Hello. This is an extremely rough proof of concept for a museum object database scraping tool.

It is loosely based around the ideas we put together for hoard.it but has been extended so that the transform file model is more flexible and democratic. Some background can be found on this blog post.


Install the bookmarklet by clicking and dragging the link below onto your browser toolbar. Then when you're on a collections site, click the button.


What does the bookmarklet do..?


  1. Look for the URL in the cachedurls table of the database
  2. Look to see if there is a transform file associated with the URL root
  3. Process the page and either display structured data from the database or generated on the fly depending on the outcome from 1.
  4. Insert the URL into the cachedurls table

Note that the code doesn't cope at all well with errors - if you give it a URL root which it doesn't recognise, it'll fall over. This will be fixed later on...