The Pipeline

  1. The Steps
  2. Pipeline Content
 

At the center of everything in Static Shock is a pipeline and a few data structures.

Broadly speaking, there are two primary paradigms for programming. One approach is object orientation, where long-lived objects attach behaviors to data, and the behaviors form the contract. The other approach is procedural, where the data structure forms the contract, and the code operates on the data.

Pipelines are fundamentally procedural. The data is the contract. All areas of Static Shock agree about the kind of data that's being manipulated. As the data moves through the pipeline, each step of the pipeline reads and writes the data. At the end, the data is written to files, which form a static site.

The core of Static Shock has two primary jobs:

  1. Define the data structures that all code reads and writes.
  2. Implement a pipeline and push the data through it.

All behavior outside of these two responsibilities are implemented by Static Shock plugins.

The Steps

The Static Shock pipeline runs a series of steps, and then writes all the static site files to the destination directory.

The Static Shock pipeline executes the following steps on each build:

  1. Delete everything in destination directory
  2. Initialize all plugins
  3. Load layout and components
  4. Load local data into the DataIndex
  5. Pick all files in the source directory that we want to process
  6. Load external data into the DataIndex
  7. Load external assets
  8. Load pages and assets from the source directory
  9. Transform pages
  10. Transform assets
  11. Filter pages
  12. Render pages
  13. Run finishers
  14. Write all files to the destination directory

The Static Shock pipeline might appear to include a lot of steps, but the reason for so many steps is to make Static Shock as pluggable as possible. As such, most Static Shock development can happen inside of plugins, instead of altering the core of Static Shock.

Pipeline Content

A collection of data structures (content) are pushed through the Static Shock pipeline to produce the final website files.

The most fundamental data structures in Static Shock are "pages" and "assets". A page represents a URL-addressable HTML page. An asset represents any URL-addressable file that's not a page, e.g., CSS, JavaScript, image, video.

To aid in assembling pages and assets, a few higher level data structures are provided, too. A DataIndex is provided, where plugins can store any data they'd like, which is made available to the rest of the pipeline. A PagesIndex holds references to all the pages that are loaded into the pipeline, which can be used to search for pages based on various criteria.