# print method doesn't change unexpectedly

    Code
      left_join(lf1, lf2, by = "x") %>% left_join(lf3, by = "x") %>% sql_build()
    Output
      <SQL JOINS>
      X:
        <dbplyr_table_ident[1]>
        [1] `lf1`
      Type: left
      By:
        x-x
      Y:
        <dbplyr_table_ident[1]>
        [1] `lf2`
      Type: left
      By:
        x-x
      Y:
        <dbplyr_table_ident[1]>
        [1] `lf3`

# generated sql doesn't change unexpectedly

    Code
      inner_join(lf, lf)
    Message
      Joining with `by = join_by(x, y)`
    Output
      <SQL>
      SELECT `df_LHS`.*
      FROM `df` AS `df_LHS`
      INNER JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`x` = `df_RHS`.`x` AND `df_LHS`.`y` = `df_RHS`.`y`)

---

    Code
      left_join(lf, lf)
    Message
      Joining with `by = join_by(x, y)`
    Output
      <SQL>
      SELECT `df_LHS`.*
      FROM `df` AS `df_LHS`
      LEFT JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`x` = `df_RHS`.`x` AND `df_LHS`.`y` = `df_RHS`.`y`)

---

    Code
      right_join(lf, lf)
    Message
      Joining with `by = join_by(x, y)`
    Output
      <SQL>
      SELECT `df_RHS`.*
      FROM `df` AS `df_LHS`
      RIGHT JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`x` = `df_RHS`.`x` AND `df_LHS`.`y` = `df_RHS`.`y`)

---

    Code
      full_join(lf, lf)
    Message
      Joining with `by = join_by(x, y)`
    Output
      <SQL>
      SELECT
        COALESCE(`df_LHS`.`x`, `df_RHS`.`x`) AS `x`,
        COALESCE(`df_LHS`.`y`, `df_RHS`.`y`) AS `y`
      FROM `df` AS `df_LHS`
      FULL JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`x` = `df_RHS`.`x` AND `df_LHS`.`y` = `df_RHS`.`y`)

# only disambiguates shared variables

    Code
      left_join(lf1, lf2)
    Message
      Joining with `by = join_by(x)`
    Output
      <SQL>
      SELECT `df_LHS`.*, `z`
      FROM `df` AS `df_LHS`
      LEFT JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`x` = `df_RHS`.`x`)

---

    Code
      left_join(lf1, lf2, by = c(y = "z"))
    Output
      <SQL>
      SELECT `df_LHS`.`x` AS `x.x`, `y`, `df_RHS`.`x` AS `x.y`
      FROM `df` AS `df_LHS`
      LEFT JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`y` = `df_RHS`.`z`)

# disambiguate variables that only differ in case

    Code
      left_join(lf1, lf2, by = "y")
    Output
      <SQL>
      SELECT `df_LHS`.*, `df_RHS`.`X` AS `X`
      FROM `df` AS `df_LHS`
      LEFT JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`y` = `df_RHS`.`y`)

---

    Code
      full_join(lf1, lf2, by = "y")
    Output
      <SQL>
      SELECT
        `df_LHS`.`x` AS `x`,
        COALESCE(`df_LHS`.`y`, `df_RHS`.`y`) AS `y`,
        `df_RHS`.`X` AS `X`
      FROM `df` AS `df_LHS`
      FULL JOIN `df` AS `df_RHS`
        ON (`df_LHS`.`y` = `df_RHS`.`y`)

# sql_on query doesn't change unexpectedly

    Code
      inner_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.`x` AS `x.x`, `y`, `RHS`.`x` AS `x.y`, `z`
      FROM `df` AS `LHS`
      INNER JOIN `df` AS `RHS`
        ON (LHS.y < RHS.z)

---

    Code
      left_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.`x` AS `x.x`, `y`, `RHS`.`x` AS `x.y`, `z`
      FROM `df` AS `LHS`
      LEFT JOIN `df` AS `RHS`
        ON (LHS.y < RHS.z)

---

    Code
      right_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.`x` AS `x.x`, `y`, `RHS`.`x` AS `x.y`, `z`
      FROM `df` AS `LHS`
      RIGHT JOIN `df` AS `RHS`
        ON (LHS.y < RHS.z)

---

    Code
      full_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.`x` AS `x.x`, `y`, `RHS`.`x` AS `x.y`, `z`
      FROM `df` AS `LHS`
      FULL JOIN `df` AS `RHS`
        ON (LHS.y < RHS.z)

---

    Code
      semi_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.*
      FROM `df` AS `LHS`
      WHERE EXISTS (
        SELECT 1 FROM `df` AS `RHS`
        WHERE (LHS.y < RHS.z)
      )

---

    Code
      anti_join(lf1, lf2, sql_on = "LHS.y < RHS.z")
    Output
      <SQL>
      SELECT `LHS`.*
      FROM `df` AS `LHS`
      WHERE NOT EXISTS (
        SELECT 1 FROM `df` AS `RHS`
        WHERE (LHS.y < RHS.z)
      )