diff --git a/Language/SQL/SimpleSQL/Parser.lhs b/Language/SQL/SimpleSQL/Parser.lhs
index 6d255cd..7f34703 100644
--- a/Language/SQL/SimpleSQL/Parser.lhs
+++ b/Language/SQL/SimpleSQL/Parser.lhs
@@ -420,7 +420,7 @@ TODO: this code needs heavy refactoring
> typeName :: Parser TypeName
> typeName =
-> (rowTypeName <|> intervalTypeName <|> ref <|> otherTypeName)
+> (rowTypeName <|> intervalTypeName <|> otherTypeName)
> >>= tnSuffix
> <?> "typename"
> where
@@ -432,11 +432,6 @@ TODO: this code needs heavy refactoring
> intervalTypeName =
> keyword_ "interval" >>
> uncurry IntervalTypeName <$> intervalQualifier
-> ref =
-> keyword_ "ref" >>
-> RefTypeName
-> <$> parens (names)
-> <*> optionMaybe (keyword_ "scope" *> names)
> -- other type names, which includes:
> -- precision, scale, lob scale and units, timezone, character
> -- set and collations
@@ -478,6 +473,8 @@ TODO: this code needs heavy refactoring
> x <- choice [Just LobK <$ keyword_ "k"
> ,Just LobM <$ keyword_ "m"
> ,Just LobG <$ keyword_ "g"
+> ,Just LobT <$ keyword_ "t"
+> ,Just LobP <$ keyword_ "p"
> ,return Nothing]
> return (p,x)
> lobUnits = choice [LobCharacters <$ keyword_ "characters"
diff --git a/Language/SQL/SimpleSQL/Pretty.lhs b/Language/SQL/SimpleSQL/Pretty.lhs
index 52a7f22..1752d77 100644
--- a/Language/SQL/SimpleSQL/Pretty.lhs
+++ b/Language/SQL/SimpleSQL/Pretty.lhs
@@ -263,7 +263,9 @@ which have been changed to try to improve the layout of the output.
> <> me (\x -> case x of
> LobK -> text "K"
> LobM -> text "M"
-> LobG -> text "G") m
+> LobG -> text "G"
+> LobT -> text "T"
+> LobP -> text "P") m
> <+> me (\x -> case x of
> LobCharacters -> text "CHARACTERS"
> LobOctets -> text "OCTETS") u)
@@ -297,11 +299,6 @@ which have been changed to try to improve the layout of the output.
> typeName (MultisetTypeName tn) =
> typeName tn <+> text "multiset"
-> typeName (RefTypeName rt sc) =
-> text "ref"
-> <> parens (names rt)
-> <+> me (\x -> text "scope" <+> names x) sc
-
> intervalTypeField :: IntervalTypeField -> Doc
> intervalTypeField (Itf n p) =
> text n
diff --git a/Language/SQL/SimpleSQL/Syntax.lhs b/Language/SQL/SimpleSQL/Syntax.lhs
index 20c60d6..4dc7900 100644
--- a/Language/SQL/SimpleSQL/Syntax.lhs
+++ b/Language/SQL/SimpleSQL/Syntax.lhs
@@ -167,8 +167,6 @@
> | UQName String
> deriving (Eq,Show,Read,Data,Typeable)
-TODO: add ref and scope, any others?
-
> -- | Represents a type name, used in casts.
> data TypeName
> = TypeName [Name]
@@ -182,13 +180,12 @@ TODO: add ref and scope, any others?
> | IntervalTypeName IntervalTypeField (Maybe IntervalTypeField)
> | ArrayTypeName TypeName (Maybe Integer)
> | MultisetTypeName TypeName
-> | RefTypeName [Name] (Maybe [Name])
> deriving (Eq,Show,Read,Data,Typeable)
> data IntervalTypeField = Itf String (Maybe (Integer, Maybe Integer))
> deriving (Eq,Show,Read,Data,Typeable)
-> data LobMultiplier = LobK | LobM | LobG
+> data LobMultiplier = LobK | LobM | LobG | LobT | LobP
> deriving (Eq,Show,Read,Data,Typeable)
> data LobUnits = LobCharacters
> | LobOctets
diff --git a/TODO b/TODO
index 468201e..dd1fb8b 100644
--- a/TODO
+++ b/TODO
@@ -1,4 +1,4 @@
-continue 2003 review and tests
+continue 2011 review and tests
1. start replacing the 2003 stuff with 2011
2. create an error message document for the website
@@ -62,76 +62,58 @@ rules for changing the multi keyword parsing:
recursive, then don't do this.
