a1eb940 · typeable/postgresql-simple

git.iko.soy/typeable/postgresql-simple

Add Values constructor

Closes #61 and #91. See #61 in particular for a discussion of the design

commita1eb940108eb660b92bcd3792ccd66f685748329 authorLeon P Smith <leon@melding-monads.com> dateSat 22 Mar 2014, 17:11 (13 years ago) parentf0bfbb9

4 files changed+124−1

compare with master

unifiedsplit

src/Database/PostgreSQL/Simple/ToField.hs+57 src/Database/PostgreSQL/Simple/ToField.hs-boot+23 src/Database/PostgreSQL/Simple/ToRow.hs-boot+2−1 src/Database/PostgreSQL/Simple/Types.hs+42

src/Database/PostgreSQL/Simple/ToField.hs+57

@@ -33,6 +33,7 @@ import Data.Monoid (mappend)

3333import Data.Time (Day, TimeOfDay, LocalTime, UTCTime, ZonedTime)

3434import Data.Typeable (Typeable)

3535import Data.Word (Word, Word8, Word16, Word32, Word64)

36+import {-# SOURCE #-} Database.PostgreSQL.Simple.ToRow

3637import Database.PostgreSQL.Simple.Types

3738import qualified Blaze.ByteString.Builder.Char.Utf8 as Utf8

3839import qualified Data.ByteString as SB

@@ -279,3 +280,59 @@ toJSONField = toField . JSON.toJSON

279280inQuotes :: Builder -> Builder

280281inQuotes b = quote `mappend` b `mappend` quote

281282 where quote = Utf8.fromChar '\''

283+

284+interleaveFoldr :: (a -> [b] -> [b]) -> b -> [b] -> [a] -> [b]

285+interleaveFoldr f b bs as = foldr (\a bs -> b : f a bs) bs as

286+{-# INLINE interleaveFoldr #-}

287+

288+instance ToRow a => ToField (Values a) where

289+ toField (Values types rows) =

290+ case rows of

291+ [] -> case types of

292+ [] -> error err

293+ (_:_) -> values $ typedRow (repeat (lit "null"))

294+ types

295+ [lit " LIMIT 0)"]

296+ (_:_) -> case types of

297+ [] -> values $ untypedRows rows [litC ')']

298+ (_:_) -> values $ typedRows rows types [litC ')']

299+ where

300+ err = "Database.PostgreSQL.Simple.toField :: Values -> Action either values or types must be non-empty"

301+ lit = Plain . fromByteString

302+ litC = Plain . fromChar

303+ values x = Many (lit "(VALUES ": x)

304+

305+ typedField :: (Action, QualifiedIdentifier) -> [Action] -> [Action]

306+ typedField (val,typ) rest = val : lit "::" : toField typ : rest

307+

308+ typedRow :: [Action] -> [QualifiedIdentifier] -> [Action] -> [Action]

309+ typedRow (val:vals) (typ:typs) rest =

310+ litC '(' :

311+ typedField (val,typ) ( interleaveFoldr

312+ typedField

313+ (litC ',')

314+ (litC ')' : rest)

315+ (zip vals typs) )

316+

317+ untypedRow :: [Action] -> [Action] -> [Action]

318+ untypedRow (val:vals) rest =

319+ litC '(' : val :

320+ interleaveFoldr

321+ (:)

322+ (litC ',')

323+ (litC ')' : rest)

324+ vals

325+

326+ typedRows :: ToRow a => [a] -> [QualifiedIdentifier] -> [Action] -> [Action]

327+ typedRows (val:vals) types rest =

328+ typedRow (toRow val) types (litC ',' : untypedRows vals rest)

329+

330+ untypedRows :: ToRow a => [a] -> [Action] -> [Action]

331+ untypedRows [] rest = rest

332+ untypedRows (val:vals) rest =

333+ untypedRow (toRow val) $

334+ interleaveFoldr

335+ (untypedRow . toRow)

336+ (litC ',')

337+ rest

338+ vals

src/Database/PostgreSQL/Simple/ToField.hs-boot+23

@@ -1,6 +1,29 @@

11module Database.PostgreSQL.Simple.ToField where

33import Database.PostgreSQL.Simple.Types

4+import Blaze.ByteString.Builder(Builder)

5+import Data.ByteString(ByteString)

7+-- | How to render an element when substituting it into a query.

