Opaleye.Order

Ordering, LIMIT, OFFSET and DISTINCT ON

Order by

orderBy :: Order a -> Select a -> Select a

Order the rows of a Select according to the Order.

import Data.Monoid ((<>))

-- Order by the first field ascending.  When first fields are equal
-- order by second field descending.
example :: Select (Field SqlInt4, Field SqlText)
        -> Select (Field SqlInt4, Field SqlText)
example = orderBy (asc fst <> desc snd)

data Order a

An Order a represents a sort order and direction for the elements of the type a. Multiple Orders can be composed with mappend or (<>) from Data.Monoid. If two rows are equal according to the first Order in the mappend, the second is used, and so on.

Instances

Contravariant Order

Defined in Opaleye.Internal.Order

Methods

contramap :: (a -> b) -> Order b -> Order a

(>$) :: b -> Order b -> Order a

Divisible Order

Defined in Opaleye.Internal.Order

Methods

divide :: (a -> (b, c)) -> Order b -> Order c -> Order a

conquer :: Order a

Decidable Order

Defined in Opaleye.Internal.Order

Methods

lose :: (a -> Void) -> Order a

choose :: (a -> Either b c) -> Order b -> Order c -> Order a

Semigroup (Order a)

Defined in Opaleye.Internal.Order

Methods

(<>) :: Order a -> Order a -> Order a

sconcat :: NonEmpty (Order a) -> Order a

stimes :: Integral b => b -> Order a -> Order a

Monoid (Order a)

Defined in Opaleye.Internal.Order

Methods

mempty :: Order a

mappend :: Order a -> Order a -> Order a

mconcat :: [Order a] -> Order a

Order direction

asc :: SqlOrd b => (a -> Column b) -> Order a

Specify an ascending ordering by the given expression. (Any NULLs appear last)

desc :: SqlOrd b => (a -> Column b) -> Order a

Specify an descending ordering by the given expression. (Any NULLs appear first)

ascNullsFirst :: SqlOrd b => (a -> Column b) -> Order a

Specify an ascending ordering by the given expression. (Any NULLs appear first)

descNullsLast :: SqlOrd b => (a -> Column b) -> Order a

Specify an descending ordering by the given expression. (Any NULLs appear last)

Limit and offset

limit :: Int -> Select a -> Select a

Limit the results of the given query to the given maximum number of items.

WARNING: If you're planning on using limit/offset together please use offset before you use limit, e.g.:

limit 10 (offset 50 yourQuery)

This is because Opaleye applies OFFSET and LIMIT to the query separately. The result of the query given above is the following, which will return 10 rows after skipping the first 50 (probably what you want).

SELECT * FROM (SELECT * FROM yourTable OFFSET 50) LIMIT 10

However, reversing the order of the limit/offset will result in the following, which will result in no rows being returned (probably not what you want).

SELECT * FROM (SELECT * FROM yourTable LIMIT 10) OFFSET 50

offset :: Int -> Select a -> Select a

Offset the results of the given query by the given amount, skipping that many result rows.

WARNING: Please read the documentation of limit before combining offset with limit.

Distinct on

distinctOn :: Default Unpackspec b b => (a -> b) -> Select a -> Select a

Keep a row from each set where the given function returns the same result. No ordering is guaranteed. Mutliple fields may be distinguished by projecting out tuples of Field_s. Use distinctOnBy to control how the rows are chosen.

distinctOnBy :: Default Unpackspec b b => (a -> b) -> Order a -> Select a -> Select a

Keep the row from each set where the given function returns the same result. The row is chosen according to which comes first by the supplied ordering. However, no output ordering is guaranteed. Mutliple fields may be distinguished by projecting out tuples of Field_s.

Exact ordering

exact :: [Column b] -> (a -> Column b) -> Order a

Order the results of a given query exactly, as determined by the given list of input columns. Note that this list does not have to contain an entry for every result in your query: you may exactly order only a subset of results, if you wish. Rows that are not ordered according to the input list are returned after the ordered results, in the usual order the database would return them (e.g. sorted by primary key). Exactly-ordered results always come first in a result set. Entries in the input list that are not present in result of a query are ignored.

Other

class PGOrd a

Typeclass for Postgres types which support ordering operations.

Instances

PGOrd SqlCitext

Defined in Opaleye.Order

PGOrd SqlUuid

Defined in Opaleye.Order

PGOrd SqlTimestamptz

Defined in Opaleye.Order

PGOrd SqlTimestamp

Defined in Opaleye.Order

PGOrd SqlTime

Defined in Opaleye.Order

PGOrd SqlText

Defined in Opaleye.Order

PGOrd SqlNumeric

Defined in Opaleye.Order

PGOrd SqlInt2

Defined in Opaleye.Order

PGOrd SqlInt4

Defined in Opaleye.Order

PGOrd SqlInt8

Defined in Opaleye.Order

PGOrd SqlFloat8

Defined in Opaleye.Order

PGOrd SqlFloat4

Defined in Opaleye.Order

PGOrd SqlDate

Defined in Opaleye.Order

PGOrd SqlBool

Defined in Opaleye.Order

PGOrd a => PGOrd (Nullable a)

Defined in Opaleye.Order

type SqlOrd = PGOrd