fs: port SonicBoom module to fs module as FastUtf8Stream

[jasnell]

As a first step to porting portions of the pino structured logger into the runtime, this commit ports the SonicBoom module to the fs module as FastUtf8Stream. Sonicboom is a dependency of pino.

This is a faithful port of the SonicBoom module with some modern updates, such as converting to a Class and using Symbol.dispose. The bulk of the implementation is unchanged from the original.

Refs: #49296 (comment)

/cc @mcollina @ronag @kibertoad @jsumners @mmarchini @feugy

[nodejs-github-bot]

Review requested:

• @nodejs/security-wg
• @nodejs/streams

[codecov]

Codecov Report

❌ Patch coverage is 74.48870% with 237 lines in your changes missing coverage. Please review.
✅ Project coverage is 89.96%. Comparing base (cd1a90d) to head (a5d35b0).
⚠️ Report is 209 commits behind head on main.

Files with missing lines	Patch %	Lines
lib/internal/streams/fast-utf8-stream.js	74.21%	235 Missing and 2 partials ⚠️

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #58897      +/-   ##
==========================================
- Coverage   90.06%   89.96%   -0.11%     
==========================================
  Files         648      649       +1     
  Lines      191078   192007     +929     
  Branches    37454    37631     +177     
==========================================
+ Hits       172101   172742     +641     
- Misses      11598    11885     +287     
- Partials     7379     7380       +1     

Files with missing lines	Coverage Δ
lib/fs.js	98.18% <100.00%> (+<0.01%)	⬆️
lib/internal/streams/fast-utf8-stream.js	74.21% <74.21%> (ø)

... and 30 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.

[mcollina]

huge +1. This would be beneficial for other loggers too.

[jasnell]

Tests updated. @mcollina ... So far I have not ported any of the tests that require proxyquire. I'm not sure what's the best way to port that behavior yet. I think, however, that I'd like to get this landed as experimental first and then incrementally improve the test coverage.

[mcollina]

Some of the worst bugs we fixed are being tests by the proxyquire-based tests (those are error conditions that are impossible to replicate without mocking). They are the battle scars, and they are important to keep around / being ported.

I'm ok for landing this as-is, but can you create a tracking issue for all the tests that you skipped so that they can be ported around?

[jasnell]

Absolutely. It's really all of the tests that use proxyquire. We can port those by monkey patching the fs APIs. Not as clean but doable. Will just take time

[mcollina]

lgtm

[jasnell]

@mertcanaltin ... looks like the tests here are hanging and failing consistently. Looks like switching away from using node:test and restructuring the tests helps clear it. Will need to work through refactoring a bit more.

[mertcanaltin]

@mertcanaltin ... looks like the tests here are hanging and failing consistently. Looks like switching away from using node:test and restructuring the tests helps clear it. Will need to work through refactoring a bit more.

I would love to help you with this, I'll try an arrangement

[aduh95]

Tests are failing on Windows with this change on v22.x-staging, we would need a manual backport

[joyeecheung]

parallel/test-fastutf8stream-flush-mocks and parallel/test-fastutf8stream-sync have recently failed 14 PRs out of the recent 100 CI runs: #59638