Skip to content

Result Rows

Statement rows are represented in rdbc as Row trait. This chapter describes this trait and methods it provides that give access column values.

Generic methods

Row trait declares a family of col methods that allow accessing column values of any Scala type. The methods accept a single type parameter telling rdbc what Scala type should be used to represent a database value — see the list below for their simplified declarations.

  • def col[A](idx: Int): A
  • def colOpt[A](idx: Int): Option[A]
  • def col[A](name: String): A
  • def colOpt[A](name: String): Option[A]

The methods differ from one another in two areas:

  • Named vs unnamed columns

    To fetch column value by name use method that accepts a String; to fetch value by column's index, use the version that accepts an Int. Index is 0 based.

  • Null-safety

    colOpt methods are null-safe, and col methods aren't. You can't use plain col method to get SQL NULL values. If you try that, ConversionException will be thrown. colOpt returns Option so it's fit for handling NULLs — it represents them by None. col method is intended to be used only for columns that can't hold NULL values, for example because there is a not-null constraint defined for them.

Examples below show how to use col methods on rows produced by 

select first_name, last_name, age from persons

statement.

/* fetches first_name. If value is NULL ConversionException will be thrown */
row.col[String](0) 

/* fetches last_name. If value is NULL ConversionException will be thrown */
row.col[String]("last_name")

/* fetches age. If value is NULL, None will be returned */
row.colOpt[Int](2)

Type-specific methods

Row trait, for convenience, also provides methods that aren't parametrized by type and their names reflect the type they return. They are simply shortcuts for calling generic col methods described earlier. For instance, there is a str method that is a shortcut for calling col[String] method. See the Row Scaladoc for a complete list.

Previous example could be rewritten as follows:

/* fetches first_name. If value is NULL ConversionException will be thrown */
row.str(0) 

/* fetches last_name. If value is NULL ConversionException will be thrown */
row.str("last_name")

/* fetches age. If value is NULL, None will be returned */
row.intOpt(2)

If you want to use types supported by the particular driver but not supported by default by rdbc, you must always use generic col methods.