git.iko.soy/syntax
on master help sign in
swap
64f67b739 ahead, 0 behind
Data/Syntax.hs+128−64
since splitsnapshots
unifiedsplit
@@ -1,7 +1,6 @@
1+{-# LANGUAGE TypeFamilies #-}
1{-# LANGUAGE FlexibleContexts #-}
2{-# LANGUAGE FlexibleContexts #-}
2{-# LANGUAGE DefaultSignatures #-}
3{-# LANGUAGE DefaultSignatures #-}
3−{-# LANGUAGE MultiParamTypeClasses #-}
4−{-# LANGUAGE FunctionalDependencies #-}
5{- |
4{- |
6Module : Data.Syntax
5Module : Data.Syntax
7Description : Abstract syntax description.
6Description : Abstract syntax description.
@@ -11,22 +10,28 @@ License : MIT
11Maintainer : Paweł Nowak <pawel834@gmail.com>
10Maintainer : Paweł Nowak <pawel834@gmail.com>
12Stability : experimental
11Stability : experimental
13
12
14−Abstract syntax descriptions based on semi-isomorphisms.
13+Reversible parsing and pretty-printing.
15-}
14-}
16module Data.Syntax (
15module Data.Syntax (
17 -- * Syntax.
16 -- * Syntax.
18 Syntax(..),
17 Syntax(..),
18+ Isolable(..),
19 -- * Common isomorphisms.
19 -- * Common isomorphisms.
20 packed
20 packed
21 ) where
21 ) where
22
22
23−import Prelude hiding (take, takeWhile)
23+import Prelude hiding (take, takeWhile, id, (.))
24
24
25+import Control.Category
26+import Control.Category.Reader
27+import Control.Category.Structures
25−import Control.Lens
28+import Control.Lens.Iso
26−import Control.Lens.SemiIso
29+import Control.Lens.SemiIso
30+import Control.SIArrow
27−import Data.MonoTraversable
31+import Data.MonoTraversable
28−import Data.SemiIsoFunctor
29−import Data.Sequences hiding (take, takeWhile)
32+import Data.Sequences hiding (take, takeWhile, replicate)
33+import qualified Data.Vector as V
34+import qualified Data.Vector.Unboxed as VU
30
35
31-- | An isomorphism between a sequence and a list of its elements.
36-- | An isomorphism between a sequence and a list of its elements.
32packed :: IsSequence seq => Iso' seq [Element seq]
37packed :: IsSequence seq => Iso' seq [Element seq]
@@ -34,91 +39,150 @@ packed = iso otoList fromList
34
39
35-- | An abstract syntax description based on semi-isomorphisms.
40-- | An abstract syntax description based on semi-isomorphisms.
36--
41--
37−-- This class can be implemented by both parsers and printers (and maybe more?).
42+-- This class can be implemented by both parsers and printers.
38−--
39−-- The usual use is to write a polymorphic syntax description and instantiate it
40−-- both as a parser and a printer. An example syntax description:
41--
43--
42−-- > | A simple untyped lambda calculus.
44+-- The usual use is to write a polymorphic syntax description and instantiate it
43−-- > data AST = Var Text
45+-- both as a parser and a printer. Examples are available in 'syntax-example' and
44−-- > | App AST AST
46+-- 'syntax-example-json' packages.
45−-- > | Abs Text AST
46−-- > deriving (Show)
47−-- >
48−-- > $(makePrisms ''AST)
49−-- >
50−-- > -- | A variable name.
51−-- > name :: Syntax syn Text => syn Text
52−-- > name = S.takeWhile1 isAlphaNum
53−-- >
54−-- > -- | Encloses a symbol in parentheses.
55−-- > parens :: Syntax syn Text => syn a -> syn a
56−-- > parens m = S.char '(' */ S.spaces_ */ m /* S.spaces_ /* S.char ')'
57−-- >
58−-- > -- | An atom is a variable or an expression in parentheses.
59−-- > atom :: Syntax syn Text => syn AST
60−-- > atom = _Var /$/ name
61−-- > /|/ parens expr
62−-- >
63−-- > -- | An expression of our lambda calculus.
64−-- > expr :: Syntax syn Text => syn AST
65−-- > expr = _App /$/ atom /* S.spaces1 /*/ atom
66−-- > /|/ _Abs /$/ S.char '\\' /* S.spaces_
67−-- > */ name /* S.spaces
68−-- > /* S.string "->" /* S.spaces
69−-- > /*/ expr
70−-- > /|/ atom
71--
47--
72-- Methods of this class try to mimic "Data.Attoparsec.Text" interface.
