Details
-
Sub-task
-
Status: Open
-
Major
-
Resolution: Unresolved
-
None
-
None
-
None
Description
These are the parts of the current website generation that I've identified that aren't directly handled by Antora (yet):
- Collection of example .adoc files. I've written a bash script to collect these, sort them by language, and copy them to an antora structure. This could be translated to javascript or java.
- Generation of aggregated javadoc. I think this can be done by maven?
- Incorporation of javadoc into generated site. My current plan is to write an Antora pipeline step that takes a javadoc jar and adds the content to the Antora playbook, and add a way to directly include this content without applying the Antora "skin". This sort of activity is frequently requested for Antora so this will be a useful prototype or example. I expect eventually Antora will support this without extension.
- Generation of some index pages based on attributes from the static pages. The data needed to do this will be available to the UI in the soon-to-be-released Antora 2.3. We'll want our custom UI bundle, so this code can be added there. (The needed data is being added exactly to support functionality such as this). The code will need to be in javascript, but it should be pretty simple.
- Publication of the generated site to git. At worst, the site can be generated into a locally checked out copy of the appropriate git repo and committed by hand. However, a git publisher has been planned for a long time: I believe this would at least add the generated site to a local clone of the repo (probably bare), possibly commit it, and possibly push it. See Write a git site publisher.