Haze

Haze is a minimalistic blogging engine, in the spirit of Honk, it's predecessor.

It has very few features compared to other blog engines. Exhaustively:

• Entries, stored in flat text files
• Tags
• An archive page that lists every entry
• An Atom feed
• Drafts
• A very simple comment system

Why so few? Well, I simply don't need more.

Haze's source code is very short (~200LOC) thus the app is light and quick.

Requirements

• sinatra (tested with 0.9.4)

Install

$ git clone git://github.com/madx/haze.git

Copy the files where you want to install it, then edit config.ru if you want to change the defaults. Use plain ruby to do the configuration, like this:

require File.join(File.dirname(__FILE__), 'lib', 'haze')

Haze.set :title,  "Blog title"
Haze.set :author, "Your name"
Haze.set :domain, "http://example.com/"
Haze.set :email,  "[email protected]"
# Replace KEY by a sha1sum of something
Haze.set :key,    "KEY"

Haze.reload!

run Haze::App

Create the folder where you will put your entries: mkdir entries. Then run the app with $ rackup -E production config.ru.

Source

Haze's Git repo is available on GitHub, which can be browsed at http://github.com/madx/haze and cloned with:

git clone git://github.com/madx/haze.git

Usage

Haze entry format enforces a few conventions:

An entry is split in two parts. The first is a header and the second the entry body. The separator is a succession of at least three - put on a line of their own.

The header will be the title of your entry. Every word starting with a # will create a new tag. Tags are automatically lowercase'd. You can add tags that won't show up in the title by enclosing them in braces ({}).

To determine the order of entries and their URLs, you have to give a correct name to the file they are stored into.

The format is: <date>[+<counter>]_<slug>.(hz/draft).

<date> must be parseable by Time.parse, the handiest format to use is probably <year>-<month>-<day>. The +<counter> part allows you to write multiple entries with on single <date>. Behind the scenes, it simply adds <counter> seconds to the parsed date.

<slug> will be the name of the URL for your entry. hz or draft tells if the file is a regular entry or a draft.

Drafts are viewable using the url /draft/<slug>.

Example:

$ cat entries/2009-01-08_helloworld.hz
A #Ruby Hello world program {#tutorial,#programming}
---
<p>Hello world is the most common program used to demonstrate a language's
syntax. Here is one in Ruby: </p>

<pre><code>puts "hello world"</code></pre>

• Title: A Ruby Hello world program
• Tags: ruby, tutorial, programming
• Date: 2009-01-08
• Slug: helloworld

Syncing content

To synchronize the blog with contents, just hit the url /_sync_?key=<KEY>, replacing <KEY> with the value set in config.ru.

To automatically update your post as you update the files, you may use mynyml's watchr or ttilley's fssm and integrate a script in the config.ru or using an external tool such as inotifywait.

Static pages

Haze supports static pages. Just create a static folder in the root directory and put raw HTML pages in it. They will be rendered inside the layout and are available at /static/<file name>.

Issues

There may be a bug with the Encoding class and HAML under Ruby 1.9, something about Encoding.default_internal. To fix it, add this somewhere in your config.ru:

class ::Encoding
  def self.default_internal
    "utf-8" # Or the right encoding
  end
end

License

Haze is published under the terms of the MIT license, you can find a copy of that license in the LICENSE file