48-- Methods of this class try to mimic "Data.Attoparsec.Text" interface.
73−class ( SemiIsoAlternative syn
49+class ( SIArrow syn
74− , IsSequence seq
50+ , IsSequence (Seq syn)
75− , Eq seq
51+ , Eq (Seq syn)
76− , Eq (Element seq))
52+ , Eq (Element (Seq syn)))
77− => Syntax syn seq | syn -> seq
53+ => Syntax syn
78 where
54 where
79−
55+ -- | The sequence type used by this syntax.
56+ type Seq syn :: *
57+
80 -- | Any character.
58 -- | Any character.
81− anyChar :: syn (Element seq)
59+ anyChar :: syn () (Element (Seq syn))
82
60
83 -- | A specific character.
61 -- | A specific character.
84− char :: Element seq -> syn ()
62+ char :: Element (Seq syn) -> syn () ()
85− char c = exact c /$/ anyChar
63+ char c = rev (exact c) /$/ anyChar
86
64
87 -- | Any character except the given one.
65 -- | Any character except the given one.
88− notChar :: Element seq -> syn (Element seq)
66+ notChar :: Element (Seq syn) -> syn () (Element (Seq syn))
89− notChar c = filtered (/= c) /$/ anyChar
67+ notChar c = bifiltered (/= c) /$/ anyChar
90
68
91 -- | Any character satisfying a predicate.
69 -- | Any character satisfying a predicate.
92− satisfy :: (Element seq -> Bool) -> syn (Element seq)
70+ satisfy :: (Element (Seq syn) -> Bool) -> syn () (Element (Seq syn))
93− satisfy p = filtered p /$/ anyChar
71+ satisfy p = bifiltered p /$/ anyChar
94
72
95 -- | Transforms a character using a SemiIso and filters out values
73 -- | Transforms a character using a SemiIso and filters out values
96 -- not satisfying the predicate.
74 -- not satisfying the predicate.
97− satisfyWith :: ASemiIso' a (Element seq) -> (a -> Bool) -> syn a
75+ satisfyWith :: ASemiIso' a (Element (Seq syn)) -> (a -> Bool) -> syn () a
98− satisfyWith ai p = filtered p . ai /$/ anyChar
76+ satisfyWith ai p = bifiltered p . ai /$/ anyChar
99
77
100 -- | A specific string.
78 -- | A specific string.
101− string :: seq -> syn ()
79+ string :: Seq syn -> syn () ()
102− string s = exact s /$/ take (olength s)
80+ string s = rev (exact s) /$/ take (olength s)
103
81
104 -- | A string of length @n@.
82 -- | A string of length @n@.
105− take :: Int -> syn seq
83+ take :: Int -> syn () (Seq syn)
106 take n = packed /$/ sireplicate n anyChar
84 take n = packed /$/ sireplicate n anyChar
107
85
108 -- | Maximal string which elements satisfy a predicate.
86 -- | Maximal string which elements satisfy a predicate.
109− takeWhile :: (Element seq -> Bool) -> syn seq
87+ takeWhile :: (Element (Seq syn) -> Bool) -> syn () (Seq syn)
110 takeWhile p = packed /$/ simany (satisfy p)
88 takeWhile p = packed /$/ simany (satisfy p)
111
89
112 -- | Maximal non-empty string which elements satisfy a predicate.
90 -- | Maximal non-empty string which elements satisfy a predicate.
113− takeWhile1 :: (Element seq -> Bool) -> syn seq
91+ takeWhile1 :: (Element (Seq syn) -> Bool) -> syn () (Seq syn)
114 takeWhile1 p = packed /$/ sisome (satisfy p)
92 takeWhile1 p = packed /$/ sisome (satisfy p)
115
93
116 -- | Maximal string which elements do not satisfy a predicate.
94 -- | Maximal string which elements do not satisfy a predicate.
117− takeTill :: (Element seq -> Bool) -> syn seq
95+ takeTill :: (Element (Seq syn) -> Bool) -> syn () (Seq syn)
118 takeTill p = takeWhile (not . p)
96 takeTill p = takeWhile (not . p)
119
97
120 -- | Maximal non-empty string which elements do not satisfy a predicate.
