preprocess

#Overview

Since at least 2006, we have been performing XML markup on our journal articles. The workflow has gone through a number of iterations. Though we're going to be moving to a Markdown-based system for creating OJS galleys, we still use the Érudit Article 3.0 DTD (for detailed markup) and Schema (for light markup via Tournesol) heavily in our workflow. All of the scripts in the preprocess stack are used at varying stages of the XML markup process.

##buildImportFile

buildImportFile runs from the issue-level of a journal in our working directory. It scrapes the HTML, PDF, and XML directories within that folder to create an xml file in OJS's native.dtd. This XML file contains all the article and issue metadata necessary for batch upload via the command line. It will prompt users if HTML or PDF galleys are missing, as well as preserve older ojs_import files if run additional times.

##cleanUp

cleanUp makes XML markup more tolerable and reasonable for people who aren't robots. It automatically numbers things that require sequential IDs like paragraphs, footnotes, images, and the like. It also automatically creates xlink connections between image tags and the frontmatter. It should be run towards the end of a markup process as defined by the user guide for markup. Without it, we'd have lost many brave souls to the mundane.

##eruditify

eruditify is a script that does a bunch of initial replacement on an XML file before markup starts. Since the starting point for markup is actually usually HTML, this script crosswalks some basic HTML into the Erudit Article 3.0 DTD.

##transform

transform is the basic script we use to apply our XSLT to our XML. It creates HTML for every galley we have an XML file for.