Recipe import cache

[MadeInPierre]

When importing a recipe, create a dedicated row with author_id = NULL, then duplicate the row again with the right author_id.

Also add a column with the app version, LLM/scraping strategies used, some stats, etc... so that in the future we can decide if and when to invalidate the cache (e.g. new app version which produces better results, hence previous cached recipes should be ignored/deleted).

Hence, when a user imports a recipe from a URL, we can first check if it has already been imported and skip to the duplication. This enables:

• Less waiting time
• Keeps full customization for users after the recipe has been imported