Skip to main content

Class-based view and mixins for handling CSV with Django.

Project description

Building Status

Class-based view and mixins for responding with CSV in Django. django-separated supports Django 1.3+.


$ pip install django-separated


django-separated provides many tools for generating CSV files based on querysets.



You can use the ColumnSerializer to generate CSV. It accepts a column definition, and returns a callable that you can use to serialize to CSV.

from separated.utils import ColumnSerializer

serialize_books = ColumnSerializer([
    ('title', 'Title'),
    ('pub_date', 'Publication Date'),
    ('isbn', 'ISBN'),
with open('/tmp/books.csv', 'wb') as f:
    books = Book.objects.all()
    serialize_books(books, file=f)

The column definition is an iterable of 2-tuples where the first item is an accessor to get the value off of an object and the second item is the column header.

The accessor can be a string or a callable. If it isn’t a callable, it will be passed into attrgetter to turn into a callable. If the accessor returns a callable, it will be called. All of the following are valid examples of accessors:

  • 'first_name'
  • 'first_name.upper'
  • 'get_absolute_url'
  • lambda x: x.upvotes.count() - x.downvotes.count()

You can even stretch across relationships:

  • 'author'
  • ''
  • 'author.book_count'
  • 'author.user.username'

The header value is optional, if you want a header to be generated from the accessor, you can write a simpler columns definition:

serialize_books = ColumnSerializer([
    'title',  # Header will be 'Title'
    'pub_date',  # Header will be 'Pub date'

You can mix and match the two styles as well.

By default, ColumnSerializer will output the headers as the first line. If you want to suppress this behavior, set output_headers to False.



A ListView that returns will present the user with a CSV file download. It requires a column definition that will be passed to the ColumnSerializer

class UserCsvView(CsvView):
    model = User
    columns = [
        ('first_name', 'First name'),
        ('last_name', 'Last name'),
        ('email', 'Email'),

There is a corresponding get_columns method if you need to have more dynamic behavior.

Additionally, you can specify the filename of the CSV file that will be downloaded. It will default to the model name + _list.csv if you don’t provide one. For example:

class UserCsvView(CsvView):
    model = User

will have a filename of user_list.csv. But you can override it by settings the filename attribute. There is a corresponding get_filename that you can override for more complicated behavior.

CsvView will forward the value of output_headers to the ColumnSerializer. To turn off the headers, you can do this:

class UserCsvView(CsvView):
    model = False
    output_headers = False


A MultipleObjectMixin subclass that returns a CsvResponse.

This is useful in instances where you want to substitute BaseListView for a ListView of your own. CsvResponseMixin supports all the behavior mentioned in CsvView, the only machinery you need to hook it up is a View class that calls render_to_response with a context that has a queryset available in the object_list key.

class MyWeirdBaseListView(View):
    def get(self, request, *args, **kwargs):
        return self.render_to_response({
            'object_list': User.objects.all(),

class MyWeirdCsvView(CsvResponseMixin, MyWeirdBaseListView):


A subclass of HttpResponse that will download as CSV. CsvResponse requires a filename as the first argument of the constructor.


You can use django-separated in the admin center to export CSV from the admin site.

from separated.admin import CsvExportModelAdmin

class NewsAdmin(CsvExportModelAdmin):
    csv_export_columns = [

This adds an action to the change list.

csv_export_columns corresponds to the CsvView.columns attribute. If you want more fine-grained control, you can override csv_export_view_class instead:

from datetime import datetime

from separated.admin import CsvExportModelAdmin
from separated.views import CsvView

class NewsCsvView(CsvView):
    columns = [
    output_headers = False

    def get_filename(self, model):
        return '%s-news-export.csv' %'Y-m-d')

class NewsAdmin(CsvExportModelAdmin):
    csv_export_view_class = NewsCsvView

csv_export_columns and csv_export_view_class also exist as methods (get_csv_export_columns and get_csv_export_view_class respectively) if you need change them based on request.

from separated.admin import CsvExportModelAdmin

class NewsAdmin(CsvExportModelAdmin):
    staff_export_columns = (

    superuser_export_columns = staff_export_columns + (

    def get_csv_export_columns(self, request):
        if request.user.is_superuser:
            return self.superuser_export_columns
            return self.staff_export_columns


django-separated provides a couple of helpers for normalizing the data that comes off of the model before sending it to the CSV writer. These are all based on a Getter class which handles the different types of accessors.


If you have a boolean value that you wish to be transformed into Yes or No, you can use the BooleanGetter:

from separated.utils import BooleanGetter

user_serializer = ColumnSerializer([


If you have a model field that has choices and you want the human readable display to appear in the CSV, you can use the DisplayGetter:

from separated.utils import BooleanGetter

class User(models.Model):
    favorite_color = models.CharField(max_length=255,
            ('blue', 'Blue'),
            ('green', 'Green'),
            ('red', 'Red'),

user_serializer = ColumnSerializer([

This will end up using the get_favorite_color_display method that Django automatically adds.

Project details

Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Files for django-separated, version 1.1.0
Filename, size File type Python version Upload date Hashes
Filename, size django_separated-1.1.0-py2-none-any.whl (12.7 kB) File type Wheel Python version 2.7 Upload date Hashes View
Filename, size django-separated-1.1.0.tar.gz (11.6 kB) File type Source Python version None Upload date Hashes View

Supported by

Pingdom Pingdom Monitoring Google Google Object Storage and Download Analytics Sentry Sentry Error logging AWS AWS Cloud computing DataDog DataDog Monitoring Fastly Fastly CDN DigiCert DigiCert EV certificate StatusPage StatusPage Status page