98 -- | Maximal non-empty string which elements do not satisfy a predicate.
121− takeTill1 :: (Element seq -> Bool) -> syn seq
99+ takeTill1 :: (Element (Seq syn) -> Bool) -> syn () (Seq syn)
122 takeTill1 p = takeWhile1 (not . p)
100 takeTill1 p = takeWhile1 (not . p)
123
101
102+ -- | Constant size vector. The default implementation uses lists, but
103+ -- "syntax-attoparsec" and "syntax-printer" override it with an efficient
104+ -- implementation that works on a vector directly.
105+ --
106+ -- @vecN n e@ describes a size @n@ vector with elements @e@.
107+ --
108+ -- Also see 'Data.Syntax.Combinator.vec'.
109+ vecN :: Int -> syn () a -> syn () (V.Vector a)
110+ vecN n e = packed /$/ sireplicate n e
111+
112+ -- | Constant size vector with index-aware element. The default implementation
113+ -- uses lists, but "syntax-attoparsec" and "syntax-printer" override it with an
114+ -- efficient implementation that works on a vector directly.
115+ --
116+ -- @ivecN n e@ describes a size @n@ vector with elements @e@. Each element
117+ -- gets its index and should output a value and the index unchanged.
118+ --
119+ -- Also see 'Data.Syntax.Combinator.ivec'.
120+ ivecN :: Int -> syn Int (Int, a) -> syn () (V.Vector a)
121+ ivecN n e = (packed /$/)
122+ $ sisequence
123+ $ map (\(i, e') -> constant i ^>> e'
124+ >>> first (sipure (constant i))
125+ >># unit . swapped)
126+ $ zip [0 .. n-1] (replicate n e)
127+
128+ -- | Constant size unboxed vector. The default implementation uses lists, but
129+ -- "syntax-attoparsec" and "syntax-printer" override it with an efficient
130+ -- implementation that works on a vector directly.
131+ --
132+ -- @vecN n e@ describes a size @n@ vector with elements @e@.
133+ --
134+ -- Also see 'Data.Syntax.Combinator.vec'.
135+ uvecN :: VU.Unbox a => Int -> syn () a -> syn () (VU.Vector a)
136+ uvecN n e = packed /$/ sireplicate n e
137+
138+ -- | Constant size unboxed vector with index-aware element. The default implementation
139+ -- uses lists, but "syntax-attoparsec" and "syntax-printer" override it with an
140+ -- efficient implementation that works on a vector directly.
141+ --
142+ -- @ivecN n e@ describes a size @n@ vector with elements @e@. Each element
143+ -- gets its index and should output a value and the index unchanged.
144+ --
145+ -- Also see 'Data.Syntax.Combinator.ivec'.
146+ uivecN :: VU.Unbox a => Int -> syn Int (Int, a) -> syn () (VU.Vector a)
147+ uivecN n e = (packed /$/)
148+ $ sisequence
149+ $ map (\(i, e') -> constant i ^>> e'
150+ >>> first (sipure (constant i))
151+ >># unit . swapped)
152+ $ zip [0 .. n-1] (replicate n e)
153+
124 {-# MINIMAL anyChar #-}
154 {-# MINIMAL anyChar #-}
155+
156+instance Syntax syn => Syntax (ReaderCT env syn) where
157+ type Seq (ReaderCT env syn) = Seq syn
158+ anyChar = clift anyChar
159+ char = clift . char
160+ notChar = clift . notChar
161+ satisfy = clift . satisfy
162+ satisfyWith ai = clift . satisfyWith ai
163+ string = clift . string
164+ take = clift . take
165+ takeWhile = clift . takeWhile
166+ takeWhile1 = clift . takeWhile1
167+ takeTill = clift . takeTill
168+ takeTill1 = clift . takeTill1
169+ vecN n e = ReaderCT $ vecN n . runReaderCT e
170+ ivecN n e = ReaderCT $ ivecN n . runReaderCT e
171+ uvecN n e = ReaderCT $ uvecN n . runReaderCT e
172+ uivecN n e = ReaderCT $ uivecN n . runReaderCT e
173+
174+-- | Execute a computation in an isolated context.
