I am having 2 tables with PK and FK reference as below.

CREATE TABLE IF NOT EXISTS User
(
    userid VARCHAR(255) PRIMARY KEY
);


CREATE TABLE IF NOT EXISTS UserType
(
    email       VARCHAR(255) PRIMARY KEY,
    userId      VARCHAR(255),
    type        VARCHAR(255),
    FOREIGN KEY (userId) REFERENCES User (userId)
);

Table: User

Table: UserType

I am generating userid from code and inserting using sqlx – sqlite as below

async fn create_user(
        &self,
        user: &User,
    ) -> Result<(), CreateError> {
        let query = "insert into User (id) values (?1)";
        let query_local = "insert or ignore into UserType (email,user_id) values (?1,?2)";
        let mut conn = self
            .pool
            .acquire()
            .map_err("xxx")
            .await?;

        let mut tx = conn.begin().map_err("xxx").await?;

        sqlx::query(query)
            .bind(user.userid.as_str())
            .execute(&mut *tx)
            .map_err("xxx")
            .await?;
        sqlx::query(query_local)
            .bind(user.email.as_str())
            .bind(user.userid.as_str())
            .execute(&mut *tx)
            .map_err("xxx")
            .await?;
        tx.commit().map_err("xxx").await?;
        Ok(())
    }

Now in case of getting duplicate email, on conflict while inserting ‘UserType’ (second) table, How can i avoid the UserId inserted in ‘User’ (first) table? I am new to relational DB and SQL. Is there idiomatic way to do this through query or through transaction? Please help

Note: I know this design is vague but its required for processing further for application reasons