Django允许空字段模型对象保存，如何禁止Dataset实例空字段创建？

Great question! Let's break down why this is happening and how to fix it properly.

Why Your Current Setup Allows Empty Objects #

First, let's clarify the critical difference between two common Django field arguments—this is the root of your issue:

• null=False: This is a database-level constraint that blocks the field from storing a NULL value. But in SQLite (and most databases), an empty string '' isn't the same as NULL—so saving '' for name/desc and None for data_file (which gets stored as an empty string in the database) doesn't violate this rule.
• blank=False: This is a form-level constraint that tells Django forms to reject empty inputs. However, when you create and save a model instance directly in the shell (without using a form), Django doesn't enforce blank=False checks automatically.

That's why your empty Dataset instance gets saved—neither the database nor the model's default save logic is stopping it.

Fixes to Enforce Non-Empty Fields #

1. Override the Model's clean() Method (Most Reliable) #

The best way to lock in validation at the model level is to override the clean() method. This runs automatically before saving (if you use Django's built-in workflows) and lets you add custom checks that apply no matter how the instance is created:

from django.db import models
from django.core.exceptions import ValidationError

class Dataset(models.Model):
    name = models.CharField(max_length=256)
    desc = models.TextField()
    data_file = models.FileField(upload_to='datasets/')

    def clean(self):
        # Always call the parent class's clean method first
        super().clean()

        # Validate name isn't empty or just whitespace
        if not self.name.strip():
            raise ValidationError({'name': 'Name cannot be empty or just whitespace.'})
        
        # Validate description has actual content
        if not self.desc.strip():
            raise ValidationError({'desc': 'Description cannot be empty.'})
        
        # Ensure a file was uploaded
        if not self.data_file:
            raise ValidationError({'data_file': 'Please upload a data file.'})

Now, when you try to save an empty instance, Django will throw a ValidationError before writing to the database:

>>> d = Dataset()
>>> d.save()
# Raises ValidationError with your custom error messages

2. Use full_clean() for Manual Validation #

If you don't want to override clean(), you can explicitly call full_clean() on the instance before saving. This triggers all model-level validations (including field validators and the default clean() logic):

>>> d = Dataset()
>>> d.full_clean()  # Raises ValidationError right away
>>> d.save()  # This line won't run if full_clean() fails

Note: This requires remembering to call full_clean() every time you create an instance directly—so it's less reliable than overriding clean() for long-term use.

3. Add Field Validators #

You can also use Django's built-in validators to enforce minimum length for text fields:

from django.db import models
from django.core.validators import MinLengthValidator

class Dataset(models.Model):
    name = models.CharField(
        max_length=256,
        validators=[MinLengthValidator(1, message="Name cannot be empty.")]
    )
    desc = models.TextField(
        validators=[MinLengthValidator(1, message="Description cannot be empty.")]
    )
    data_file = models.FileField(upload_to='datasets/')

Just keep in mind: these validators only run when full_clean() is called (or when using a form), so combining this with the clean() method is the safest approach.

Key Takeaway #

blank=False only applies to forms—if you're creating instances directly (via shell, scripts, etc.), you need model-level validation to block empty fields. Overriding clean() is the most robust way to ensure your Dataset objects always have valid data, no matter how they're created.

内容的提问来源于stack exchange，提问作者Hemant Mehra