sszuev / jena-site

Website sources for the Apache Jena website

Apache Jena website

This is the source code for the website of Apache Jena, hosted at:

https://jena.apache.org/

Repository structure

This repository uses 3 branches for building the published website.

• The main branch, which contains all the sources for the website.
• The asf-site branch, which contains the generated website being used for the actual website.
• The javadoc branch, which has the javadoc to be published as part of the main website under documentation/javadoc/

When contributing patches, please create pull requests for the main branch.

Additionally the repository also has a asf-staging branch that can be used to preview website changes prior to publishing them to the main website. Any branch other than main that has a Jenkinsfile present in it will automatically be staged at https://jena.staged.apache.org, the Javadoc from the javadoc branch is also automatically staged into this staging site.

Content Management System

The website uses Hugo as static website generator. See Hugo for more info and for details how to install Hugo.

Generate the website

To generate the static website, execute hugo to generate and serve the website on localhost:1313.

During development, it may be useful to run an incremental build. For this to work, execute hugo server -D to continuously generate and serve the website on localhost:1313.

Building and publishing the website

The ASF Jenkins Jena_Site job is used for generating the website and committing the generated site to the asf-site branch.

Separately, javadoc for a release is committed into the javadoc branch during the release process.

gitpubsub is used to publish the site, using the content from the asf-site and javadoc branches.

ASF Jenkins job

The Jenkinsfile was contributed in apache#17 (July 2020).

Steps to set up the Jenkins job:

• Create a new multibranch pipeline (e.g. 'Jena_Site').

• Branch source -> git

• Set the gitbox url -> https://gitbox.apache.org/repos/asf/jena-site.git and use the jenkins (pub key) credentials.

• In the 'Scan Multibranch Pipeline Triggers' check the 'Periodically if not otherwise run' checkbox and enter a sane value (e.g. 15 minutes). This is needed because webhooks are not delivered to ci-builds (yet - 2020-07-28).

• Save the job and click the 'Scan Multibranch Pipeline Now' button to trigger an initial scan. A first run may also happen as the SCM polls.

It is at this point that it gets the label git-websites so the first job may have run on the wrong node.