Custom validation in the clean method triggers before checking the fields exist or not.

[Tahmid]

I have the following code in my models file in Django:

class MyModel(models.Model):
        ...
    foo = models.IntegerField()
    bar = models.IntegerField()
    
    def validate_foo_bar(self):
        self._validation_errors = {}
        if self.foo > self.bar:
            self._validation_errors['foo'] = ['Must be greater than bar.']
            self._validation_errors['bar'] = ['Must be less than foo.']
            
    def clean(self):
        self.validate_foo_bar()
        if bool(self._validation_errors):
    		raise ValidationError(self._validation_errors)
    	super(MyModel, self).clean()

Hopefully the idea is clear. I check for errors in the clean method and raise them if they occur. When I use an admin form to create an object, if I leave the foo and bar fields empty, I get the following error:

if self.foo > self.bar:
TypeError: '>' not supported between instances of 'NoneType' and 'NoneType'

Why is this happening? Shouldn't the requirement check trigger before the method I wrote? Thanks for any help.

[Mariusz Felisiak]

Please don't use Trac as a support channel. If you accept None values you need to handle them in comparisons, see also ​documentation.

Closing per TicketClosingReasons/UseSupportChannels.