Skip to main content

set_by

set_{field}[_and_{field2}...]_by_{filter} — an UPDATE that sets one or more fixed columns on every row matching a filter, without needing a full values struct:

async fn set_email_verified_at_by_id(&self, email_verified_at: DateTime<Utc>, id: i32) -> sqlx::Result<()>;
// UPDATE "user" SET email_verified_at = $1 WHERE id = $2

Multiple SET columns are chained with _and_ before _by_:

async fn set_verified_and_verified_at_by_id(
    &self, verified: bool, verified_at: DateTime<Utc>, id: i32,
) -> sqlx::Result<()>;
// UPDATE "user" SET verified = $1, verified_at = $2 WHERE id = $3

Parameter order​

Parameters are positional, not name-matched: the first N parameters (N = number of SET columns) are the new values, in the order the SET columns appear in the method name; the rest are the filter's bound values, in filter order. Arity is checked at compile time against (SET column count) + (filter parameter count).

What this is for vs. update​

set_by_* updates a fixed set of columns to values you pass directly — no values struct, no #[derive(SqlxUpdate)], and it can touch many rows (it's filter-based, not keyed to one row). Use update instead when you're updating a whole row by its primary key from a values struct.

Constraints​

  • _in in the filter works as usual, switching that clause to the QueryBuilder path.