odoo-constraints

name	Odoo Constraints
description	Ensure data consistency using SQL Constraints and Python Constraints.

Odoo Constraints

Goal

Prevent invalid data from being saved to the database.

1. SQL Constraints

Enforced by the PostgreSQL database. Efficient and atomic.

• Defined via _sql_constraints model attribute.
• List of tuples: ('constraint_name', 'SQL CHECK/UNIQUE', 'Error Message').

Example:

class EstateProperty(models.Model):
    _name = "estate.property"

    _sql_constraints = [
        ('check_expected_price', 'CHECK(expected_price > 0)', 'The expected price must be positive.'),
        ('check_selling_price', 'CHECK(selling_price >= 0)', 'The selling price must be positive.'),
        ('name_uniq', 'UNIQUE(name)', 'Property name must be unique!'),
    ]

• Note: If existing data violates the constraint, the module update/install will fail. You must fix the data first.

2. Python Constraints

Enforced by Odoo server application logic. Used for complex validations that SQL cannot handle easily.

• Decorator: @api.constrains('field1', 'field2')
• Exception: Raise odoo.exceptions.ValidationError if check fails.

Example:

from odoo.exceptions import ValidationError
from odoo.tools.float_utils import float_compare

@api.constrains('selling_price', 'expected_price')
def _check_selling_price(self):
    for record in self:
        # Skip check if no selling price is set
        if float_is_zero(record.selling_price, precision_digits=2):
            continue

        # Check if selling price is lower than 90% of expected
        if float_compare(record.selling_price, record.expected_price * 0.9, precision_digits=2) == -1:
            raise ValidationError("Selling price cannot be lower than 90% of the expected price.")

• Float Comparison: ALWAYS use float_compare and float_is_zero when working with monetary fields to avoid rounding errors.
• Trigger: The method runs whenever any of the fields in @api.constrains are modified.

3. SQL vs Python

• SQL: Faster, strictly enforced at DB level. Use for simple checks (ranges, uniqueness).
• Python: Slower, enforced at app level. Use for complex logic involving multiple records or python-specific calculations.