8+data Action =

9+ Plain Builder

10+ -- ^ Render without escaping or quoting. Use for non-text types

11+ -- such as numbers, when you are /certain/ that they will not

12+ -- introduce formatting vulnerabilities via use of characters such

13+ -- as spaces or \"@'@\".

14+ | Escape ByteString

15+ -- ^ Escape and enclose in quotes before substituting. Use for all

16+ -- text-like types, and anything else that may contain unsafe

17+ -- characters when rendered.

18+ | EscapeByteA ByteString

19+ -- ^ Escape binary data for use as a @bytea@ literal. Include surrounding

20+ -- quotes. This is used by the 'Binary' newtype wrapper.

21+ | EscapeIdentifier ByteString

22+ -- ^ Escape before substituting. Use for all sql identifiers like

23+ -- table, column names, etc. This is used by the 'Identifier' newtype

24+ -- wrapper.

25+ | Many [Action]

26+ -- ^ Concatenate a series of rendering actions.

427

528class ToField a

629

src/Database/PostgreSQL/Simple/ToRow.hs-boot+2−1

@@ -3,6 +3,7 @@ module Database.PostgreSQL.Simple.ToRow where

33import Database.PostgreSQL.Simple.Types

44import {-# SOURCE #-} Database.PostgreSQL.Simple.ToField

6−class ToRow a

6+class ToRow a where

7+ toRow :: a -> [Action]

89instance ToField a => ToRow (Only a)

src/Database/PostgreSQL/Simple/Types.hs+42

@@ -27,6 +27,7 @@ module Database.PostgreSQL.Simple.Types

2727 , (:.)(..)

2828 , Savepoint(..)

2929 , PGArray(..)

30+ , Values(..)

3031 ) where

3132

3233import Blaze.ByteString.Builder (toByteString)

@@ -173,3 +174,44 @@ infixr 3 :.

173174

174175newtype Savepoint = Savepoint Query

175176 deriving (Eq, Ord, Show, Read, Typeable)

177+

178+-- | Represents a @VALUES@ table literal, usable as an alternative

179+-- to @executeMany@ and @returning@. For example:

180+--

181+-- > execute c "INSERT INTO table (key,val) ?"

182+-- > (Only (Values ["int4","text"]

183+-- > [(1,"hello"),(2,"world")]))

184+--

185+-- Issues the following query:

186+--

187+-- > INSERT INTO table (key,val) (VALUES (1::"int4",'hello'::"text"),(2,'world'))

188+--

189+-- When the list of values is empty, the following query will be issued:

190+--

191+-- > INSERT INTO table (key,val) (VALUES (null::"int4",null::"text") LIMIT 0)

192+--

193+-- By contrast, @executeMany@ and @returning@ don't issue the query

194+-- in the empty case, and simply return @0@ and @[]@ respectively.

195+--

196+-- The advantage over @executeMany@ is in cases when you want to

197+-- parameterize table literals in addition to other parameters, as can

198+-- occur with writable common table expressions, for example.

199+--

200+-- The first argument is a list of postgresql type names. Because this

201+-- is turned into a properly quoted identifier, the type name is case

202+-- sensitive and must be as it appears in the @pg_type@ table. Thus,

203+-- you must write @timestamptz@ instead of @timestamp with time zone@,

204+-- @int4@ instead of @integer@, @_int8@ instead of @bigint[]@, etcetera.

205+--

206+-- You may omit the type names, however, if you do so the list

207+-- of values must be non-empty, and postgresql must be able to infer

208+-- the types of the columns from the surrounding context. If these

209+-- conditions are not met, postgresql-simple will throw an exception

210+-- without issuing the query in the former case, and in the latter

211+-- the postgres server will return an error which will be turned into

212+-- a @SqlError@ exception.

213+--

214+-- See <http://www.postgresql.org/docs/9.3/static/sql-values.html> for

215+-- more information.

216+data Values a = Values [QualifiedIdentifier] [a]

217+ deriving (Eq, Ord, Show, Read, Typeable)