name pending
This is a little experiment for building a static site generator that is based around routes.
Think of it like you define the desired routes, but instead of controllers (from a typical mvc application) the "routes" are matched from files.
For example
const routes = {
// the / route will be matched from content/index.html
'/': {
from: 'content/index.html'
},
// :slug is a variable that will be taken from the file name
'/blog/:slug/': {
from: 'content/posts/:slug.md'
},
// :path+ means a repeating pattern, so we can deeply nest
// for example categories
'/categories/:path+/': {
from: 'content/categories/:path+.md'
}
// parameters can also be optional using the :param? syntax
// see https://www.npmjs.com/package/path-to-regexp for all
// available types of parameters.
};By default easy file will be run through a default pipeline, which will basically read the contents of the file, and then write a new file in the dist folder at the given path.
In many cases we want to apply different transformations to a file, this is done by specifying a pipeline in the route definition.
The pipeline property can be an array for async functions (or promises)
or an instance of the lib/Pipeline class.
A pipeline is just a simple chain of functions. The order is important, because every function in the pipeline will be run one after another.
Important: always return a value from a pipeline function, because the next function in the line will receive it as the parameter, and if you forget to return, the data will be "lost".
const routes = {
'/blog/:slug/': {
from: 'content/posts/:slug.md',
pipeline: [
matter // parses front matter
markdown // transforms markdown to html
async file => {
// a step in the pipeline is just an async function
// file.content is a Buffer by default, to allow
// working with binary files
file.content = file.content.toString().toUpperCase();
return file;
}
]
},
}Released under the MIT license.