Page MenuHomePhabricator

Add SPARQL as input for Massviews
Open, Needs TriagePublic

Description

First of all, link to SPARQL for those who doesn't know what it is.

So it would be nice to include it as input for Massviews.

Some notes/ideas:

  • if user doesn't include ?item variable, error should be showed
  • probably SPARQL queries will include ?item variable, but people will want to get pageview data for some specific Wikipedia, so there should be some internal converter from Wikidata item to Wikipedia article
  • I would suggest inputting SPARQL query itself, not link to it (as they are pretty long) and because you can't directly get results. You have to use another URL (example in Python)

Event Timeline

Ahh, just noticed T164113. Well, this or that task may be subtask of other one

@Edgars2007 For starters, I am myself new to SPARQL but I've been wanting to get into it, so this sounds like a good opportunity to get my feet wet :)

Regarding your third bullet point – ideally anything you do with Massviews can be permalinked to. Would the payload in this code be short enough to put in a URL alongside the current Massviews URL parameters? Maybe we could have it automatically create a Pagepile, then link to that? That's assuming the results are all for a single project. Or is there another service that saves the SPARQL query? I also worry about syntax highlighting – something you'd probably want if you are able to directly input the SPARQL code.

Sorry for late reply.

Some "documentation".

Sintax highlight. Hmm, that of course would be cool, but two tools, that currently offer some SPARQL support (for some needs) doesn't offer that, so... Those tools: Descriptioner, Petscan ("Other sources" tab).