-rough SQL 2003 todo, including tests to write:
+rough SQL 2011 todo, including tests to write:
-switch to SQL 2011
-
-now:
review the commented out reserved keyword entries and work out how to
fix
-go through all? the functions
-go through almost all the predicates
-window functions missing bits, window clauses
-from: more tests, review missing
- tablesample, unnest, etc.
-rows review
-match missing bit
-between symmetric
-case review
-detail list from the grammar, LNR = maybe leave until after next
- release, otherwise planned for next release
+big areas:
+window functions
+nested window functions
+case
+table ref
+joined table
+group by
+window clause
-LNR support needed MODULE syntax in identifiers - already covered?
-LNR decide how to represent special identifiers including the session
- variables or whatever they are called like current_user
-LNR grouping - needs special syntax?
-LNR next value for
-LNR subtypes, methods, new /routine, dereference
-LNR double check associativity, precedence (value exprs, joins, set ops)
-position expressions
-length expressions
-extract expression
-cardinality expression?
-check concatenations
-substring expressions
-regular expression substring function
-convert
-translate
-trim
-overlay
-LNR specifictype
-row value constructors, expressions review
-review table value constructor exactly what is allowed
-lots more tests for from clause variations
-tablesamples
-unnest
-table function derived table
-only spec
-join variations, including union join
-review group by
-window clauses
-LNR all fields reference with alias 'select * as (a,b,c) ... '
-search or cycle clause
-between symmetric/asymmetric
-in predicate review
-escape for like
-escape for [not] similar to
-LNR regular expression syntax?
-normalized predicate
-overlaps predicate
-distinct from predicate
-member predicate
-submultiset predicate
-set predicate
-LNR type predicate
-additional stuff review:
-complete the list of keywords/reserved keywords and check everything
- still works ok. The parser will reject all unquoted identifiers
- which are the same as reserved or unreserved keywords.
-LNR select into
-LNR other language format identifiers for host params?
+other areas:
+unicode escape, strings and idens
+character set behaviour review
+datetime literals
+mixed quoting identifier chains
+names/identifiers careful review
+typenames: lengths, binary
+general value bits
+ collate for
+numeric val fn
+string exp fn
+datetime exp fn
+interval exp fn
+rows
+table value constructor
+interval qualifier
+query spec
+with
+setop
+explicit table
+order/offset/fetch
+search/cycle
+preds:
+between
+in
+like
+similar
+regex like?
+null
+normalize
+match
+overlaps
+distinct
+member
+submultiset
+period
+create list of unsupported syntax: xml, ref, subtypes, modules?
---
@@ -248,6 +230,10 @@ string literals
full number literals -> other bases?
apply, pivot
+maybe add dml and ddl, source poses, quasi quotes
+
+leave: type check, dialects, procedural, separate lexing?
+
other dialect targets:
postgres
oracle
diff --git a/simple-sql-parser.cabal b/simple-sql-parser.cabal
index b13095b..fa8ea32 100644
--- a/simple-sql-parser.cabal
+++ b/simple-sql-parser.cabal
@@ -1,7 +1,10 @@
name: simple-sql-parser
version: 0.4.0
synopsis: A parser for SQL queries
-description: A parser for SQL queries. Please see the homepage for more information <http://jakewheat.github.io/simple-sql-parser/>.
