A simple static site loader for webpack that you can customize to your liking. It makes it easy to iterate over a file system, or interject content from any source you desire. Then you control the output path of your generated files. I built this so I could easily hook what I needed into the webpack build step.
The following is an example use case, you can change and tweak it to your liking, or do something crazy on your own.
Let's say you have a project with a folder content
that contains your sites files in markdown(or whatever you want to write your content in). These files happen to be structured how you would like your site to be structured.
You will need to add a index.js file to content: content/index.js
. What's in it is up to you, it just needs to be there so webpack can open the folder.
Here's a basic setup that will recursively parse those files and dump them into the built
directory as a static site to do with as you please... upload to s3, view with webpack-dev-server --content-base public/
, the possibilites are endless ⛵
webpack.config.js:
var pathUtil = require('path');
var marked = require('marked');
var jade = require('jade');
module.exports = {
...
entry: {
'site-generator': 'static-site-loader!./content',
...
},
output: {
path: 'built',
...
}
...
//These are custom options used to configure your site generator instance
//all of these function are called via .apply, so you will have this available to you
//as you would with a normal loader
staticSiteLoader: {
//perform any preprocessing tasks you might need here.
//compile a template to use, read some config settings from ./conten/index.js as source
preProcess: function(source, path) {
//watch the content directory for changes
this.addContextDependency(path);
//define the template file we'll use
var template = 'template.jade';
//watch the template for changes
this.addDependency(template);
//Compile the template and store it to this for later use
this.template = jade.compileFile(template, { pretty: false });
}
//Test if a file should be processed or not, should return a Boolean;
testToInclude: function(path, stats, absPath) {
//only use files that have the markdown extentsion
return pathUtil.extname(path) === '.md';
},
//Rewrite the url path used when written to output.path
rewriteUrlPath: function(path, stats, absPath) {
//strip out the extension
var urlPath = path.slice(0, -3);
//rewrite /index to be just /, making index.md files become the folder index properly
urlPath = urlPath.replace('index', '');
return urlPath;
},
processFile: function(file, content) {
var content = marked(content.replace(picoCMSMetaPattern, ''));
return this.template({content: content});
}
....
// or you can do the following for async content generation
processFile: function(file, content, callback) {
var content = marked(content.replace(picoCMSMetaPattern, ''));
callback(this.template({content: content}));
}
....
postProcess: function(files) {
// do something after all the files have been processed
// ex. collect links in previous steps and output an rss feed or site map
}
}
};