fix: merge partial route options with global options

[Vinayak1337]

Fixes #340

What changed

• Merge raw global plugin options with partial route-level compress and decompress options before processing them once.
• Preserve route option precedence without changing the existing option processors or public API.
• Add regression coverage for inherited handlers and settings, including the forceRequestEncoding scenario raised during review.

Why

The previous Object.assign({}, globalParams, processParams(routeOptions)) flow allowed defaults and undefined fields from the processed route object to overwrite configured global values. Processing one coherent { ...globalOptions, ...routeOptions } object applies the intended route > global > default precedence and follows the simpler approach demonstrated in #398.

Tests

• node --test test/routes-compress.test.js test/routes-decompress.test.js
• npm run lint
• npm test
• git diff --check
• npm run benchmark --if-present (executed; the harness rejected the measurements because the unchanged control varied by 12% on a busy machine)

No documentation or type changes are needed because this fixes existing option precedence without changing the public API.

Checklist

• run npm run test && npm run benchmark --if-present
• tests and/or benchmarks are included
• documentation is changed or added
• commit message and code follows the Developer Certificate of Origin and the Code of conduct

[ranjeetcao]

👋 I just closed #398 in favor of this one (it hadn't surfaced in my issue-thread search). One small suggestion if you take another pass: this regression also affects the compress-side onUnsupportedEncoding hook (not just the decompress hooks), and the current test coverage here doesn't exercise that path. A minimal test sketch:

describe('When routes `compress` options are partial — issue #340 :', async () => {
  test('it should fall back to global `onUnsupportedEncoding` when the route only sets `threshold`', async (t) => {
    // ... (paste the test from my branch)
  })
})

Full version on my closed PR if useful: https://github.com/fastify/fastify-compress/pull/398/files

[Vinayak1337]

Thanks for catching this, @ranjeetcao. That path was a real coverage gap.

I added a focused regression test for the route-level compress: { threshold: 1 } case to make sure it falls back to the global onUnsupportedEncoding hook.

I also re-ran:

• node --test test/routes-compress.test.js
• npm run lint
• npm test

Appreciate the pointer.

[mcollina]

lgtm

[Copilot]

Pull request overview

This PR fixes option resolution precedence so that partial route-level compress/decompress option objects inherit from processed global plugin options first, and only fall back to plugin defaults when neither route nor global provided a value (fixing #340).

Changes:

• Refactored option processing to support process*Params(routeOpts, globalParams) merging without re-expanding missing route values to plugin defaults.
• Added regression tests to ensure partial route overrides preserve global handlers/settings (onInvalidRequestPayload, removeContentLengthHeader: false).

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated no comments.

File	Description
index.js	Refactors compress/decompress parameter processing to properly merge route-level partial options with processed global options.
test/routes-decompress.test.js	Adds coverage ensuring a partial route decompress override preserves the global onInvalidRequestPayload handler.
test/routes-compress.test.js	Adds coverage ensuring a partial route compress override preserves global removeContentLengthHeader: false (and adds a related partial-options test).

---

💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.

[Copilot]

Pull request overview

Copilot reviewed 3 out of 3 changed files in this pull request and generated no new comments.

Suppressed comments (1)

index.js:262

• baseParams.encodings may already have been narrowed to the global forceRequestEncoding below. Consequently, with global { requestEncodings: ['gzip', 'deflate'], forceRequestEncoding: 'gzip' }, a route override { forceRequestEncoding: 'deflate' } inherits only ['gzip'] and rejects every request, although deflate is allowed by the global list. Preserve the unforced request-encoding list separately and narrow only the effective route list after resolving its force option; this case should also have regression coverage.

    encodings: encodingsOrBase(opts, 'requestEncodings', baseParams),
    forceEncoding: forceEncodingOrBase(opts, baseParams)
  }

  if (params.forceEncoding && params.encodings.includes(params.forceEncoding)) {
    params.encodings = [params.forceEncoding]

[Fdawgs]

Incredibly hard to review compared to #398. Needs simplifying.

[Vinayak1337]

Thanks @Fdawgs — I simplified the implementation to follow the approach in #398: raw global and route options are merged first, then passed through the existing processor once. The processor refactor and merge commit are gone, and the branch is rebased onto current main.

I also trimmed the tests to the minimum focused coverage: the issue #340 invalid-payload-handler case, compression-handler inheritance, removeContentLengthHeader: false, and the forceRequestEncoding / global requestEncodings scenario raised by Copilot.

Local validation:

• node --test test/routes-compress.test.js test/routes-decompress.test.js (22 passed)
• npm run lint
• npm test (238 unit tests plus tstyche passed)
• git diff --check

The fresh GitHub Actions run is currently marked action_required before any jobs were created, so it appears to need maintainer approval to run the fork workflow.

[Tony133]

@Fdawgs as soon as you have time, could you review the latest changes? If everything looks good to you, we'll merge it.

[Copilot]

🟡 Changes recommended

Explicitly undefined route fields still erase global values, and the request-encoding inheritance test does not exercise inheritance.

Once you've addressed the issues Copilot identified, you can request another Copilot review.

Review details

• Files reviewed: 3/3 changed files
• Comments generated: 3
• Review effort level: Balanced

[Vinayak1337]

Good catches, I missed the explicit undefined case, and the original requestEncodings test was not actually testing inheritance.

I have fixed the option merge, updated the regression tests, and rebased the branch onto the latest main.

I also checked the earlier review feedback again and added coverage for both allowed and rejected route-level force encoding cases.

Local tests, lint, TypeScript checks, coverage, and git diff --check are all passing.

[Tony133]

@Fdawgs do you agree with merging this PR? 😄

[Fdawgs]

@Fdawgs do you agree with merging this PR? 😄

I've lost the will with this PR and similar.
#340 seems to be an AI bot magnet and I don't really care to review them.

I'd much prefer to fix this myself?

[Tony133]

@Fdawgs do you agree with merging this PR? 😄

I've lost the will with this PR and similar. #340 seems to be an AI bot magnet and I don't really care to review them.

I'd much prefer to fix this myself?

If you'd rather fix it yourself, that's fine with me — otherwise I'd suggest we merge this one. I'll keep this PR open for now so the contributor's work isn't lost either way, and we can close it once your fix lands if you go that route. 😉