Research
Security News
Malicious npm Package Targets Solana Developers and Hijacks Funds
A malicious npm package targets Solana developers, rerouting funds in 2% of transactions to a hardcoded address.
Django Garnett is a field level translation library that allows you to store strings in multiple languages for fields in Django - with minimal changes to your models and without having to rewrite your code.
Want a demo? https://django-garnett.herokuapp.com/
Made with ♥ by Aristotle Metadata
In summary it allows you to do this:
models.py | You can do this! |
By changing your models from this...
to this...
|
|
Tested on:
Tested with the following libraries:
Pros:
Model.field_name = "some string"
and print(Model.field_name)
work as you would expectF()
and Q()
objects within the ORM - and when it doesn't we have a language aware LangF()
replacement you can use.Cons:
queryset.values_list
or queryset.values
natively - but we have an easy Queryset Mixin to add language support for both below.A few reasons:
Note: Field language is different to the django display language. Django can be set up to translate your pages based on the users browser and serve them with a user interface in their preferred language.
Garnett does not use the browser language by design - a user with a French browser may want the user interface in French, but want to see content in English or French based on their needs.
Add django-garnett
to your dependencies. eg. pip install django-garnett
Convert your chosen field using the Translated
function
title = fields.Translated(models.CharField(*args))
Add GARNETT_TRANSLATABLE_LANGUAGES
(a callable or list of language codes) to your django settings.
Note: At the moment there is no way to allow "a user to enter any language".
Add GARNETT_DEFAULT_TRANSLATABLE_LANGUAGE
(a callable or single language code) to your settings.
Re-run django makemigrations
, perform a data migration to make your existing data translatable (See 'Data migrations') & django migrate
for any apps you've updated.
Thats mostly it.
You can also add a few optional value adds:
(Optional) Add a garnett middleware to take care of field language handling:
You want to capture the garnett language in a context variable available in views use: garnett.middleware.TranslationContextMiddleware
You want to capture the garnett language in a context variable available in views, and want to raise a 404 if the user requests an invalid language use: garnett.middleware.TranslationContextNotFoundMiddleware
(Future addition) You want to capture the garnett language in a context variable available in views, and want to redirect to the default language if the user requests an invalid language use: garnett.middleware.TranslationContextRedirectDefaultMiddleware
If you want to cache the current language in session storage use garnett.middleware.TranslationCacheMiddleware
after one of the above middleware (this is useful with the session selector mentioned below)
(Optional) Add the garnett
app to your INSTALLED_APPS
to use garnett's template_tags. If this is installed before django.contrib.admin
it also include a language switcher in the Django Admin Site.
(Optional) Add a template processor:
garnett.context_processors.languages
this will add garnett_languages
(a list of available Language
s) and garnett_current_language
(the currently selected language).(Optional) Add a custom translation fallback:
By default, if a language isn't available for a field, Garnett will show a mesage like:
No translation of this field available in English
You can override this either by creating a custom fallback method:
Translated(CharField(max_length=150), fallback=my_fallback_method))
Where my_fallback_method
takes a dictionary of language codes and corresponding strings, and returns the necessary text.
Additionally, you can customise how django outputs text in templates by creating a new
TranslationStr
class, and overriding the __html__
method.
If you have lots of existing data (and if you are using this library you probably do) you will need to perform data migrations to make sure all of your existing data is multi-lingual aware. Fortunately, we've added some well tested migration utilities that can take care of this for you.
Once you have run django-admin makemigrations
you just need to add the step_1_safe_encode_content
before and step_2_safe_prepare_translations
after your schema migrations, like in the following example:
# Generated by Django 3.1.13 on 2022-01-11 10:13
from django.db import migrations, models
import garnett.fields
import library_app.models
#### Add this line in
from garnett.migrate import step_1_safe_encode_content, step_2_safe_prepare_translations
#### Define the models and fields you want ot migrate
model_fields = {
"book": ["title", "description"],
}
class Migration(migrations.Migration):
dependencies = [
("library_app", "0001_initial"),
]
operations = [
## Add this operation at the start
step_1_safe_encode_content("library_app", model_fields),
## These are the automatically generated migrations
migrations.AlterField( # ... migrate title to TranslatedField),
migrations.AlterField( # ... migrate description to TranslatedField),
## Add this operation at the start
step_2_safe_prepare_translations("library_app", model_fields),
]
Language
vs languageDjango Garnett uses the python langcodes
library to determine more information about the languages being used - including the full name and local name of the language being used. This is stored as a Language
object.
GARNETT_DEFAULT_TRANSLATABLE_LANGUAGE
'en-AU'
GARNETT_TRANSLATABLE_LANGUAGES
:
[GARNETT_DEFAULT_TRANSLATABLE_LANGUAGE]
GARNETT_REQUEST_LANGUAGE_SELECTORS
:
garnett.selector.query
: Checks the GARNETT_QUERY_PARAMETER_NAME
for a language to displaygarnett.selector.cookie
: Checks for a cookie called GARNETT_LANGUAGE_CODE
for a language to display.
Note: you cannot change this cookie name.garnett.selector.session
: Checks for a session key GARNETT_LANGUAGE_CODE
for a language to display.
