toto

the tiniest blogging engine in Oz!

introduction

toto is a git-powered, minimalist blog engine for the hackers of Oz. The engine weighs around ~300 sloc at its worse. There is no toto client, at least for now; everything goes through git.

blog in 10 seconds

$ git clone git://github.com/cloudhead/dorothy.git myblog
$ cd myblog
$ heroku create myblog
$ git push heroku master

philosophy

Everything that can be done better with another tool should be, but one should not have too much pie to stay fit. In other words, toto does away with web frameworks or DSLs such as sinatra, and is built right on top of rack. There is no database or ORM either, we use plain text files.

Toto was designed to be used with a reverse-proxy cache, such as Varnish. This makes it an ideal candidate for heroku.

Oh, and everything that can be done with git, is.

how it works

content is entirely managed through git; you get full fledged version control for free.
articles are stored as .txt files, with embeded metadata (in yaml format).
articles are processed through a markdown converter (rdiscount) by default.
templating is done through ERB.
toto is built right on top of Rack.
toto was built to take advantage of HTTP caching.
toto was built with heroku in mind.
comments are handled by disqus
individual articles can be accessed through urls such as /2009/11/21/blogging-with-toto
the archives can be accessed by year, month or day, wih the same format as above.
arbitrary metadata can be included in articles files, and accessed from the templates.
summaries are generated intelligently by toto, following the :max setting you give it.
you can also define how long your summary is, by adding ~ at the end of it (:delim).

dorothy

Dorothy is toto's default template, you can get it at http://github.com/cloudhead/dorothy. It comes with a very minimalistic but functional template, and a config.ru file to get you started. It also includes a .gems file, for heroku.

synopsis

One would start by installing toto, with sudo gem install toto, and then forking or cloning the dorothy repo, to get a basic skeleton:

$ git clone git://github.com/cloudhead/dorothy.git weblog
$ cd weblog/

One would then edit the template at will, it has the following structure:

templates/
|
+- layout.rhtml      # the main site layout, shared by all pages
|
+- index.builder     # the builder template for the atom feed
|
+- pages/            # pages, such as home, about, etc go here
   |
   +- index.rhtml    # the default page loaded from `/`, it displays the list of articles
   |
   +- article.rhtml  # the article (post) partial and page
   |
   +- about.rhtml

One could then create a .txt article file in the articles/ folder, and make sure it has the following format:

title: The Wonderful Wizard of Oz
author: Lyman Frank Baum
date: 1900/05/17

Dorothy lived in the midst of the great Kansas prairies, with Uncle Henry,
who was a farmer, and Aunt Em, who was the farmer's wife.

If one is familiar with webby or aerial, this shouldn't look funny. Basically the top of the file is in YAML format, and the rest of it is the blog post. They are delimited by an empty line /\n\n/, as you can see above. None of the information is compulsory, but it's strongly encouraged you specify it. Note that one can also use rake to create an article stub, with rake new.

Once he finishes writing his beautiful tale, one can push to the git repo, as usual:

$ git add articles/wizard-of-oz.txt
$ git commit -m 'wrote the wizard of oz.'
$ git push remote master

Where remote is the name of your remote git repository. The article is now published.

deployment

Toto is built on top of Rack, and hence has a rackup file: config.ru.

on your own server

Once you have created the remote git repo, and pushed your changes to it, you can run toto with any Rack compliant web server, such as thin, mongrel or unicorn.

With thin, you would do something like:

$ thin start -R config.ru

With unicorn, you can just do:

$ unicorn

on heroku

Toto was designed to work well with heroku, it makes the most out of it's state-of-the-art caching, by setting the Cache-Control and Etag HTTP headers. Deploying on Heroku is really easy, just get the heroku gem, create a heroku app with heroku create, and push with git push heroku master.

$ heroku create weblog
$ git push heroku master
$ heroku open

configuration

You can configure toto, by modifying the config.ru file. For example, if you want to set the blog author to 'John Galt', you could add set :author, 'John Galt' inside the Toto::Server.new block. Here are the defaults, to get you started:

set :author,      ENV['USER']                               # blog author
set :title,       Dir.pwd.split('/').last                   # site title
set :url,         'http://example.com'                      # site root URL
set :prefix,      ''                                        # common path prefix for all pages
set :permalink,   '/:year/:month/:day/:title'               # structure of article's URLs
set :root,        "index"                                   # page to load on /
set :date,        lambda {|now| now.strftime("%d/%m/%Y") }  # date format for articles
set :markdown,    :smart                                    # use markdown + smart-mode
set :disqus,      false                                     # disqus id, or false
set :summary,     :max => 150, :delim => /~\n/              # length of article summary and delimiter
set :ext,         'txt'                                     # file extension for articles
set :cache,       28800                                     # cache site for 8 hours
set :github,      :user => "", :repos => [], :ext => 'md'   # get READMEs from this user's repositories

set :to_html   do |path, page, ctx|                         # returns an html, from a path & context
  ERB.new(File.read("#{path}/#{page}.rhtml")).result(ctx)
end

set :error     do |code|                                    # The HTML for your error page
  "<font style='font-size:300%'>toto, we're not in Kansas anymore (#{code})</font>"
end

permalinks

Toto includes supports for custom URLs for the articles. Prior to the inclusion of this feature, all article's URLs were forced to have the structure "/year/month/day/title". Now, you can modify that path, to include (or exclude) whatever data you want. The custom permalink's syntax follows that of Jekyll's, so anyone familiar with it should feel right at home with Toto's permalinks. Toto provides 4 preexisting variables:

:year - The year the article was made, derived from the date
:month - The month the article was made, derived from the date
:day - The day the article was made, derived from the date
:title - The title of the article, with URL un-safe characters removed. (The slug)

Also, any extra YAML front matter can be used in the structure. If you define one of the pre-existing variables in the front matter (excluding title), the value in the front-matter will override the derived variable. :title will always be the slug.

thanks

To heroku for making this easy as pie. To adam wiggins, as I stole a couple of ideas from Scanty. To the developpers of Rack, for making such an awesome platform.