Honor forwarded headers when creating management URLs

[andystaples]

🐛 Describe the bug

A clear and concise description of what the bug is.

DurableOrchestrationClient.create_check_status_response() and get_client_response_links() do not consider Forwarded, X-Forwarded-Host, or X-Forwarded-Proto when constructing management URLs.

The SDK copies the extension-provided management URL templates, then unconditionally replaces their origin with the scheme and authority from request.url. Behind Azure Front Door, Application Gateway, or another reverse proxy, request.url can contain the internal origin. The resulting Location, statusQueryGetUri, and other management URLs can therefore expose an internal hostname or use HTTP even when forwarded headers describe the public HTTPS origin.

The Durable Functions extension added opt-in forwarded-header support in Azure/azure-functions-durable-extension#3083. Out-of-process binding templates are generated without an HTTP request, however, and the Python SDK subsequently replaces their origin from request.url, so that extension behavior does not cover this Python path.

This is separate from #355: current Azure Linux managed hosting correctly returns HTTPS management URLs for ordinary HTTPS-only Function Apps.

🤔 Expected behavior

What should have happened?

When trusted forwarded-header handling is enabled, Python-generated management URLs should use the client-facing scheme and host from Forwarded or X-Forwarded-Proto/X-Forwarded-Host, consistent with the Durable Functions extension behavior. Without opt-in, existing request.url behavior should remain unchanged.

☕ Steps to reproduce

What Durable Functions patterns are you using, if any?
Any minimal reproducer we can use?
Are you running this locally or on Azure?

1. Run a Python Durable Functions HTTP starter behind a reverse proxy.
2. Have the proxy forward a public HTTPS origin using X-Forwarded-Proto: https and X-Forwarded-Host: public.example.com, while the worker receives an internal request.url such as http://internal.example/api/start.
3. Return client.create_check_status_response(req, instance_id).
4. Observe that the management URL origins use http://internal.example rather than https://public.example.com.

A source-level reproduction is also sufficient: initialize the client with an HTTPS management URL template, call get_client_response_links() with an HTTP request.url, and observe that _replace_url_origin() downgrades the template to HTTP.

Relevant implementation: azure/durable_functions/models/DurableOrchestrationClient.py, in get_client_response_links() and _replace_url_origin().

⚡If deployed to Azure

We have access to a lot of telemetry that can help with investigations. Please provide as much of the following information as you can to help us investigate!

• Timeframe issue observed: N/A; identified through source analysis of reverse-proxy behavior
• Function App name: N/A
• Function name(s): N/A
• Azure region: N/A
• Orchestration instance ID(s): N/A
• Azure storage account name: N/A

If you don't want to share your Function App or storage account name GitHub, please at least share the orchestration instance ID. Otherwise it's extremely difficult to look up information.