Note: you cannot change this key name.garnett.selector.header
: Checks for a HTTP Header called X-Garnett-Language-Code
for a language to display.
Note: you cannot change this Header name.garnett.selector.browser
: Uses Django's get_language
function to get the users browser/UI language as determined by Django.['garnett.selectors.header', 'garnett.selectors.cookie']
.['garnett.selectors.header', 'garnett.selectors.query', 'garnett.selectors.cookie']
GARNETT_QUERY_PARAMETER_NAME
:
glang
GARNETT_ALLOW_BLANK_FALLBACK_OVERRIDE
Advanced Settings (you probably don't need to adjust these)
GARNETT_TRANSLATABLE_FIELDS_PROPERTY_NAME
:
translatable_fields
GARNETT_TRANSLATIONS_PROPERTY_NAME
:
translations
If you did everything above correctly, garnett should for the most part "just work".
Garnett comes with a handy context manager that can be used to specify the current language. In any place where you want to manually control the current language, wrap your code in set_field_language
and garnett will correctly store the language. This can be nested, or you can change the language for a context multiple times before saving.
from garnett.context import set_field_language
greeting = Greeting(text="Hello", target="World")
with set_field_language("en"):
greeting.text = "Hello"
with set_field_language("fr"):
greeting.text = "Bonjour"
greeting.save()
values_list
or values
This is one of the areas that garnett doesn't work immediately, but there is a solution.
In the places you are using values lists or values
, wrap any translated field in an L-expression and the values list will return correctly. For example:
from garnett.expressions import L
Book.objects.values_list(L("title"))
Book.objects.values(L("title"))
As TranslationField
s are based on JSONField, by default Django-Rest-Framework renders these as a JSONField, which may not be ideal.
You can get around this by using the TranslatableSerializerMixin
as the first mixin, which adds the necessary hooks to your serializer. This will mean class changes, but you won't need to update or override every field.
For example:
from rest_framework import serializers
from library_app import models
from garnett.ext.drf import TranslatableSerializerMixin
class BookSerializer(TranslatableSerializerMixin, serializers.ModelSerializer):
class Meta:
model = models.Book
fields = "__all__"
This will allow you to set the value for a translatable as either a string for the active langauge, or by setting a dictionary that has all languages to be saved (note: this will override the existing language set). For example:
To override just the active language:
curl -X PATCH ... -d "{ \"title\": \"Hello\"}"
To specifically override a single language (for example, Klingon):
curl -X PATCH ... -H "X-Garnett-Language-Code: tlh" -d "{ \"title\": \"Hello\"}"
To override all languages:
curl -X PATCH ... -d "{ \"title\": {\"en\": \"Hello\", \"fr\": \"Bonjour\"}}"
There are a few minor tweaks required to get Garnett to operate properly with django-reversion and django-reversion-compare based on how they serialise and display data.
This is because Garnett does not use the same 'field.attname' and 'field.name' which means serialization in Django will not work correctly.
To get django-reversion to work you will need to use a translation-aware serialiser and apply a patch to ensure that django-reversion-compare can show the right information.
An example json translation-aware serializer is included with Garnett and this can be applied with the following two settings in settings.py
:
# In settings.py
GARNETT_PATCH_REVERSION_COMPARE = True
SERIALIZATION_MODULES = {"json": "garnett.serializers.json"}
TranslatedFields will list the history and changes in json, but it does do comparisons correctly.
contains
acts like icontains
on SQLite only when doing a contains query
it does a case insensitive search. I don't know why - https://www.youtube.com/watch?v=PgGNWRtceagAdminTextAreaWidget
on translated fields in the django admin site by default. They can however
be specified explicitly on the corresponding admin model form.There is a /dev/
directory with a docker-compose stack you can ues to bring up a database and clean development environment.
There are a few good options for adding translatable strings to Django that may meet other use cases. We've included a few other options here, their strengths and why we didn't go with them.
Pros: this library lets you apply translations to external apps, without altering their models.
Cons: each translation adds an extra column, which means languages are specified in code, and can't be altered by users later.
Pros: uses a great context processor for switching lanugages (which is where we got the idea for ours).
Cons: Languages are specified in django-settings, and can't be altered by users later.
Pros: Django admin site support.
Cons: Languages are stored in a separate table and can't be altered by users later. Translated fields are specified in model meta, away from the fields definition which makes complex lookups harder.
FAQs
Simple translatable Django fields
We found that django-garnett demonstrated a healthy version release cadence and project activity because the last version was released less than a year ago. It has 1 open source maintainer collaborating on the project.
Did you know?
Socket for GitHub automatically highlights issues in each pull request and monitors the health of all your open source dependencies. Discover the contents of your packages and block harmful activity before you install or update your dependencies.
Research
Security News
A malicious npm package targets Solana developers, rerouting funds in 2% of transactions to a hardcoded address.
Security News
Research
Socket researchers have discovered malicious npm packages targeting crypto developers, stealing credentials and wallet data using spyware delivered through typosquats of popular cryptographic libraries.
Security News
Socket's package search now displays weekly downloads for npm packages, helping developers quickly assess popularity and make more informed decisions.