175+--
176+-- The motivating example: you want to write a function
177+--
178+-- > serializeList :: Syntax syn => syn () a -> syn () [a]
179+--
180+-- Notice that we cannot just use simany, because the first @syn () a@
181+-- could eat the entire sequence (even though we printed more than 1 value!), so
182+-- we have to insert some kind of separators between the element. But how can we
183+-- be sure that @syn () a@ will not eat our separators? We can't! Thats why we
184+-- have to do the parsing in two stages: first extract the sequence between separators,
185+-- then run the @syn () a@ on this sequence.
186+class Syntax syn => Isolable syn where
187+ -- | Turns a computation into one executed in an isolated context.
188+ isolate :: syn () b -> syn (Seq syn) b
Data/Syntax/Char.hs+81−24
@@ -11,44 +11,101 @@ Stability : experimental
11
11
12Common combinators that work with sequences of chars.
12Common combinators that work with sequences of chars.
13
13
14−There are A LOT of combinators missing.
15-}
14-}
16module Data.Syntax.Char (
15module Data.Syntax.Char (
17− SyntaxChar,
16+ SyntaxChar(..),
17+ SyntaxText,
18+ signed,
18 spaces,
19 spaces,
19 spaces_,
20 spaces_,
20 spaces1,
21 spaces1,
21− spaces1_,
22− endOfLine
22+ endOfLine,
23+ digitDec,
24+ digitOct,
25+ digitHex
23 ) where
26 ) where
24
27
28+import Control.Category.Reader
29+import Control.Category.Structures
25−import Control.Lens.SemiIso
30+import Control.Lens.SemiIso
31+import Control.SIArrow
32+import Data.Bits
26−import Data.Char
33+import Data.Char
27−import Data.MonoTraversable
34+import Data.MonoTraversable
28−import Data.Monoid
35+import Data.Scientific (Scientific)
29−import Data.SemiIsoFunctor
36+import Data.Syntax
30−import Data.Syntax (Syntax)
37+import Data.Syntax.Combinator
31−import qualified Data.Syntax as S
38+import Data.Text (Text)
32
39
33-- | Syntax constrainted to sequences of chars.
40-- | Syntax constrainted to sequences of chars.
41+--
42+-- Note: methods of this class do not have default implementations (for now),
43+-- because their code is quite ugly and already written in most parser libraries.
34−type SyntaxChar syn seq = (Syntax syn seq, Element seq ~ Char)
44+class (Syntax syn, Element (Seq syn) ~ Char) => SyntaxChar syn where
45+ -- | An unsigned decimal number.
46+ decimal :: Integral a => syn () a
47+
48+ -- | An unsigned hexadecimal number.
49+ hexadecimal :: (Integral a, Bits a) => syn () a
50+
51+ -- | A signed real number.
52+ realFloat :: RealFloat a => syn () a
53+
54+ -- | A scientific number.
55+ scientific :: syn () Scientific
56+
57+ {-# MINIMAL decimal, hexadecimal, realFloat, scientific #-}
58+
59+instance SyntaxChar syn => SyntaxChar (ReaderCT env syn) where
60+ decimal = clift decimal
61+ hexadecimal = clift hexadecimal
62+ scientific = clift scientific
63+ realFloat = clift realFloat
64+
65+-- | An useful synonym for SyntaxChars with Text sequences.
66+type SyntaxText syn = (SyntaxChar syn, Seq syn ~ Text)
67+
68+-- | A number with an optional leading '+' or '-' sign character.
69+signed :: (Real a, SyntaxChar syn) => syn () a -> syn () a
70+signed n = _Negative /$/ char '-' */ n
71+ /+/ opt_ (char '+') */ n
35
72
36-- | Accepts zero or more spaces. Generates a single space.
73-- | Accepts zero or more spaces. Generates a single space.
37−spaces :: SyntaxChar syn seq => syn ()
74+spaces :: SyntaxChar syn => syn () ()
38−spaces = constant (opoint ' ') /$/ S.takeWhile isSpace
75+spaces = opt spaces1
39
76
40-- | Accepts zero or more spaces. Generates no output.
77-- | Accepts zero or more spaces. Generates no output.
41−spaces_ :: SyntaxChar syn seq => syn ()
78+spaces_ :: SyntaxChar syn => syn () ()
42−spaces_ = constant mempty /$/ S.takeWhile isSpace
79+spaces_ = opt_ spaces1
43
80
44-- | Accepts one or more spaces. Generates a single space.
81-- | Accepts one or more spaces. Generates a single space.
