Jekyll collections: process markdown within a plugin - jekyll

Question

I'm trying to adapt an existing Jekyll plugin (taken from here) in order to generate a .json version of every document in a collection.
However, I'm having trouble getting my content to convert from markdown to HTML (which I would then like to convert/encode into JSON). In Jekyll, a collection "document" is different from a "post", and while posts have access to a transform method that does what I need, it looks like "documents" do not.
Is there some other straight-forward way to feed content to a markdown parser in the context of a Jekyll plug-in?
Here's the plug-in code I've been working with so far. This is generating JSON, but markdown is not being converted into HTML (markdown syntax like ** remains in the file).
module Jekyll
class JSONPage < Page
def initialize(site, base, dir, name, content)
#site = site
#base = base
#dir = dir
#name = name
self.data = {}
self.content = content
process(#name)
end
def read_yaml(*)
# Do nothing
end
def render_with_liquid?
false
end
end
class JSONPageGenerator < Generator
safe true
def generate(site)
site.documents.each do |document|
# Set the path of the JSON version
path = "#{document.collection.label}" + document.cleaned_relative_path + ".json"
output = document.to_liquid
# Delete unnecessary metadata
['layout', 'output'].each { |key| output.delete(key) }
site.pages << JSONPage.new(site, site.source, File.dirname(path), File.basename(path), output)
end
end
end
end

Answer 1

Okay, turns out the answer is simple – you can just require 'kramdown' or any other markdown generator directly in the plugin.
module Jekyll
class JSONPage < Page
def initialize(site, base, dir, name, content)
#site = site
#base = base
#dir = dir
#name = name
self.data = {}
self.content = content
process(#name)
end
def read_yaml(*)
# Do nothing
end
def render_with_liquid?
false
end
end
class JSONPostGenerator < Generator
safe true
def generate(site)
require 'kramdown'
site.documents.each do |document|
# Set the path of the JSON version
path = "#{document.collection.label}" + document.cleaned_relative_path + ".json"
output = document.to_liquid
output['content'] = Kramdown::Document.new(document.content).to_html.gsub(/\n/, "")
# Delete unnecessary metadata
['layout', 'output'].each { |key| output.delete(key) }
site.pages << JSONPage.new(site, site.source, File.dirname(path), File.basename(path), output.to_json)
end
end
end
end