Notion Integration

### Describe the feature you'd like to request

I’d like floccus to support direct syncing with Notion, so bookmarks can be pushed/pulled to Notion databases without using an intermediate service.
Current workflow: I sync bookmarks from floccus to Linkwarden, then pull them into Notion. This adds complexity, and an extra point of failure.
Use case: keep a central, searchable Notion database of bookmarks with tags, notes, and original metadata (title, URL, favicons, created/updated timestamps), maintained automatically from browser bookmarks via floccus.

### Describe the solution you'd like

Add a Notion connector in floccus that:

Authenticates with Notion via OAuth (or integration token) and lets users choose target Notion database or page.
Maps bookmark fields to Notion properties (title, URL, tags, notes, created/modified timestamps, favicon, folder path).
Supports two-way sync or at least one-way push from floccus → Notion, preserving edits (detect conflicts and offer last-write-wins or configurable conflict resolution).
Allows configurable mapping of tags/folders to Notion multi-select or relation properties.
Preserves unique IDs so updates and deletes in floccus propagate to Notion.
Provides incremental sync (not full reimport each time) and respects rate limits; logs sync activity and errors in the floccus UI*

### Describe alternatives you've considered

- Continue using Linkwarden as an intermediary (current approach).
- Export/import bookmarks manually (HTML/CSV) into Notion — manual and non-incremental.
- Use a dedicated bookmarking service that has native Notion support (but that would require migrating away from floccus).

The issue names no files, tests, or entry points. Start by locating floccus’s existing connector and synchronization code, then trace how authentication, bookmark fields, incremental updates, conflicts, and UI logging are handled. Done requires a scoped Notion integration with the requested mapping and sync behavior, plus tests for its supported flows.

Written by the indexing model from the issue text.