hublo: the world needed another site generator

Configuration

(source-dir "site/sources")
(output-dir "site/out")

(handle! "\\.org$"      :org)
(handle! "\\.html$"     :noop)
(handle! "\\.css$"      :noop)
(handle! "\\.mustache$" :mustache)

(page "/entries/.*"
      (slugify)
      (route "/entries/{{year}}/{{month}}/{{day}}/{{slug}}.html")
      (group-with :entries)
      (layout "blog")
      (layout "default"))

(page "/index.mustache"
      (route "/index.html")
      (title "home")
      (layout "default"))

(page "/index.xml.mustache"
      (title "spootnik")
      (metadata :description "my blog")
      (route "/index.xml"))

(page "/static/.*")

Running

hublo clean
hublo publish
hublo phases <phase-list>

Philosophy

Hublo provides a way to build a site by publishing pages in different phases. A phase is a barrier that must be reached by all pages before advancing to the next. Each hublo command consists of a list of phases to run through, some shortcuts are defined in hublo-cli.el.

For instance, the most common hublo command is publish and translates to the following phases: :bootstrap, :metadata, :route, :content, :augment, :publish, and :clean.

This allows some steps to gather information or metadata from other pages which provides an elegant way to build lists and incrementally produce content.

In the above example, blog entries are grouped together with the label :entries which is then accessible in templates. This allows the index page and an rss feed page to go through the list. Similar techniques can be used for publishing content dynamically.

The configuration is processed as elisp code and can thus accomodate for plenty of scenarios.

There is an example configuration at https://github.com/pyr/blog with the corresponding output at: http://spootnik.org.