+
+description: A parser for SQL queries. Parses most SQL:2011
+ queries. Please see the homepage for more information
+ <http://jakewheat.github.io/simple-sql-parser/>.
homepage: http://jakewheat.github.io/simple-sql-parser/
license: BSD3
@@ -36,6 +39,7 @@ library
default-language: Haskell2010
ghc-options: -Wall
other-extensions: TupleSections,DeriveDataTypeable
+
Test-Suite Tests
type: exitcode-stdio-1.0
main-is: RunTests.lhs
@@ -67,7 +71,7 @@ Test-Suite Tests
Language.SQL.SimpleSQL.Tpch,
Language.SQL.SimpleSQL.ValueExprs
- other-extensions: TupleSections,OverloadedStrings,DeriveDataTypeable
+ other-extensions: TupleSections,DeriveDataTypeable
default-language: Haskell2010
ghc-options: -Wall
@@ -78,6 +82,7 @@ executable SQLIndent
parsec >=3.1 && <3.2,
mtl >=2.1 && <2.2,
pretty >= 1.1 && < 1.2
+ other-extensions: TupleSections,DeriveDataTypeable
default-language: Haskell2010
ghc-options: -Wall
if flag(sqlindent)
diff --git a/tools/Language/SQL/SimpleSQL/SQL2003.lhs b/tools/Language/SQL/SimpleSQL/SQL2003.lhs
index 6031114..5b6e016 100644
--- a/tools/Language/SQL/SimpleSQL/SQL2003.lhs
+++ b/tools/Language/SQL/SimpleSQL/SQL2003.lhs
@@ -1130,8 +1130,6 @@ create a list of type name variations:
> ,("interval year(4) to second(2,3)"
> ,IntervalTypeName (Itf "year" $ Just (4,Nothing))
> (Just $ Itf "second" $ Just (2, Just 3)))
-> ,("ref (t)", RefTypeName [Name "t"] Nothing)
-> ,("ref (t) scope q", RefTypeName [Name "t"] (Just [Name "q"]))
> ]
Now test each variation in both cast expression and typed literal
diff --git a/tools/Language/SQL/SimpleSQL/SQL2011.lhs b/tools/Language/SQL/SimpleSQL/SQL2011.lhs
index d4d288e..d88ea48 100644
--- a/tools/Language/SQL/SimpleSQL/SQL2011.lhs
+++ b/tools/Language/SQL/SimpleSQL/SQL2011.lhs
@@ -16,6 +16,7 @@ which parts aren't currently supported.
> [literals
> ,identifiers
> ,typeNameTests
+> ,fieldDefinition
> ,valueExpressions
> ,queryExpressions
> ,scalarSubquery
@@ -494,6 +495,11 @@ Specify a non-null value.
> ,StringLit "a quote: ', stuff")
> ,("''"
> ,StringLit "")
+
+I'm not sure how this should work. Maybe the parser should reject non
+ascii characters in strings and identifiers unless the current SQL
+character set allows them.
+
> ,("_francais 'français'"
> ,TypedLit (TypeName [Name "_francais"]) "français")
> ]
@@ -1055,6 +1061,8 @@ create a list of type name variations:
> ,("decimal(15,2)", PrecScaleTypeName [Name "decimal"] 15 2)
> -- lob prec + with multiname
> ,("blob(3M)", LobTypeName [Name "blob"] 3 (Just LobM) Nothing)
+> ,("blob(3T)", LobTypeName [Name "blob"] 3 (Just LobT) Nothing)
+> ,("blob(3P)", LobTypeName [Name "blob"] 3 (Just LobP) Nothing)
> ,("blob(4M characters) "
> ,LobTypeName [Name "blob"] 4 (Just LobM) (Just LobCharacters))
> ,("blob(6G octets) "
@@ -1145,8 +1153,6 @@ create a list of type name variations:
> ,("interval year(4) to second(2,3)"
> ,IntervalTypeName (Itf "year" $ Just (4,Nothing))
> (Just $ Itf "second" $ Just (2, Just 3)))
-> ,("ref (t)", RefTypeName [Name "t"] Nothing)
-> ,("ref (t) scope q", RefTypeName [Name "t"] (Just [Name "q"]))
> ]
Now test each variation in both cast expression and typed literal
@@ -1169,7 +1175,13 @@ Define a field of a row type.
<field definition> ::= <field name> <data type>
-Tested in the row type above.
