Content Translation Framework based on Postgresql's JSONB field
Clone or download
Latest commit 44a6f7a Nov 22, 2018
Permalink
Type Name Latest commit message Commit time
Failed to load latest commit information.
images nece meaning Jan 19, 2016
nece #24 fix bugs Sep 5, 2017
tests Add Django 2.0 & Python 3.7 support Nov 19, 2018
.gitignore Initial commit Jan 16, 2016
.travis.yml Add Django 2.0 & Python 3.7 support Nov 19, 2018
CHANGELOG.md changelog Aug 21, 2017
LICENSE Initial commit Jan 16, 2016
README.rst Add yourself to contributors Nov 19, 2018
__init__.py needed an __init__.py at root Jan 18, 2016
requirements.txt Fix setup: don't use requirements.txt Aug 24, 2017
runtests.py runtest.py -> runtests.py Jan 20, 2016
setup.py version bump Sep 21, 2017

README.rst

nece?

Introduction

nece

A “Content Translation Framework” using Postgresql’s jsonb field. It simply sets and gets translations from a jsonb field called translations.

Why?

You might ask why you should use django-nece since there are other, and more mature content translation frameworks like django-hvad and django-modeltranslation. Both of them are good in some ways, worst in others.

For instance, it is very hard for django-hvad users to get default language if there is no corresponding translation for an object. And it holds translated values in a different table, so every translation query results in another hit to the database.

On the other hand django-modeltranslation adds multiple additional fields for multiple languages. The number of fields inceases by the number of languages you need to support. At the end it becomes a huge chunk of an object if you need to add more than 20 languages.

nece? more or less works like the latter one with an important difference. It uses Postgresql’s new JSONB field to hold translation information. And overrides the original one on query.

Dependencies

postgresql >= 9.4.5
Django >= 1.9
psycopg2 >= 2.5.4

Installation

via pypi:

pip install nece

via setup.py

python setup.py install

Usage

Lets say we have a model called Fruit:

from nece.models import TranslationModel

class Fruit(TranslationModel):
    name = CharField(max_length=255)

    def __str__(self):
        return self.name

    class Meta:
        translatable_fields = ('name',)

TranslationModel adds a jsonb field to this table and sets translations in a notation like the one below:

{u'de_de': {u'name': u'Apfel'},
 u'tr_tr': {u'name': u'elma'}}

When we need the German translation we can simply choose the language and get the attribute as usual:

>> f = Fruit.objects.get(name='apple')
>> print(f.name)
apple
>> f.language('de_de')
>> print(f.name)
Apfel

You can also filter out the ones containing any language translation:

>> Fruit.objects.all()
[<Fruit: apple>, <Fruit: pear>, <Fruit: banana>]
>> Fruit.objects.language('tr_tr')
[<Fruit: elma>, <Fruit: armut>]  # there is no translation for banana
>> Fruit.objects.language_or_default('tr_tr')
[<Fruit: elma>, <Fruit: armut>, <Fruit: banana>]
>> Fruit.objects.language('tr_tr').filter(name='elma')
[<Fruit: elma>]
>> Fruit.objects.language('tr_tr').get(name='elma')
<Fruit: elma>

Updating translations

>> fruit._language_code
tr_tr
>> fruit.name
elma
>> fruit.translate(name='armut').save()
>> fruit.name
armut
>> fruit.language('en')
>> fruit.translate('it_it', name='pera')
>> fruit.language('it_it')
>> fruit.name
pera

Settings

TRANSLATIONS_DEFAULT

Default language code. Default value: `en_us`

TRANSLATIONS_MAP

Shortcuts for `languagecode_countrycode` notation.

Example:

TRANSLATIONS_MAP = {
    "en": "en_us",
    "tr": "tr_tr",
    "ar": "ar_sy",
    "bg": "bg_bg",
    "cs": "cs_cz",
    "da": "da_dk",
    ...
}

Default:

{'en': 'en_us'}

Contributors & Thanks

Change Log