Django make will apply a set of build scripts for getting a django application hosted from an empty box to production.

Yep, another build script with current features:

• Installer folder / install.py

  An app to run all the boilerplate required for a new project.

  • project creation much like django-admin startproject

    • basic apps installed including:
      • south
      • admin
      • django-admin-tools
      • django-cms
      • django-extensions
      • django-debug-toolbar

  • auto PIP, installing all dependencies

  • Initial proect templating

  • database setup, sync, migrations, supporting:

    • mysql
    • sqlite3

  • deployment to one/all of the following:

    • WSGI
    • Dev (with dev media setup)
    • cherrypy

Why am I building this:

I'm a python developer - go figure. I love Django but I don't love deployments. I must deploy offsite, of which is handled by an engineer; of which in turn is handled by myself over a remote desktop or a camera pointing at a CRT with me trying to yelp over the sound of background servers 220 miles away. At this point they don't know what a python is or what the where f*ck django has his guitar. Couple that with my eggs, a pip bundle, some south migrations and a git in the middle, it's no wonder why my collegues think I'm a foreign language.

Let's make it easy - both for me as a developer wanting to have a ready site in minutes.

Great for:

• Testbed websites for creating sites as quick as you think of them.
• Personal deployement strategy for new websites.
• Creating installation routines for remote deployment
• Creating installers for third-party deployment
• All apps are made in the same manor.
• Never miss a step.

The pip installer is a wrapper to pip of which can be managed via the code. Currently it simply installs.

# Pip install
>>> from utils import *
>>> pip_install('yolk')
# installs yolk

Yes another confimation dialog. This allows force answering, default and obviously question.

# Confirm dialog
>>> confirm(prompt='Create Directory?', resp=True)
Create Directory? [y]|n: 
True
>>> confirm(prompt='Create Directory?', resp=False)
Create Directory? [n]|y: 
False
>>> confirm()
Confirm? [n]|y: y
True

DictClass is a tiny utility to convert flat objects to class callables.

# DictClass for turning objects into class style
>>> x = DictClass({'name': 'Fishy', 'word':'w00t!'})
>>> x.name
'Fishy'

get_setting is (mostly) framework dependant. It returns a value by attempting a chain of inputs:

1. passed config files
2. command line argument
3. pipe # NOT IMPLEMENTED
4. terminal prompt

in that order.

# Get Setting functionality. Providing full waterfall fallthough
# of values using: Config File JSON, argument settings, user input
>>> name = get_setting('site_name', prompt="What's the site name? ")
# Fallthrough
'project name'    

Config files are JSON mapped variables using variable names matching arguments and application settings.

# Tiny default.conf example
{
    "name": "Jimmy",
    "verbose": true
} 

# Supply tuple, tuple of configs.

>>> configs = (
           ('default', 'config/default.conf',),
           # ('dev', 'config/development.conf',),
           # ('production', 'config/production.conf',),
        )
>>> parse_config()

Library handling - for matching and collecting using PIP missing packages. Map packages is a working progress paradim; supply the PIP package and the import string required for required functionality off the application. On initial script load and install

# A list of real imports for use with aut importing of required
# scripts
>>> map_package = (
        ('termcolor', 'from termcolor import colored', ),
        ('docopt', 'from docopt import docopt', ),
        ('fabric', 'from fabric.api import settings, run', ),
        ('clint', 'from clint import resources', ),
        ('yolk', 'import yolk', ),
)

# If package is installed
>>> have('yolk')
True

# Run install if required before continuing
>> def run(): pass # Fancy function
>>> depend_on('yolk', run)