7be7041aa2
- keep using find-exec (more bin files) - parse_convert command with spaces works |
||
---|---|---|
bin | ||
example | ||
lib | ||
config.mk | ||
LICENSE | ||
Makefile | ||
README.md |
zodiac - a static site generator
ZODIAC is a static website generator powered by sh and awk. The core features of zodiac are:
- utliization of existing tools (i.e. awk, sh, find, etc.)
- built-in support for markdown
- a simple, easy to use templating system
- supports custom helpers written in awk
- multiple layouts
- partials
- configuration, meta, helpers, etc. can be added as you need them
- support any markup format via external tools
SYNOPSIS
zod projectdir targetdir
INSTALL
git clone git://github.com/nuex/zodiac.git
Edit the config.mk file to customize the install paths. /usr/local
is the default install prefix.
Run the following (as root if necessary):
make install
DESCRIPTION
A typical Zodiac project will look something like this:
site/
index.md
index.meta
main.layout
global.meta
projects/
project-1.md
project-1.meta
project-2.md
project-2.meta
cv.md
cv.meta
stylesheets/
style.css
And it's output could look like this:
site/
index.html
projects/
project-1.html
project-2.html
cv.html
stylesheets/
style.css
Meta
.meta
files contain a key / value pair per line. A key and its value must be separated by a ": ". A metafile looks like this:
this: that
title: Contact
author: Me
Each page can have its own meta file. The only requirement is that the meta file is in the same directory as the page, has the same name as the page and has the .meta
file extension.
The optional global.meta
file contains data that is available to all of your site's pages, like a site title.
Page metadata will always override global metadata of the same key.
Templates
Templates come in two forms, page templates and layout templates. Metadata can be bound to templates by using the {{key}}
notation in your pages and layout files.
Page templates can be either markdown files with an .md
extension or plain HTML files with a .html
extension.
The main.layout
file wraps HTML content around a page template. A main.layout
file could look something like this:
<!DOCTYPE html>
<html lang="en">
<head>
<meta charset="utf-8" />
<link rel="stylesheet" href="/stylesheets/style.css" />
<title>{{page_title}}</title>
</head>
<body>
<header>
<h1><a href="/">{{site_title}}</a></h1>
</header>
<article>
{{{yield}}}
</article>
<footer>
<p>powered by static files, compiled by <a href="http://nu-ex.com/projects/zodiac">zodiac</a>.</p>
</footer>
</body>
</html>
{{{yield}}}
is a special tag that renders the page content within the layout. {{{yield}}}
can only be used in the main.layout
file.
Helpers
The helpers.awk
file is an awk script that can make custom data available to your templates. You also have access to the page and global data. Here is a peak at the script included in the examples folder:
{ helpers = "yes" }
function load_helpers() {
# your custom data settings
data["page_title"] = page_title()
}
# your custom functions
function page_title( title) {
if ("title" in data) {
return data["title"] "-" data["site_title"]
} else {
return data["site_title"]
}
}
Just be sure to set the data array in the load_helpers()
function at the top of the script to make your custom data available to the template.
Config
For more control over the parsing and conversion process, a .zod/config
file can be created within your project directory. Here is a sample config:
[parse]
htm
html
[parse_convert]
md smu
txt asciidoc -s -
[convert]
coffee coffee -s > {}.js
[ignore]
Makefile
Here we're only parsing (not converting to a different format) files matching *.htm
and *.html
.
Files matching *.md
are going to be parsed and converted using the smu
markdown parsing program.
Files matching *.txt
are going to be parsed and converted using asciidoc
.
Files matching *.coffee
are going to be converted to JavaScript before being copied into the target directory. The {}
notation will be expanded to the path and filename of the coffeescript file, but without the .coffee
extension.
Files matching Makefile
will be ignored and not copied.
Conversion programs must accept a UNIX-style pipe and send converted data to stdout.
Per-page Templates and Partials
Multiple templates and partials are also supported.
For example; a blog.md
page:
title: my blog layout: blog
A sidebar.partial
:
And blog.layout
:
{{>sidebar}}
will be replaced with the parsed contents of sidebar.partial
.
CREDITS
- zsw: for the introduction to parameter expansion and other shell scripting techniques
LICENSE
MIT