[Feature Request] Bookmark Assets API

[tnt1232007]

Scenario

Hi, I am thinking of using a LLM to summerize the articles that I saved in Linkding.

Currently I am trying to do it with n8n:

1. Fetch the bookmark article from Linkding
2. Run its content through a LLM to summerize
3. Save the response as an asset back to Linkding

But I find out there's no APIs available to read and upload assets afaik

Requests

• Modify the existing GET Bookmark APIs to expose the Asset IDs
• New set of Bookmark Assets APIs:
  • GET - Retrieve asset metadata by ID, as showed in /admin/bookmarks/bookmarkasset/<ASSET_ID>/change/
  • GET - Retrieve asset data by ID, either raw content or file download Extra if I can utilize the new reader-mode and optionally retrieve the content that has been processed
  • POST - Upload new asset to a bookmark
  • DELETE - Delete asset by ID, just for completeness

Please review and consider the implemention. Thanks for the amazing works, especially the recently added features, I love the snapshot and reader mode.

[sissbruecker]

Exposing a list of assets through the bookmark details, and then having an endpoint for downloading an asset by ID would probably be a good start. Contributing that shouldn't be too difficult if you're interested.

Regarding reader mode, converting the HTML into a readable article actually only happens in the browser when you open the reader view, so that content is not available on the server. Depending on which language you want to use for processing your bookmarks there is probably some similar library for extracting the readable content from an HTML document.