45−spaces1 :: SyntaxChar syn seq => syn ()
82+spaces1 :: SyntaxChar syn => syn () ()
46−spaces1 = constant (opoint ' ') /$/ S.takeWhile1 isSpace
83+spaces1 = constant (opoint ' ') /$/ takeWhile1 isSpace
47−
48−-- | Accepts one or more spaces. Generates no output.
49−spaces1_ :: SyntaxChar syn seq => syn ()
50−spaces1_ = constant mempty /$/ S.takeWhile1 isSpace
51
84
52-- | Accepts a single newline. Generates a newline.
85-- | Accepts a single newline. Generates a newline.
53−endOfLine :: SyntaxChar syn seq => syn ()
86+endOfLine :: SyntaxChar syn => syn () ()
54−endOfLine = S.char '\n'
87+endOfLine = char '\n'
88+
89+-- | A decimal digit.
90+digitDec :: SyntaxChar syn => syn () Int
91+digitDec = semiIso toChar toInt /$/ anyChar
92+ where toInt c | isDigit c = Right (digitToInt c)
93+ | otherwise = Left ("Expected a decimal digit, got " ++ [c])
94+ toChar i | i >= 0 && i <= 9 = Right (intToDigit i)
95+ | otherwise = Left ("Expected a decimal digit, got number " ++ show i)
96+
97+-- | An octal digit.
98+digitOct :: SyntaxChar syn => syn () Int
99+digitOct = semiIso toChar toInt /$/ anyChar
100+ where toInt c | isOctDigit c = Right (digitToInt c)
101+ | otherwise = Left ("Expected an octal digit, got " ++ [c])
102+ toChar i | i >= 0 && i <= 7 = Right (intToDigit i)
103+ | otherwise = Left ("Expected an octal digit, got number " ++ show i)
104+
105+-- | A hex digit.
106+digitHex :: SyntaxChar syn => syn () Int
107+digitHex = semiIso toChar toInt /$/ anyChar
108+ where toInt c | isHexDigit c = Right (digitToInt c)
109+ | otherwise = Left ("Expected a hex digit, got " ++ [c])
110+ toChar i | i >= 0 && i <= 15 = Right (intToDigit i)
111+ | otherwise = Left ("Expected a hex digit, got number " ++ show i)
Data/Syntax/Combinator.hsnew file+190
@@ -0,0 +1,190 @@
1+{-# LANGUAGE RankNTypes #-}
2+{-# LANGUAGE ScopedTypeVariables #-}
3+{-# LANGUAGE FlexibleContexts #-}
4+{- |
5+Module : Data.Syntax.Combinator
6+Description : Combinators that work with any sequence type.
7+Copyright : (c) Daan Leijen 1999-2001, Bryan O'Sullivan 2007-2014, Paweł Nowak 2014
8+License : MIT
9+
10+Maintainer : Paweł Nowak <pawel834@gmail.com>
11+Stability : experimental
12+
13+Combinators that work with any sequence type.
14+-}
15+module Data.Syntax.Combinator
16+ (
17+ -- * Combinators.
18+ optional,
19+ opt,
20+ opt_,
21+ choice,
22+ eitherOf,
23+
24+ -- * Lists.
25+ manyTill,
26+ sepBy,
27+ sepBy1,
28+
29+ -- * Arrowized combinators.
30+ takeArr,
31+
32+ -- * Vectors.
33+ vecNSepBy,
34+ ivecNSepBy,
35+ vec,
36+ vecSepBy,
37+ ivec,
38+ ivecSepBy,
39+
40+ -- * Unboxed vectors.
41+ uvecNSepBy,
42+ uivecNSepBy,
43+ uvec,
44+ uvecSepBy,
45+ uivec,
46+ uivecSepBy
47+ ) where
48+
49+import Control.Category
50+import Control.Category.Structures
51+import Control.Lens
52+import Control.Lens.SemiIso
53+import Control.SIArrow
54+import qualified Data.MonoTraversable as Seq
55+import Data.Syntax
56+import qualified Data.Vector as V
57+import qualified Data.Vector.Unboxed as VU
58+import Prelude hiding (id, (.), take)
59+
60+-- | One or zero occurences of @f@.
61+optional :: SIArrow cat => cat () a -> cat () (Maybe a)
62+optional f = _Just /$/ f /+/ sipure _Nothing
63+
64+-- | Like 'optional', but specialized for @()@.
65+opt :: SIArrow cat => cat () () -> cat () ()
66+opt f = f /+/ sipure id
67+
68+-- | Parser one or zero occurences of @f@, but prints nothing.
