What's actually getting passed to Drupal once the AWS workflow is complete?

[Ian Mulvany]

Q: What's actually getting passed to Drupal once the AWS workflow is complete? Is the XML transformed into a JSON representation before being piped to Drupal? Or rendered all the way down to HTML?

It is a subset of the article in JSON, with a pointer to the XML file stored on a CDN. The HTML rendering happens on the Drupal site. 

We are working on a full JSON representation of the article. 

[na...@notthisway.com]

The idea of a full JATS XML -> JATS JSON converter is pretty exciting to me. Would this likely be developed as a Python script executed by Simple Workflow? Is there any chance the converter could be written as a standalone component or class that could be easily deployed in a range of server environments?

I can see a JSON representation being a key step in getting more interesting, meaningful web-based reading experiences for journal articles (Lens being a great example of that).

[Ian Mulvany]

Hi Nate,

It's looking like the XML -> JSON converter is going to be developed as a stand alone python library that SWF can call, but that could also be used in other contexts. 

[Nate Wright]

That's great to hear!

--
You received this message because you are subscribed to a topic in the Google Groups "elife-continuum-list" group.
To unsubscribe from this topic, visit https://groups.google.com/d/topic/elife-continuum-list/awGqtQ86qqQ/unsubscribe.
To unsubscribe from this group and all its topics, send an email to elife-continuum-list+unsub...@googlegroups.com.
To post to this group, send email to elife-continuum-list@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/elife-continuum-list/bc7add0f-7fef-4c0e-a741-20404663148f%40googlegroups.com.
For more options, visit https://groups.google.com/d/optout.

[Luke Skibinski]

Hi Nate,

What code I do have regarding xml to json conversion is here: https://github.com/elifesciences/bot-lax-adaptor

Valid article json (as we have defined it[1]) requires more data than is being captured by the xml alone so it's embedded within a small application whose purpose is to receive article xml from a queue and output article json into Lax, where other applications are also feeding article data. From Lax we'll be able to output valid article json.

However if you don't care for what elife considers valid (and that's totally fine) you can use the adaptor as-is. Once this initial development churn has ended we'll look at separating it out into a standalone library. Right now it's more efficient to develop these parts as a single application.

Converting the article body is the next big (huge) task.

[1] https://github.com/elifesciences/api-raml/blob/master/dist/model/article-vor.v1.json

Best,

Luke