As MapStruct itself, this web site is open-source, too. You can find it’s source code here.
The website is built using the Hugo static site generator. So it’s recommended to spend a few minutes reading the Getting Started section of the Hugo docs if you are not familiar with it yet.
The site is deployed to GitHub Pages (as a “project page”, using the “gh-pages” branch).
Install Hugo by following the instructions given on its homepage. E.g. using brew if you are on OS X:
brew update && brew install hugo
Make sure that you are using Git 2.5 or later due the worktree feature used for publication (see below). Check out the web site’s source code and assign the name “upstream” to the remote repository by running:
git clone git@github.com:mapstruct/mapstruct.org.git git remote rename origin upstream
Hugo comes with a built-in server for previewing the web site locally. Start it by running:
hugo server --buildDrafts
The website can then be accessed at http://localhost:1313/. The server injects some magic JavaScript which enables auto-reload of rendered pages in the browser as the source files are edited.
Web site contents are written using Markdown, source files are located under content. You can mark a new page, e.g. a news post, as a draft which will cause it to be ignored when generating the production website:
--- title: "My blog post" draft: true --- Lorem ipsum dolor sit amet, consectetur adipiscing elit...
For news posts a short summary is shown on the home page and the news landing page. Add the “more” comment to your posts after the part of the post that should be shown in the summary, e.g. an introductory first paragraph:
--- title: "My blog post" --- Two or three sentences as a teaser. And then the remainder of the post.
The team page is a data-driven page. If you want to be listed on the page, add a new file to data/team following the structure of the existing files in that directory.
The website is automatically published for every commit to master via this GitHub Workflow.
The workflow generates the HTML with the hugo command in the public directory.
The contents of that directory are then pushed to the “gh-pages” branch of the website repository.
To make publication as smooth as possible, the script scripts/generate-site.sh is taking care of the regeneration. It calls hugo and afterwards removes some generated files which are not needed.
A typical editing process will look like this (assuming you start on the master directory):
# update local branches git fetch upstream git merge upstream/master # do your changes... # commit git commit -a -m "Some website change" # deploy git push upstream master