🍌 Django Bananas - Django extensions the monkey way

Install

django-bananas is on PyPI, so just run:

pip install django-bananas

Compatibility

Currently tested only for

• Django 1.8-1.11b1 under Python 3.4-3.6

pull requests welcome!

Examples

Models

TimeStampedModel

Abstract TimeStampedModel with date created/modified fields:

Use TimeStampedModel as base class for your model

from bananas.models import TimeStampedModel

class Book(TimeStampedModel):
    pass

the timestamps can be accessed on the model as

>>> book.date_created
>>> book.date_modified

UUIDModel

Abstract model that uses a Django 1.8 UUID field as the primary key.

from bananas.models import UUIDModel

class User(UUIDModel):
    display_name = models.CharField(max_length=255)
    email = models.EmailField()

>>> user.id
UUID('70cf1f46-2c79-4fc9-8cc8-523d67484182')

>>> user.pk
UUID('70cf1f46-2c79-4fc9-8cc8-523d67484182')

SecretField

Can be used to generate and store "safe" random bytes for authentication.

from bananas.models import SecretField

class User(models.Model):
    # Ask for 32 bytes and require 24 bytes from urandom
    token = SecretField(num_bytes=32, min_bytes=24)

>>> User.objects.create()  # Token is generated automatically
>>> user.token
'3076f884da827809e80ced236e8da20fa36d0c27dd036bdd4afbac34807e5cf1'

URLSecretField

An implementation of SecretField that generates an URL-safe base64 string instead of a hex representation of the random bytes.

from bananas.models import URLSecretField


class User(models.Model):
    # Generates an URL-safe base64 representation of the random value
    token = URLSecretField(num_bytes=32, min_bytes=24)

>>> user.token
'WOgrNwqFKOF_LsHorJy_hGpPepjvVH7Uar-4Z_K6DzU-'

ORM

New queryset.dicts() with field renaming through kwargs, and dot-dict style results:

from bananas.query import ExtendedQuerySet

class Book(TimeStampedModel):
    author = ForeignKey(Author)
    objects = Manager.from_queryset(ExtendedQuerySet)()

>>> book = Book.objects.dicts('id', author='author__name').first()
{'id': 1, 'author': 'Jonas'}
>>> book.author
'Jonas'

Admin

Custom django admin stylesheet.

Warning

Work in progress. Only a few views styled completely as of now.

# settings.py
INSTALLED_APPS = (
    'bananas',  # Needs to be before 'django.contrib.admin'
    'django.contrib.admin',
    ...
)

ADMIN = {
    'SITE_HEADER': 'Bananas',
    'SITE_TITLE': 'Bananas Admin',
    'INDEX_TITLE': 'Admin Panel',
    # 'BACKGROUND_COLOR': '#363c3f',
}

# your main urls.py
from bananas import admin

urlpatterns = [
    ...
    url(r'^admin/', include(admin.site.urls)),
]

Database URLs

Parse database information from a URL, kind of like SQLAlchemy.

Engines

Currently supported engines are:

URI scheme	Engine
pgsql, postgres, postgresql	django.db.backends.postgresql_psycopg2
mysql	django.db.backends.mysql
oracle	django.db.backends.oracle
sqlite, sqlite3	django.db.backends.sqlite3
mysqlgis	django.contrib.gis.db.backends.mysql
oraclegis	django.contrib.gis.db.backends.oracle
postgis	django.contrib.gis.db.backends.postgis
spatialite	django.contrib.gis.db.backends.spatialite

You can add your own by running register(scheme, module_name) before parsing.

database_conf_from_url(url)

Return a django-style database configuration based on url.

param url:	Database URL
return:	Django-style database configuration dict

Example:

>>> from bananas.url import database_conf_from_url
>>> conf = database_conf_from_url(
...     'pgsql://joar:[email protected]:4242/tweets/tweetschema'
...     '?hello=world')
>>> sorted(conf.items())  # doctest: +NORMALIZE_WHITESPACE
[('ENGINE', 'django.db.backends.postgresql_psycopg2'),
 ('HOST', '5monkeys.se'),
 ('NAME', 'tweets'),
 ('PARAMS', {'hello': 'world'}),
 ('PASSWORD', 'hunter2'),
 ('PORT', 4242),
 ('SCHEMA', 'tweetschema'),
 ('USER', 'joar')]

bananas.environment - Helpers to get setting values from environment variables

bananas.environment.env is a wrapper around os.environ, it provides the standard .get(key, value), method to get a value for a key, or a default if the key is not set - by default that default is None as you would expect. What is more useful is the additional type-parsing .get_* methods it provides:

• get_bool
• get_int
• get_list, get_set, get_tuple

get_int:	>>> # env ONE=1 >>> env.get_int('ONE') 1 >>> env.get_int('TWO') # Not set None >>> env.get_int('TWO', -1) # Not set, default to -1 -1
get_bool:	returns True if the environment variable value is any of, case-insensitive: "true" "yes" "on" "1" returns False if the environment variable value is any of, case-insensitive: "false" "no" "off" "0" if the value is set to anything other than above, the default value will be returned instead. e.g.: >>> # env CAN_DO=1 NO_THANKS=false NO_HABLA=f4lse >>> env.get_bool('CAN_DO') True >>> env.get_bool('NO_THANKS') False >>> env.get_bool('NO_HABLA') # Set, but not valid None >>> env.get_bool('NO_HABLA', True) # Set, but not valid, with default True >>> env.get_bool('IS_NONE') # Not set None >>> env.get_bool('IS_NONE', False) # Not set, default provided False
get_tuple, get_list, get_set:	Returns a tuple, list or set of the environment variable string, split by the ascii comma character. e.g.: >>> # env FOOS=foo,foo,bar >>> get_list('FOO') ['foo', 'foo', 'bar'] >>> get_set('FOO') set(['foo', 'bar'])