Web interface / browser plugin for interactive scraper creation

[blahah]

Take ideas from

• PeerLibrary

cc @mitar

[mitar]

So the idea is to use a similar process that we use for annotation to let user highlight parts of a page and then we can store that as an open annotation standard targets. But instead of attaching annotations, we would use it to extract data.

This could be then integrated with nice user interface, maybe reusing parts of this feedback tool, or simply Annotator.

[mitar]

So this is related also to the question how to define scrappers. I would advise using Xpath as only one of available options. I think storing also other information similar to open annotation standard would be helpful:

• offsets in the page
• prefix/suffix + regex to match the content (instead of direct quote as used in annotations)
• xpath
• DOM path (my addition)