Consider discussing about Validation Data strategies

[ddahan]

Location within the Book

• Chapter or Appendix: 11.1: Validate All Incoming Data With Django Forms

Description

While the book already explains well how to validate models with ModelForm class in this section, some may prefer using validation at a model level (with validators in field declaration and/or overriding clean/clean_field methods in model).

To my view, setting validation at this lower level can decrease the risk of inserting bad data in database, especially when a project has many sources of input data, because you could forget to call the validation process.

In the case of you choose model-level validation, it could be well to warn people about the strange Django behaviour in which a call to the save method does not trigger the model validation, because the full_clean method is actually not called, for backward-compatibility purposes. I personally find this behaviour very counter-intuitive.

For this purpose, the mixin bellow is often used:

class ValidateOnSaveMixin(object):

    def save(self, force_insert=False, force_update=False, **kwargs):
        if not (force_insert or force_update):
            self.full_clean()
        super(ValidateOnSaveMixin, self).save(force_insert, force_update, **kwargs)

The problem and different solutions are well explained here: https://www.xormedia.com/django-model-validation-on-save/