================================
- Hits         4675     4588      -87     
- Misses       1095     1096       +1     
  Partials        7        7              

Impacted Files	Coverage Δ
src/flatbuffers/generated/message_generated.ts	78.46% <0.00%> (-1.02%)	⬇️
src/flatbuffers/generated/gossip_generated.ts	71.42% <0.00%> (-0.71%)	⬇️
...buffers/generated/name_registry_event_generated.ts	57.52% <0.00%> (-0.68%)	⬇️
src/eth/ethEventsProvider.ts	61.74% <0.00%> (-0.51%)	⬇️
...atbuffers/generated/id_registry_event_generated.ts	82.31% <0.00%> (-0.32%)	⬇️
src/flatbuffers/generated/rpc_generated.ts	65.85% <0.00%> (-0.18%)	⬇️
src/utils/hubErrors.ts	90.90% <0.00%> (ø)
src/network/p2p/node.ts	78.99% <0.00%> (ø)
src/rpc/client/index.ts	92.59% <0.00%> (ø)
src/storage/engine/index.ts	83.62% <0.00%> (ø)
... and 11 more

Help us with your feedback. Take ten seconds to tell us how you rate us. Have a feature suggestion? Share it here.

☔ View full report at Codecov.
📢 Do you have feedback about the report comment? Let us know in this issue.

[deodad]

The security check that is failing is for the install script from @swc/core, here's what the script does:

It checks if corresponding optional dependencies for native binary is installed and can be loaded properly.
If it fails, it'll internally try to install @swc/wasm as fallback.

This alert can be ignored by replying with a comment starting @SocketSecurity ignore @swc/core@1.3.24

[varunsrin]

I ran some benchmarks locally, and the improvement seems closer to ~ 15%, not 100%. I still think its worth merging in, but I want to make sure this discrepancy isn't the result of some other config issue before we do that.

I used hyperfine to benchmark, and here's the output when running yarn test from main and clearing the cache each time

➜  hub git:(main) ✗ hyperfine 'rm -rf build && yarn test'
Benchmark 1: rm -rf build && yarn test
  Time (mean ± σ):     12.751 s ±  0.262 s    [User: 60.905 s, System: 3.453 s]
  Range (min … max):   12.155 s … 13.006 s    10 runs

Here's the output when doing it on your branch:

➜  hub git:(ad/swc-build) ✗ hyperfine 'rm -rf build && yarn test'
Benchmark 1: rm -rf build && yarn test
  Time (mean ± σ):     11.166 s ±  0.222 s    [User: 40.538 s, System: 2.851 s]
  Range (min … max):   11.004 s … 11.657 s    10 runs

[deodad]

Jest has its own cache for transformed modules separate from ./build, what do you get w hyperfine 'yarn test --no-cache'?

ts-jest:

Benchmark 1: yarn test --no-cache
  Time (mean ± σ):     51.285 s ±  0.889 s    [User: 282.376 s, System: 15.140 s]
  Range (min … max):   50.079 s … 52.389 s    5 runs

swc:

Benchmark 1: rm -rf build && yarn test --no-cache
  Time (mean ± σ):     21.337 s ±  0.384 s    [User: 82.215 s, System: 7.419 s]
  Range (min … max):   20.936 s … 21.719 s    5 runs

[varunsrin]

That comes out much better, shows closer to a 40% improvement on an M1. I think its worth merging

✨  Done in 19.70s.
  Time (mean ± σ):     18.909 s ±  0.646 s    [User: 109.488 s, System: 6.586 s]
  Range (min … max):   17.801 s … 19.797 s    10 runs

✨  Done in 12.46s.
  Time (mean ± σ):     12.653 s ±  0.185 s    [User: 41.741 s, System: 3.459 s]
  Range (min … max):   12.454 s … 13.080 s    10 runs

[varunsrin]

@SocketSecurity ignore @swc/core@1.3.24