Django CSV Downloads

Django app for tracking queryset-backed CSV downloads

Version support

The current version of the this app support Python 3.12+ and Django 5.2-6.0

What does this app do?

This app is used to track user downloads of CSVs that are derived from Django QuerySets. You provide the filename, queryset and the list of columns that you want to output.

It has a single model (CsvDownload) that tracks downloads and stores the user, filename, row count and timestamp.

Usage

The recommended way to use this app is to rely on django_csv.views.download_csv, which wraps up the creation of the download object and the generation of the CSV itself:

# DISPLAY PURPOSES ONLY: DO NOT ENABLE USER DATA DOWNLOADS IN PRODUCTION
def download_users(request: HttpRequest) -> HttpResponse:
    data = User.objects.all()
    columns = ("first_name", "last_name", "email")
    return download_csv(request.user, "users.csv", data, *columns)

Settings

There is a CSV_DOWNLOAD_MAX_ROWS setting that is used to truncate output. Defaults to 10000. This is a backstop, and can be overridden on a per use basis.

Examples

Caution: All of these examples invåolve the User model as it's ubiquitous - DO NOT DO THIS ON A PRODUCTION ENVIRONMENT.

Example of writing a QuerySet to a file:

>>> data = User.objects.all()
>>> columns = ("first_name", "last_name", "email")
>>> with open('users.csv', 'w') as csvfile:
>>>     csv.write_csv(csvfile, data, *columns)
10  #<--- row count

adding a custom header:

>>> data = User.objects.all()
>>> columns = ("first_name", "last_name", "email")
>>> column_headers = ("given_name", "family_name", "email_address")
>>> with open('users.csv', 'w') as csvfile:
>>>     csv.write_csv(csvfile, data, *columns, column_headers=column_headers)
10

Example of writing to an HttpResponse:

>>> response = HttpResponse(content_type="text/csv")
>>> response["Content-Disposition"] = 'attachment; filename="users.csv"'
>>> csv.write_csv(response, data, *columns)
10

Example of writing to an in-memory text buffer:

>>> buffer = io.StringIO()
>>> csv.write_csv(buffer, data, *columns)
10

Example of writing directly to S3:

>>> with s3.s3_upload("bucket_name", "object_key") as buffer:
...     csv.write_csv(fileobj, queryset, *columns)
10
>>> # one-line convenience function
>>> s3.write_csv_s3("bucket_name/object_key", queryset, *columns)
10

Example of writing directly to SFTP:

# requires a paramiko.SFTPClient to have been created / connected.
>>> with sft.sftp_upload(client, remote_filepath) as fileobj:
...     write_csv(fileobj, queryset, *columns)
10
>>> # one-line convenience function
>>> sftp.write_csv_sftp("sftp://user:pass@host:port/path", queryset, *columns)
10

Example of a custom admin action to download User data:

class CustomUserAdmin(UserAdmin):

    actions = ['download']
    csv_fields = ("first_name", "last_name", "email", "is_staff")
    csv_filename = "users.csv"

    def download(self, request, queryset):
        """Download selected users as a CSV."""
        return download_csv(
            user=request.user,
            filename=CustomUserAdmin.csv_filename,
            queryset=queryset,
            *CustomUserAdmin.csv_fields
        )

    download.short_description = "Download selected users"

Example CBV that restricts queryset based on request.user:

class DownloadUsers(CsvDownloadView):

    def has_permission(self, request: HttpRequest) -> bool:
        return request.user.is_authenticated

    def get_queryset(self, request: HttpRequest) -> QuerySetWriter:
        """Allow superusers to download Users."""
        if request.user.is_superuser:
            return User.objects.all().order_by("first_name", "last_name")
        return User.objects.none()

    def get_filename(self, request: HttpRequest) -> str:
        return "users.csv"

    def get_columns(self, request: HttpRequest) -> str:
        return ("first_name", "last_name")