django-qs2csv

A simple package to convert a Django QuerySet to a CSV file through an HttpResponse object.

Getting started

Prerequisites

• Python >= 3.9
• Django >= 3.2

Optional Dependencies

• pandas >= 1.5

Installation

Recommended

pip install django-qs2csv

Full

pip install django-qs2csv[pd]

Note: this will install pandas, which is used with queryset_to_csv_pd(). It is not recommended to install the pandas library unless you will use this function or already use pandas. Using queryset_to_csv_pd() will likely increase the export speed of large QuerySets (2M+ results) and decrease file size for all QuerySets.

Usage

views.py

from qs2csv import qs_to_csv

from .models import SampleModel

...

def export_csv(request):
    ...

    my_queryset = SampleModel.objects.all()

    response = qs_to_csv(
        my_queryset,
        filename="all_sample_models",
    )

    return response

Return type

All functions return a django.http.HttpResponse with the Content-Type and Content-Disposition headers. Additional headers can be added to the response before returning:

...

response = qs_to_csv(my_queryset)
response["Another-Header"] = "This is another header for the HttpResponse."

...

Parameters

Universal

qs : QuerySet - Required. The QuerySet to be exported as a CSV file. This can be passed as QuerySet[object], QuerySet[dict] (values()), or QuerySet[list[tuple]] (values_list()). See the note in the Limitations about QuerySet evaluation.

filename : str - The name of the exported CSV file. You do not need to include .csv, it will be added once the filename is evaluated. File names can not end in a period, include the symbols (< > : " / \ | ? *), or be longer than 251 characters (255 w/ ".csv"). Default: "export"

only : list[str] - List the field names that you would like to include in the exported file. An empty list will include all fields, other than those in defer. Field names listed in both only and defer will not be included. See the note in the Limitations section for details how this works with a QuerySet that calls only(), defer(), values(), or values_list(). Default: []

defer : list[str] - List the field names that you do not want to include in the exported file. An empty list will include all fields, or just those mentioned in only. Field names listed in both only and defer will not be included. See the note in the Limitations section for details how this works with a QuerySet that calls only(), defer(), values(), or values_list(). Default: []

header : bool - Include a header row with field names. Default: False

verbose : bool - Determine if the header row uses the fields' verbose_name or just the field names. This only applies when header=True. Default: True

qs_to_csv() and qs_to_csv_pd()

values : bool - Only enable this if your QuerySet was already evaluated (no longer lazy) and called values(). You must ensure your fields are properly selected in the original QuerySet, because this will skip applying the only and defer parameters. Default: False

qs_to_csv_rel_str()

values : bool - Only enable this if the QuerySet is passed to the function after calling values() or values_list(). This will convert the QuerySet back to a list of model objects, instead of a list of dicts/lists. See note in Limitations for an IMPORTANT WARNING about performance. Default: False

Limitations

If the QuerySet was already evaluated before being passed to qs_to_csv then it will be re-evaluated by the function. Depending on the size of the QuerySet, complexity of the query and the database setup, this may add a noticeable delay. It is recommended to monitor the impact of database queries using django.db.connection.queries or django-debug-toolbar during development. If the QuerySet must be evaluated before the function is called, it would be most efficient to use values() with the QuerySet (if possible) then pass values=True to qs_to_csv.

If your QuerySet uses only() / defer() then you must include those same fields in the only / defer parameters when calling qs_to_csv. The function transforms all QuerySets into a list of dicts using values(), which is incompatible with only() and defer().

ForeignKey and OneToOneField will always return the primary key, because the function uses values().

ManyToManyField is not supported.

Passing values=True to qs_to_csv_rel_str() will create a new query, checking for primary keys (PKs) that are in a list of all PKs from your original QuerySet. This will add significant time if your QuerySet is large and will potentially not work, depending on the size of your QuerySet and your database's capabilities. It is recommended to avoid this by not using values() or values_list() when calling this function. It would be more efficient to create a brand new QuerySet than to do use this. Note: if you make this change, ensure values is False or the issue will remain.

License

This project is licensed under the MIT License - see the LICENSE.md file for details.