OASIS Darwin Information Typing Architecture (DITA) TC

 View Only
  • 1.  RNG-to-DTD and XSD Status

    Posted 02-16-2014 18:38
    I have implemented the initial DTD and XSD generation process such that the resulting DTDs and XSDs for basetopic and basemap correctly validate my test documents. I am now setting up a continuous integration server to test the transforms. I have a working Jenkins build that does just the generation and I'm working on setting up validation of the output. I'm using CloudBees as the Jenkins server (cloudbees.com). CloudBees offers free Jenkins servers for non-commercial use. I'm currently doing it under my drmacro CloudBees user but we could set up a TC-specific CloudBees account if necessary. I can make my Jenkins build available to other CloudBees users, so if you create a CloudBees account I can add you to the build if you want to play along or at least be able to observe. I'm out of time to work on this today (Sunday) but I'll return to it on Monday. All my materials are committed to the OASIS SVN repository. The continuous integration process can be configured in many ways, but the way I'm initially setting it up is such that any commits to the doctypes/ subtree of the SVN repository will trigger the test process. All the results of the test are then captured as archived artifacts for inspection. The test process generates the DTDs and XSDs (URN- and URL-based versions) and then (once I get it set up) validates all the test documents that we're setting up (my thanks to Scott Hudson for putting together the initial set of test documents). I also plan to set up some log analysis so look for things like "unhandled element" messages from the transforms, but the main tests will simply be success/fail checks for validation (or non-validation as the case may be). I can certainly use help setting up test documents and additional test result analysis (log analyzers, etc.). I also need to create a complete Schematron for the RNGs themselves--George Bina created one for his initial RNG effort but that is now out of date with the final RNG organization and details. My Schematron skills are weak so if anyone has Schematron fu they'd like to contribute I would be most appreciative. Cheers, Eliot ————— Eliot Kimber, Owner Contrext, LLC http://contrext.com