fix(shared): stop Queue.pull treating undefined as the empty sentinel

[dinwwwh]

Queue.pull() shifted the buffer first and then tested the result against undefined, so a legitimately buffered undefined value was consumed and discarded while the pull parked despite later buffered items — the value was lost and pull ordering inverted. pull() now checks the buffer length before shifting, matching push(), which already hands undefined straight to a waiting resolver.

Fixes

• Queue<string | undefined> (and similar) no longer loses pushed undefined values; pulls return them in order instead of parking.
• Latent in-repo: all current consumers use non-optional element types, but Queue is publicly exported with unconstrained T, so external consumers were exposed.

Testing

• New regression test: push(undefined) then push('a') pulls back undefined then 'a' in order.
• All 9 queue tests pass.

[pkg-pr-new]

• @standardserver/hono-node-playground
• @standardserver/node-http-playground

@standardserver/aws-lambda

npm i https://pkg.pr.new/@standardserver/aws-lambda@50

@standardserver/bun

npm i https://pkg.pr.new/@standardserver/bun@50

@standardserver/core

npm i https://pkg.pr.new/@standardserver/core@50

@standardserver/deno

npm i https://pkg.pr.new/@standardserver/deno@50

@standardserver/fastify

npm i https://pkg.pr.new/@standardserver/fastify@50

@standardserver/fetch

npm i https://pkg.pr.new/@standardserver/fetch@50

@standardserver/node

npm i https://pkg.pr.new/@standardserver/node@50

@standardserver/peer

npm i https://pkg.pr.new/@standardserver/peer@50

@standardserver/shared

npm i https://pkg.pr.new/@standardserver/shared@50

commit: 266c7a9

[codecov]

Codecov Report

✅ All modified and coverable lines are covered by tests.

📢 Thoughts on this report? Let us know!

[codspeed-hq]

Merging this PR will not alter performance

✅ 26 untouched benchmarks
⏩ 108 skipped benchmarks¹

---

_{Comparing claude/queue-undefined-sentinel-shift-eda977 (266c7a9) with main (db2dd09)}

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. ↩