Filtering

where() adds a filter to the plan. Filters compile to WHERE clauses with bound parameters. Values are never interpolated into SQL.

Comparators

db.from(users).where("age", ">", 18);
db.from(users).where("age", ">=", 18);
db.from(users).where("age", "<", 65);
db.from(users).where("age", "<=", 65);
db.from(users).where("email", "=", "a@b.c");
db.from(users).where("email", "!=", "a@b.c");

Operator	Meaning
= !=	equality / inequality
> >= < <=	numeric / date comparison
like	pattern match (% wildcard)
in	value in a list
is is not	null checks
between	range (inclusive)

IN

db.from(users).where("id", "in", [1, 2, 3]);
// → WHERE "id" IN (?, ?, ?)

LIKE

db.from(users).where("email", "like", "ada%");

IS NULL

db.from(users).where("age", "is", null);
db.from(users).where("age", "is not", null);

BETWEEN

db.from(users).where("age", "between", [18, 65]);

OR groups

Pass a function to or() to group alternatives:

db.from(users).or((b) =>
  b.where("age", "<", 18).where("age", ">", 65),
);
// → WHERE ("age" < ? OR "age" > ?)

Subqueries (EXISTS / IN)

db.from(users).whereExists(
  db.from(posts).where("authorId", "=", /* column ref */ "users.id"),
);

FilterNode for RLS

You can pass a raw FilterNode to where() — this is how RLS policies inject filters:

import { applyPolicy } from "@mountsqli/auth";
const scoped = applyPolicy(db.from(files), allowOwner("ownerId"), { userId });

Best practices

• Chain where() to AND multiple conditions.
• Use in for membership instead of multiple or equals.
• Keep filters in the plan; don’t filter rows in app code.

Common mistakes

• Using = with null — use is null instead.
• String-interpolating a value into a raw expression — bind it as a param.

Related

• Select — running filtered queries.
• Raw SQL & RLS Gate — whereExpr.
• RLS Policy DSL — policies as FilterNode[].