test: cover leading BOM stripping in json and url-search-params bodies

[dinwwwh]

Adds tests that a leading UTF-8 BOM (U+FEFF) is stripped before json and url-search-params bodies are parsed, in the aws-lambda, node and fetch adapters. All three already strip it, so this adds no source changes; the tests keep a future decoding change from silently letting a BOM through, which would make JSON.parse throw.

Coverage

• aws-lambda: plain JSON, base64-encoded JSON, and URL-encoded forms
• node and fetch: JSON and URL-encoded forms
• peer is not covered: its bodies arrive already parsed or as string fields, so it has no bytes to decode

Testing

• The new tests pass on current code, and fail when each adapter's decoder is switched to keep the BOM (ignoreBOM: true)

Note for reviewers

The commit history includes an earlier aws-lambda body decoding change that was reverted before opening this PR; the net diff is test files only.

[pkg-pr-new]

• @standard-server/hono-node-playground
• @standard-server/node-http-playground

@standard-server/aws-lambda

npm i https://pkg.pr.new/@standard-server/aws-lambda@135

@standard-server/core

npm i https://pkg.pr.new/@standard-server/core@135

@standard-server/fastify

npm i https://pkg.pr.new/@standard-server/fastify@135

@standard-server/fetch

npm i https://pkg.pr.new/@standard-server/fetch@135

@standard-server/node

npm i https://pkg.pr.new/@standard-server/node@135

@standard-server/peer

npm i https://pkg.pr.new/@standard-server/peer@135

@standard-server/shared

npm i https://pkg.pr.new/@standard-server/shared@135

commit: 5ff8381

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[codspeed]

Merging this PR will not alter performance

✅ 26 untouched benchmarks
⏩ 108 skipped benchmarks¹

---

_{Comparing claude/dazzling-hopper-c5a634 (5ff8381) with main (39c1ce1)}

Footnotes

1. 108 benchmarks were skipped, so the baseline results were used instead. If they were deleted from the codebase, click here and archive them to remove them from the performance reports. ↩

[pullfrog]

✅ No new issues found.

Reviewed changes

Test-only PR adding coverage that a leading UTF-8 BOM (U+FEFF) is stripped before json and url-search-params bodies are parsed; no source changes ship.

• aws-lambda — BOM cases for plain JSON, base64-encoded JSON, and URL-encoded forms, all hinted via Content-Type.
• fetch — BOM cases for json and url-search-params, hinted via the standard-server header.
• node — BOM cases for json and url-search-params driven through a real supertest round trip.

I confirmed the tests are discriminating rather than tautological: all 104 tests in the three files pass as written, and patching each adapter's decoder to { ignoreBOM: true } makes every new BOM case fail — including url-search-params, since new URLSearchParams('\uFEFFfoo=bar') keeps the BOM in the key (\uFEFFfoo) rather than discarding it. The implementations lean on platform defaults (new TextDecoder() with ignoreBOM false in aws-lambda/node, Request.text() in fetch), which is exactly what these assertions pin. The peer-exclusion rationale in the description also checks out: packages/peer/src/body.ts hands message.json.body (an already-decoded JS string) straight to new URLSearchParams, so it has no decoding step to guard.

  ^{｜ View workflow run ｜ Using DeepSeek Flash (default — pick a model for stronger reviews) ｜ 𝕏}