69+opt_ :: SIArrow cat => cat () () -> cat () ()
70+opt_ f = semiIso (const (Left "opt_")) Right /$/ f
71+ /+/ sipure id
72+
73+-- | Tries to apply the actions in the list in order, until one of
74+-- them succeeds. Returns the value of the succeeding action.
75+choice :: SIArrow cat => [cat () a] -> cat () a
76+choice = foldr (/+/) (sifail "choice: all alternatives failed")
77+
78+-- | Combine two alternatives.
79+eitherOf :: SIArrow cat => cat () a -> cat () b -> cat () (Either a b)
80+eitherOf a b = _Left /$/ a /+/ _Right /$/ b
81+
82+-- | @manyTill p end@ applies action p zero or more times until action
83+-- end succeeds, and returns the list of values returned by p.
84+manyTill :: SIArrow cat => cat () a -> cat () () -> cat () [a]
85+manyTill p end = _Empty /$/ end
86+ /+/ _Cons /$/ p /*/ manyTill p end
87+
88+-- | Zero or more occurences of @v@ separated by @s@.
89+sepBy :: SIArrow cat => cat () a -> cat () () -> cat () [a]
90+sepBy v s = sepBy1 v s /+/ sipure _Empty
91+
92+-- | One or more occurences of @v@ separated by @s@.
93+sepBy1 :: SIArrow cat => cat () a -> cat () () -> cat () [a]
94+sepBy1 v s = _Cons /$/ v /*/ (s */ sepBy1 v s /+/ sipure _Empty)
95+
96+-- | A string of given length.
97+takeArr :: Syntax syn => syn Int (Seq syn)
98+takeArr = sibind $ iso (\n -> constant n #>> take n)
99+ (\s -> constant (Seq.olength s) #>> take (Seq.olength s))
100+
101+-- | Constant size vector with separators.
102+--
103+-- @vecNSepBy n e sep@ describes a size @n@ vector with elements @e@ separated by @sep@.
104+vecNSepBy :: Syntax syn => Int -> syn () a -> syn () () -> syn () (V.Vector a)
105+vecNSepBy n e sep = ivecNSepBy n (unit ^>> second e) sep
106+
107+-- | Constant size vector with separators and index-aware elements.
108+--
109+-- @ivecNSepBy n e sep@ describes a size @n@ vector with elements @e@ separated by @sep@.
110+-- Each element gets its index and should output a value and the index unchanged.
111+ivecNSepBy :: Syntax syn => Int -> syn Int (Int, a) -> syn () () -> syn () (V.Vector a)
112+ivecNSepBy n e sep = ivecN n $ sibind $ iso el (el . fst)
113+ where
114+ el k | k < n - 1 = unit ^>> e *** sep >># unit
115+ | otherwise = e
116+
117+-- | Runtime sized vector. The size can depend on the result of some computation.
118+--
119+-- @vec e@ describes a vector with elements @e@.
120+vec :: Syntax syn => syn () a -> syn Int (V.Vector a)
121+vec e = vecSepBy e (siarr id)
122+
123+-- | Runtime sized vector with separators. The size can depend on the result of some
124+-- computation.
125+--
126+-- @vecSepBy e sep@ describes a vector with elements @e@ separated by @sep@.
127+vecSepBy :: Syntax syn => syn () a -> syn () () -> syn Int (V.Vector a)
128+vecSepBy e sep = ivecSepBy (unit ^>> second e) sep
129+
130+-- | Runtime sized vector with index-aware elements. The size can depend on the result
131+-- of some computation.
132+--
133+-- @ivec e@ describes a vector with elements @e@.
134+ivec :: Syntax syn => syn Int (Int, a) -> syn Int (V.Vector a)
135+ivec e = ivecSepBy e (siarr id)
136+
137+-- | Runtime sized vector with index-aware elements and separators. The size can depend
138+-- on the result of some computation.
139+--
140+-- @ivecSepBy e sep@ describes a vector with elements @e@ separated by @sep@.
141+ivecSepBy :: Syntax syn => syn Int (Int, a) -> syn () () -> syn Int (V.Vector a)
142+ivecSepBy e sep = sibind $ iso (\n -> constant n #>> ivecNSepBy n e sep)
143+ (\v -> constant (V.length v) #>> ivecNSepBy (V.length v) e sep)
144+
145+-- | Constant size unboxed vector with separators.
