stream: accept ArrayBuffer in CompressionStream and DecompressionStream

[suuuuuuminnnnnn]

Per the WHATWG Compression Streams spec, CompressionStream and
DecompressionStream must accept BufferSource
(ArrayBuffer | ArrayBufferView) as valid chunk types.

Currently, writing a plain ArrayBuffer to the writable side throws
ERR_INVALID_ARG_TYPE, because the adapter in
newWritableStreamFromStreamWritable passes chunks directly to
stream.write(), which only accepts ArrayBufferView types.

This change normalizes plain ArrayBuffer chunks to a Uint8Array view
before forwarding them to the underlying Node.js stream.

Changes included:

• normalize ArrayBuffer to Uint8Array in lib/internal/webstreams/adapters.js
• add a regression test for ArrayBuffer input
• unskip the relevant WPT compression BufferSource test (if passing)

Fixes: #43433

[nodejs-github-bot]

Review requested:

• @nodejs/web-standards

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.
✅ Project coverage is 89.76%. Comparing base (ae2ffce) to head (d791f39).
⚠️ Report is 105 commits behind head on main.

Additional details and impacted files

@@            Coverage Diff             @@
##             main   #61913      +/-   ##
==========================================
+ Coverage   89.75%   89.76%   +0.01%     
==========================================
  Files         674      674              
  Lines      204416   204893     +477     
  Branches    39285    39378      +93     
==========================================
+ Hits       183472   183930     +458     
- Misses      13227    13232       +5     
- Partials     7717     7731      +14     

Files with missing lines	Coverage Δ
lib/internal/webstreams/adapters.js	85.75% <100.00%> (+0.17%)	⬆️

... and 79 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.

[suuuuuuminnnnnn]

Thanks @MattiasBuelens and @Renegade334 — applied both suggestions.

• added the object mode guard before normalizing ArrayBuffer
• replaced the manual read loop with Array.fromAsync(...)
• replaced the manual piping/write loop with await cs.readable.pipeTo(ds.writable)

I kept the PR scoped to the ArrayBuffer handling fix + regression coverage.
Appreciate the review.

[github-actions]

Failed to start CI

   ⚠  No approving reviews found
   ✘  Refusing to run CI on potentially unsafe PR

https://github.com/nodejs/node/actions/runs/22255119973

[mcollina]

lgtm

[nodejs-github-bot]

CI: https://ci.nodejs.org/job/node-test-pull-request/71388/

[Renegade334]

I think this is pretty much good to go.

One request, though: although testing the compression/decompression stream behaviour is appropriate, the underlying change is to the adapter behaviour, which should also be tested.

Could you please add something like the following to test/parallel/test-whatwg-webstreams-adapters-to-writablestream.js:

{
  const duplex = new PassThrough();
  const writableStream = newWritableStreamFromStreamWritable(duplex);
  const ec = new TextEncoder();
  const arrayBuffer = ec.encode('hello').buffer;
  writableStream
    .getWriter()
    .write(arrayBuffer)
    .then(common.mustCall());

  duplex.on('data', common.mustCall((chunk) => {
    assert(chunk instanceof Buffer);
    assert(chunk.equals(Buffer.from('hello')));
  }));
}

{
  const duplex = new PassThrough({ objectMode: true });
  const writableStream = newWritableStreamFromStreamWritable(duplex);
  const ec = new TextEncoder();
  const arrayBuffer = ec.encode('hello').buffer;
  writableStream
    .getWriter()
    .write(arrayBuffer)
    .then(common.mustCall());

  duplex.on('data', common.mustCall((chunk) => {
    assert(chunk instanceof ArrayBuffer);
    assert.strictEqual(chunk, arrayBuffer);
  }));
}

to test the ArrayBuffer conversion behaviour and honouring objectMode.