+> fieldDefinition :: TestItem
+> fieldDefinition = Group "field definition"
+> $ map (uncurry TestValueExpr)
+> [("cast('(1,2)' as row(a int,b char))"
+> ,Cast (StringLit "(1,2)")
+> $ RowTypeName [(Name "a", TypeName [Name "int"])
+> ,(Name "b", TypeName [Name "char"])])]
== 6.3 <value expression primary>
@@ -1214,6 +1226,8 @@ Specify a value that is syntactically self-delimited.
> [generalValueSpecification
> ,parameterSpecification
> ,contextuallyTypedValueSpecification
+> ,identifierChain
+> ,columnReference
> ,setFunctionSpecification
> ,windowFunction
> ,nestedWindowFunction
@@ -1384,7 +1398,10 @@ Disambiguate a <period>-separated chain of identifiers.
<basic identifier chain> ::= <identifier chain>
-tested with the identifier tests above
+> identifierChain :: TestItem
+> identifierChain = Group "identifier chain"
+> $ map (uncurry TestValueExpr)
+> [("a.b", Iden [Name "a",Name "b"])]
== 6.7 <column reference>
@@ -1395,7 +1412,10 @@ Reference a column.
<basic identifier chain>
| MODULE <period> <qualified identifier> <period> <column name>
-tested with the identifier tests above
+> columnReference :: TestItem
+> columnReference = Group "column reference"
+> $ map (uncurry TestValueExpr)
+> [("module.a.b", Iden [Name "module",Name "a",Name "b"])]
== 6.8 <SQL parameter reference>
@@ -1404,8 +1424,6 @@ Reference an SQL parameter.
<SQL parameter reference> ::= <basic identifier chain>
-tested with the identifier tests above
-
== 6.9 <set function specification>
Function
@@ -1590,6 +1608,11 @@ Specify a conditional value.
| <set predicate part 2>
| <type predicate part 2>
+I haven't seen these part 2 style when operands in the wild. It
+doesn't even allow all the binary operators here. We will allow them
+all, and parser and represent these expressions by considering all the
+binary ops as unary prefix ops.
+
<result> ::= <result expression> | NULL
<result expression> ::= <value expression>
@@ -1611,8 +1634,6 @@ Specify a data conversion.
<cast target> ::= <domain name> | <data type>
-This is already tested with the type name tests
-
> castSpecification :: TestItem
> castSpecification = Group "cast specification"
> $ map (uncurry TestValueExpr)
@@ -2230,7 +2251,6 @@ Specify an interval value.
> intervalValueExpression :: TestItem
> intervalValueExpression = Group "interval value expression"
> [-- todo: interval value expression
-> intervalValueFunction
> ]
@@ -2334,7 +2354,6 @@ Specify an array value.
> arrayValueExpression :: TestItem
> arrayValueExpression = Group "array value expression"
> [-- todo: array value expression
-> arrayValueFunction
> ]
== 6.37 <array value function>
@@ -3080,7 +3099,9 @@ the result of recursive query expressions.
== 7.15 <subquery>
Function
-Specify a scalar value, a row, or a table derived from a <query expression>.
+
+Specify a scalar value, a row, or a table derived from a <query
+expression>.
<scalar subquery> ::= <subquery>
@@ -3169,9 +3190,24 @@ Specify a comparison of two row values.
> comparisonPredicates :: TestItem
> comparisonPredicates = Group "comparison predicates"
-> [-- todo: comparison predicates
+> $ map (uncurry TestValueExpr)
+> $ map mkOp ["=", "<>", "<", ">", "<=", ">="]
+> ++ [("ROW(a) = ROW(b)"
+> ,BinOp (App [Name "ROW"] [a])
+> [Name "="]
+> (App [Name "ROW"] [b]))
+> ,("(a,b) = (c,d)"
+> ,BinOp (SpecialOp [Name "rowctor"] [a,b])
+> [Name "="]
+> (SpecialOp [Name "rowctor"] [Iden [Name "c"], Iden [Name "d"]]))
> ]
+> where
+> mkOp nm = ("a " ++ nm ++ " b"
+> ,BinOp a [Name nm] b)
+> a = Iden [Name "a"]
+> b = Iden [Name "b"]
+TODO: what other tests, more complex expressions with comparisons?
== 8.3 <between predicate>