Description

Users find images to be engaging, so articles on Wikipedia increasingly feature multi-media content. Initial interest suggests that users may want to pull the attribution information for all of the files contained within an article, in addition to making one off requests for specific files.

Conditions of acceptance

• Create a new endpoint that pulls the attribution information for all images/multi-media files contained within an article.
  • Suggested endpoint: /pages/{title}/files/signals
  • Short description: Get the attribution information media files on a Wikipedia article
  • Long description: Returns the collection of attribution information for all images and other media files contained within the article. This endpoint only works for Wikipedia article content; other Wikimedia project types are not supported at this time.
• Returns a list of attribution objects for all of the media files contained within the page
  • Media file links should point to the canonical Commons version, instead of the Wikimedia file version. This will help with subsequent lookup.
• Supports the 'expand' query parameter options for additional signals.
• Consider throwing a 501 not-implemented response for non-Wikipedia projects.

Implementation details

Assume this feature will only work for Wikipedia projects, unless we can reasonably assume reusability across other project types.

This will likely need a light research spike and additional exploration.

OPEN QUESTIONS:

1. Should the links go to the commons file page, for consistency/ability to request in isolation? Probably?
2. Should we assume this only works for Wikipedia articles? I'm assuming yes, for the sake of scoping and reliability.
3. What concerns might we have about performance?
4. I assume that articles have a reasonable number of images. Are there outliers where we might want to consider pagination?