DITA-OT Day 2018
What's next for DITA-OT?
Robert Anderson, Roger Sheen, Jarno Elovirta, George Bina
Who are you calling “stale” ‽
The DITA-OT issue tracker contains a backlog of open issues, many of which have not been updated in
several years, and may no longer apply to recent versions. This session will discuss options for removing
outdated issues from the backlog to help focus resources on issues that are most important to users.
Robert Anderson, Roger Sheen, Jarno Elovirta
Twisted XSLT Tricks: Making Column Switching Work for FOP
Switching from single column to two-column layout is almost impossible using the PDF2 transform and FOP
because FOP enforces XSL-FO's rule that only direct children of fo:flow can change the column spanning.
Likewise, splitting page sequences to change from portrait to landscape is hard to do with the PDF2 transform
because it provides no easy way to change the page sequence within the context of a topic's body (e.g., to put
rotated tables on landscape pages). This talk presents a general XSLT technique for splitting a single tree into
multiple trees, enabling changing column spanning and splitting page sequences with a minimum of rework of
normal templates.
Eliot Kimber
DITA-OT and DocOps
Developing documentation concurrent with the software using DITA-OT in VSTS environment at
Shell.
Quick van Rijt
Validation meets publication - Apply your style guide rules during the publication
DITA-OT warns about technical issues in the input. Why not warn about style guide violations? This talk
shows how Schematron can be used to check topics and maps after the preprocessing phase and stop the build if
content does not pass the defined quality gates.
Stefan Eike
A fast path for building data assets from DITA
Structured content can save the day when product complexity gets in the way. Let's see how DITA content
fits into an external database using the org.dita.normalize transformation as base. DITA-OT can be packaged in
different ways to prepare dev/test environment setups for IT and developer contributors.
François Violette
Overview of dita-semia open-source plugins for DITA-OT
I have created some DITA-OT plugins that are available as open-source (github.com/dita-semia). And I'd
like to give an overview of them: 1) postprocessing: Zip the output or set the pdf filename dynamically
depending on some content. 2) topic-num: Add numbers to topics, figures and tables (especially for HTML output).
3) image-convert: Convert images for compatibility (e.g. SVG to PNG for HTML output). 4) pdf: Yet another pdf
layout working with FOP. 5) diff: Compare two versions of a document and highlight the differences.resolver:
Resolve custom attributes (usually set as defaults be the schema) to standard DITA content, e.g. by applying an
XSLT transformation to some content, add static text content or resolve identifiers to cross
references.
Patrik Stellmann
Unit Testing DITA-OT Plugin Extensions
I have created a Unit Testing Framework for DITA-OT Plugins
https://github.com/jason-fox/fox.jason.unit-test - This is a DITA-OT Plugin to test DITA-OT Plugins. The
complete functionality would include HTML and command line test results, ANT script profiling, XSL template code
coverage, e automated CI testing - integration with Travis, and how to write maintainable tests swiftly and
painlessly.
Jason Fox
Step by step implementation of a DITA Open Toolkit plugin
There are lots of DITA-OT extension points, maybe list some of the most common ones, then based on 3-4
use cases implement small plugins which use those extension points. For example: a plugin which changes the
default label value for "Note"a plugin which adds an XHTML customization matching a DITA element with a certain
@status attribute value and producing a certain HTML spana plugin which customizes the PDF processing in some
waya plugin which adds a specialization DTD with an XML catalog
Radu Coravu
Meta DITA samples: testing around the edge cases
Over the past few years I've developed a number of test cases that are exhaustive but on very specific
features. The sample files are not well suited for automated testing, but can often be used to manually test
custom plugins. In this lightning talk, I'll describe the samples that exist today, and also solicit feedback on
other similar test sets that might prove useful.
Robert Anderson
Plug-in installation made easier
Description and a demo of the new DITA-OT plug-in repository.
Jarno Elovirta
Installing DITA-OT on macOS via Homebrew
A new alternative installation method can now be used to install DITA-OT on macOS via the platform's
most popular open-source package manager. This talk explains the benefits of this approach, highlights key
differences to the default installation via .zip archive and demonstrates the new installation
process.
Roger Sheen
Comparison between the DocBook and DITA publishing systems
Let's compare side by side look at how the Docbook and DITA publishing systems work. This could give us
some ideas about future directions for DITA publishing. The presentation could be structured like this: Main
target audiences for Docbook and DITA vocabularies. Maybe initially discuss a little the Docbook and DITA
vocabularies, the ways in which Docbook handled new versions, how DITA tried to have all versions backward
compatible, the choice not to have a namespace for DITA elements. Show how Docbook solved the problem with
supporting both Docbook 4 and Docbook 5 standards. The Docbook @version attribute. 1) Pure XSLT-based processing
versus the melange of ANT + Java + XSLT which is the DITA Open Toolkit. How parameters can be discovered and
set. Also discuss how many parameters are available by default to be set for Docbook. Maybe ask people if they
want more params available by default for DITA processing. Show some examples with how a simple XSLT
customization for the HTML output can be accomplished in both systems. Also discuss disadvantages of pure XSLT
based publishing approaches, impossibility of copying resources (images for example) to an output folder for
example. 2) Adding new elements/attributes to Docbook and DITA. If you change Docbook in any way, it can no
longer be named Docbook 3) User's guides. A very decent user's manual structured a bit like a tutorial binding
together the Docbook specs, extending the vocabulary + publishing part:. The DITA-OT documentation only contains
documentation about the publishing part. 4) Profiling (filtering content). For example Docbook allows you to
filter content based on any attribute (even on xml:lang). 5) Discuss new ideas to be implemented in DITA, after
they were originally used in the DocBook publishing system:a) Index pages creating when using Apache FOP b) The
Docbook WebHelp project (Javascript-based search). c) Syntax highlight in code blocks d) Olinks (linking between
publications) e) You can add xlink:href on any Docbook element converting it to a link. f) Custom Processing
instructions which can change the publishing.
Radu Coravu
DITA, Gradle and Git: a small team approach to publishing
At L-Acoustics we adopted an open source toolchain to automate our publishing process. With the CI
capabilities of GitLab and the dita-ot-gradle and saxon-gradle plugins developed by Eero Helenius, we have
created a comprehensive publishing solution suited to the needs of our small teams. One important feature of our
system is the handling of DITA-OT plugin dependencies. This architecture has facilitated the use of DITA by our
SMEs and marketing writers since from their point of view they just have to produce content and the output
"magically" appears.
Lionel Moizeau
DITA-OT Website Survey Results
This session presents the results of a survey of www.dita-ot.org website users conducted in summer
2018. Topic areas included user happiness with the current website, how often and how the site is used, in what
formats users desire the DITA-OT documentation, etc.
Lief Erickson
Highlights in recent DITA-OT versions
The core developers present an overview of the changes in DITA Open Toolkit within the past year and
highlight new features that may be of particular interest to users.
Robert Anderson/Jarno Elovirta