146+--
147+-- @uvecNSepBy n e sep@ describes a size @n@ vector with elements @e@ separated by @sep@.
148+uvecNSepBy :: (Syntax syn, VU.Unbox a) => Int -> syn () a -> syn () () -> syn () (VU.Vector a)
149+uvecNSepBy n e sep = uivecNSepBy n (unit ^>> second e) sep
150+
151+-- | Constant size unboxed vector with separators and index-aware elements.
152+--
153+-- @uivecNSepBy n e sep@ describes a size @n@ vector with elements @e@ separated by @sep@.
154+-- Each element gets its index and should output a value and the index unchanged.
155+uivecNSepBy :: (Syntax syn, VU.Unbox a) => Int -> syn Int (Int, a)
156+ -> syn () () -> syn () (VU.Vector a)
157+uivecNSepBy n e sep = uivecN n $ sibind $ iso el (el . fst)
158+ where
159+ el k | k < n - 1 = unit ^>> e *** sep >># unit
160+ | otherwise = e
161+
162+-- | Runtime sized unboxed vector. The size can depend on the result of some computation.
163+--
164+-- @uvec e@ describes a vector with elements @e@.
165+uvec :: (Syntax syn, VU.Unbox a) => syn () a -> syn Int (VU.Vector a)
166+uvec e = uvecSepBy e (siarr id)
167+
168+-- | Runtime sized unboxed vector with separators. The size can depend on the result of some
169+-- computation.
170+--
171+-- @uvecSepBy e sep@ describes a vector with elements @e@ separated by @sep@.
172+uvecSepBy :: (Syntax syn, VU.Unbox a) => syn () a -> syn () () -> syn Int (VU.Vector a)
173+uvecSepBy e sep = uivecSepBy (unit ^>> second e) sep
174+
175+-- | Runtime sized unboxed vector with index-aware elements. The size can depend on the result
176+-- of some computation.
177+--
178+-- @uivec e@ describes a vector with elements @e@.
179+uivec :: (Syntax syn, VU.Unbox a) => syn Int (Int, a) -> syn Int (VU.Vector a)
180+uivec e = uivecSepBy e (siarr id)
181+
182+-- | Runtime sized unboxed vector with index-aware elements and separators. The size can depend
183+-- on the result of some computation.
184+--
185+-- @uivecSepBy e sep@ describes a vector with elements @e@ separated by @sep@.
186+uivecSepBy :: (Syntax syn, VU.Unbox a) => syn Int (Int, a)
187+ -> syn () () -> syn Int (VU.Vector a)
188+uivecSepBy e sep = sibind $ iso (\n -> constant n #>> uivecNSepBy n e sep)
189+ (\v -> constant (VU.length v)
190+ #>> uivecNSepBy (VU.length v) e sep)
Data/Syntax/Indent.hsnew file+73
@@ -0,0 +1,73 @@
1+{-# LANGUAGE GeneralizedNewtypeDeriving #-}
2+{-# LANGUAGE FlexibleInstances #-}
3+{-# LANGUAGE TypeFamilies #-}
4+{- |
5+Module : Data.Syntax.Indent
6+Description : Simple indentation.
7+Copyright : (c) Paweł Nowak
8+License : MIT
9+
10+Maintainer : Paweł Nowak <pawel834@gmail.com>
11+Stability : experimental
12+
13+Provides a very simple indentation as a category transformer.
14+-}
15+module Data.Syntax.Indent (
16+ Indent,
17+ runIndent,
18+ breakLine,
19+ indented
20+ ) where
21+
22+import Control.Category
23+import Control.Category.Reader
24+import Control.Category.Structures
25+import Control.SIArrow
26+import Data.Syntax
27+import Data.Syntax.Char
28+import Data.Syntax.Combinator
29+import Prelude hiding (takeWhile, take, id, (.))
30+
31+-- | Adds indentation to a syntax description.
