Support timeouts in `send_request`

[vdusek]

SendRequestFunction (src/crawlee/_types.py) takes only url, method, payload and headers, so a request handler cannot bound how long an extra HTTP call may take. Add a timeout option.

JS implementation: apify/crawlee#3457

✍️ Drafted by Claude Code

[Battleplus]

Hi, I'd like to work on this issue.

I plan to add the timeout option to the public SendRequestFunction contract, pass it through the crawler context helper to the existing HTTP client timeout support, and add focused regression tests. I will keep the change scoped to #2138 and account for the open HTTP-client cookie/connection-pool work when reviewing conflicts.

OpenAI Codex will assist with implementation, test design, and review. I will inspect, understand, and verify the changes before submitting the PR.