Allow deleting moved messages

[00-kat]

Closes half of section three of #155, and lays the groundwork for the second half.

It's not perfect, but I doubt the few edge cases would ever occur.

[Copilot]

Pull Request Overview

This PR adds support for deleting moved messages while refining the moved message processing workflow. Key changes include updates to message data handling (switching to a MessageData.scrape method and using files instead of attachments), the addition of test cases for snowflake extraction and moved message author detection, and the introduction of a context menu command for deleting moved messages.

Reviewed Changes

Copilot reviewed 8 out of 8 changed files in this pull request and generated 1 comment.

Show a summary per file

File	Description
tests/test_message_moving.py	Adds comprehensive tests for snowflake extraction and moved message author logic.
app/utils/webhooks.py	Updates moved message formatting, replaces attachment handling with files, and renames a parameter.
app/utils/message_data.py	Switches to a MessageData subclass with asynchronous file scraping.
app/utils/init.py	Updates imports to reflect changes in message data handling.
app/core.py	Minor comment update for consistency.
app/components/zig_codeblocks.py	Adjusts call to move_message_via_webhook with a flag to disable move markers.
app/components/move_message.py	Introduces a context menu command to delete moved messages and updates permissions handling.
app/components/github_integration/mentions/resolution.py	Updates comment capitalization for consistency.

[00-kat]

I'll deal with the CI failure tomorrow (I didn't catch it before because Ruff 0.9.10, which is what I use, handles it correctly, while Ruff 0.9.6, which is what the CI uses as it is what is in the lockfile, doesn't).

I would like a suggestion on what to do:

• Bump Ruff.
• Add a noqa.

[trag1c]

Let's bump ruff to v0.11.78 :)

[00-kat]

Note: there is currently a known issue, where moved moved messages cannot be deleted, but I don't want to fix it now since it's an exceptionally rare case and it requires parsing the subtext, which I started in another commit and am not in the mood to do the commit surgery to reorder into this one.