32+newtype Indent cat a b = Indent { getIndent :: ReaderCT (Int, cat () ()) cat a b }
33+ deriving (Category, Products, Coproducts, CatPlus, SIArrow)
34+
35+instance CatTrans Indent where
36+ clift = Indent . clift
37+
38+-- Generalized newtype deriving cannot derive this :(
39+
40+instance Syntax syn => Syntax (Indent syn) where
41+ type Seq (Indent syn) = Seq syn
42+ anyChar = Indent anyChar
43+ char = Indent . char
44+ notChar = Indent . notChar
45+ satisfy = Indent . satisfy
46+ satisfyWith ai = Indent . satisfyWith ai
47+ string = Indent . string
48+ take = Indent . take
49+ takeWhile = Indent . takeWhile
50+ takeWhile1 = Indent . takeWhile1
51+ takeTill = Indent . takeTill
52+ takeTill1 = Indent . takeTill1
53+
54+instance SyntaxChar syn => SyntaxChar (Indent syn) where
55+ decimal = Indent decimal
56+ hexadecimal = Indent hexadecimal
57+ scientific = Indent scientific
58+ realFloat = Indent realFloat
59+
60+-- | @runIndent m tab@ runs the 'Indent' transformer using @tab@ once for each
61+-- level of indentation.
62+runIndent :: Indent cat a b -> cat () () -> cat a b
63+runIndent (Indent m) tab = runReaderCT m (0, tab)
64+
65+-- | Inserts a new line and correct indentation, but does not
66+-- require any formatting when parsing (it just skips all white space).
67+breakLine :: SyntaxChar syn => Indent syn () ()
68+breakLine = Indent . ReaderCT $ \(i, tab) ->
69+ opt (char '\n') /* opt (sireplicate_ i tab) /* spaces_
70+
71+-- | Increases the indentation level of its argument by one.
72+indented :: Indent cat a b -> Indent cat a b
73+indented (Indent f) = Indent . ReaderCT $ \(i, tab) -> runReaderCT f (i + 1, tab)
syntax.cabal+34−3
@@ -1,6 +1,35 @@
1name: syntax
1name: syntax
2−version: 0.1.0.0
2+version: 1.0.0.0
3−synopsis: Abstract syntax descriptions for parsing and pretty-printing.
3+synopsis: Reversible parsing and pretty-printing.
4+description:
5+ "syntax" allows you to write a single syntax description and instantiate is both as a parser and a pretty printer.
6+ .
7+ Syntax descriptions are written in applicative or arrow style. The library uses a custom typeclass hierarchy, provided
8+ by the "semi-iso" package. Most of the time you will be using operators like '/$/', '/*/' and '/+/' (= '<|>'), just
9+ like parser combinators. When more power is needed - e.g. when the syntax depends on the parsed or printed value -
10+ you turn to arrows.
11+ .
12+ Semi-isomorphisms from "semi-iso" are the basic building block of syntax descriptions. I recommend reading
13+ the hackage page of "semi-iso" first, as it contains much more information.
14+ .
15+ Once you write a syntax description (polymorphic in the syntax category) you can instantiate it both as a parser or as
16+ a pretty-printer. The library "syntax-attoparsec" gives you the ability to extract an Attoparsec parser. Pretty-printing
17+ is implemented by the "syntax-printer" library, which uses Text and ByteString builders. (Note that formatting is handled
18+ by "syntax" itself, not by the printer library)
19+ .
20+ Advanced formatting and parsing (for example indentation, haskell layout rule) is implemented as category transformers
21+ (similar to monad transformers). Currently only simple indentation is implemented (in "Data.Syntax.Indent") - basically
22+ a reader category transformer that tracks current indentation level. I plan on implementing Haskell layout rule in the
23+ future.
24+ .
25+ The library can work with both text and binary data. Alas, there are no binary combinators implemented yet.
26+ I will implement them when i have the time (but these category transformers look so much more interesting for now ;).
27+ .
28+ EXAMPLES! See @syntax-example@ and @syntax-example-json@ for examples.
29+ .
30+ * "syntax-example" implements a simple lambda calculus.
31+ .
32+ * "syntax-example-json" implements a json parser and pretty printer.
4license: MIT
33license: MIT
5license-file: LICENSE
34license-file: LICENSE
6author: Paweł Nowak
35author: Paweł Nowak
@@ -17,5 +46,7 @@ source-repository head
17library
46library
18 exposed-modules: Data.Syntax
47 exposed-modules: Data.Syntax
19 Data.Syntax.Char
48 Data.Syntax.Char
49+ Data.Syntax.Combinator
50+ Data.Syntax.Indent
20− build-depends: base, mono-traversable, lens, semi-iso
51+ build-depends: base >= 4 && < 5, mono-traversable, lens >= 4, semi-iso >= 1, scientific >= 0.3, text, vector
21 default-language: Haskell2010
52 default-language: Haskell2010