Rules

Every rule has a code, a name and a category. Rules in the correctness category report queries the database rejects (or that fail at run time) and are errors by default; rules in the suspicious category report valid queries that almost certainly don't do what was meant, and are warnings by default. Rules in the restriction category ban constructs some codebases don't want and are off until you enable them. See Rules and levels to change their levels and Suppressing diagnostics for exceptions.

The E in a rule code is sqlsift's prefix, not a severity: a rule's level comes from its category and your configuration, so a warning can carry an E code too. Codes are assigned in order as rules are added and never change or get reused, so they are safe to keep in config files, suppression comments and baselines. Names are easier to read in config and comments; codes are shorter.

The examples on these pages use this schema:

CREATE TYPE order_status AS ENUM ('open', 'paid', 'shipped');
CREATE TABLE users (
  id SERIAL PRIMARY KEY,
  name TEXT NOT NULL,
  email TEXT NOT NULL UNIQUE,
  created_at TIMESTAMP NOT NULL DEFAULT now()
);
CREATE TABLE orders (
  id SERIAL PRIMARY KEY,
  user_id INTEGER NOT NULL REFERENCES users(id),
  status order_status NOT NULL DEFAULT 'open',
  total NUMERIC(10, 2) NOT NULL
);
CREATE TABLE coupons (
  code TEXT PRIMARY KEY,
  order_id INTEGER REFERENCES orders(id)
);