feat: implement ipv6 addr flags

[dignifiedquire]

This implements detection of IPv6 flags and adding them to the interface.

I tried to mirror the existing code on the OS specific queries, happy to adjust as you would like.

I have tested this on linux, macos and windows 11 manually, and confirmed it working using the list_interfaces example.

I need this to detect deprecated addresses in iroh, n0-computer/iroh#4071 for better information about addresses.

[chatgpt-codex-connector]

💡 Codex Review

Here are some automated review suggestions for this pull request.

Reviewed commit: 3f13d71460

ℹ️ About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you

• Open a pull request for review
• Mark a draft as ready
• Comment "@codex review".

If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".

[chatgpt-codex-connector]

Align IPv6 flags with deduplicated address entries

This loop always appends to ipv6_scope_ids and ipv6_addr_flags even when push_ipv6 rejects a duplicate (addr, prefix) pair. In the duplicate case, iface.ipv6 is not extended but ipv6_addr_flags is, so indexing flags by IPv6 position can shift and report incorrect flags for later addresses. That violates the documented 1:1 alignment between Interface::ipv6 and Interface::ipv6_addr_flags (the Android path mirrors this pattern as well).

Useful? React with 👍 / 👎.

[shellrow]

Thanks! Looks great overall.

Before merging: push_ipv6 can skip entries (e.g. duplicates or invalid ones), but ipv6_scope_ids and ipv6_addr_flags are still always appended. This can break the intended 1:1 alignment with Interface::ipv6.

Could you update this so scope IDs and flags are only pushed when the IPv6 entry is actually inserted?
It looks like the Android path may need the same adjustment as well.

[dignifiedquire]

Could you update this so scope IDs and flags are only pushed when the IPv6 entry is actually inserted?
It looks like the Android path may need the same adjustment as well.

yes, I missed that, should be fixed now