Build Status Coverage Status

An HTML parser/emitter for CommonDoc.


(defvar node
    (:title "My Document"
     :creator "me"
     :keywords (list "test" "test1"))
      (:text "test"))))))

(common-html.emitter:node-to-html-string node) ;; => "<p>test</p>"


CommonHTML let's you customize the HTML output by inserting attributes (such as class names or data attributes) in node metadata. All CommonDoc nodes hold optional metadata, as key-value pairs. Every pair where the key starts with html: will be emitted with this prefix removed.

So, for instance, if a node has a metadata pair like "html:class" => "theorem", the resulting HTML for that node will have class="theorem".

Multi-file emission

Normally, a document is emitted into HTML as a single file. You can also perform Texinfo/Sphinx style emission, where a document is broken up into sections, and each section (Up to a certain depth, or any depth) is emitted as a different file.

To emit a document into multiple files, simply do:

(common-html.multi-emit:multi-emit doc #p"output-directory/")

An optional keyword argument, :max-depth, can be provided to choose at what section depth to stop emitting each section in a different file. For instance, if you have a document that looks like this:

  1. Intro
  2. Overview
  3. History
    1. Motivation
  4. Tutorial

Emitting it with the default :max-depth of nil will produce 5 files, while emitting it with a :max-depth of 2 will produce four files: One for each of the Intro, Overview and Tutorial subsections, and another for both the History section and its Motivation subsection.

How it Works

Multi-part file emission can be complicated.

First, some obvious choices, and how CommonHTML chooses:

  1. Should the directory structure of the HTML output mirror that of the sections? Or should all HTML files be emitted within the same directory? Answer: For simplicity (Users might not expect or want nested files), all HTML files are emitted into the same directory.

  2. What should the name of the resulting HTML files be? The pure name of the section? The result of calling common-doc.util:string-to-slug on the section? Or an autogenerated ID? Answer: The first option might produce invalid pathnames, and the last option is an inconceivable abomination, so we just go with slugifying the section text.


Copyright (c) 2014-2015 Fernando Borretti

Licensed under the MIT License.

Fernando Borretti <>
Fernando Borretti <>