1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
|
module Parser (
parseSourceFile,
) where
import Control.Monad
import Control.Monad.Reader
import Data.Char
import Data.Either
import Data.List (intersperse)
import qualified Data.Text.Lazy as T
import Data.Text.Lazy (Text)
import qualified Data.Text.Lazy.Builder as B
import Data.Text.Lazy.Builder (Builder)
import Text.Parsec
import Pattern
import SourceFile
type Parser = ParsecT String () (Reader Pattern)
type Parser' = Parsec String ()
parseSourceFile :: FilePath -> String -> Pattern -> Either ParseError SourceFile
parseSourceFile fname source pat = runReader (runParserT pSourceFile () fname source) pat
pSourceFile :: Parser SourceFile
pSourceFile = do
SourceFile . concat <$> many (pDataDef <|> pInstanceDef <|> pFunDef <|> skipLine)
where
skipLine :: Parser [Chunk]
skipLine =
fmap (pure . CSkip . build) $
(\s -> B.fromString (s ++ "\n")) <$> manyTill anyChar newline
-- Parses a function definition on the current line.
-- Assumes the cursor is at the start of a line, and ends at the start of a line again.
-- Fails if no function definition can be found on this line.
pFunDef :: Parser [Chunk]
pFunDef = try $ do
prefix <- build <$> combine
[horSpaces
,option mempty (text "=" +++ horSpaces)
,option mempty (text "let" +++ horSpaces1)
,option mempty (text "where" +++ horSpaces1)]
let indent = T.length prefix
s1 <- combine
[pIdentifier
,horSpaces
,text "::"]
let limiter = lookAhead $ try $ do
_ <- newline
_ <- atmost (fromIntegral indent) (char ' ')
_ <- satisfy (not . isSpace)
return ()
typeText <- manyTill anyChar limiter
s2 <- B.singleton <$> newline
pat <- ask
return (CSkip (prefix `T.append` build s1) : parseTypeText pat typeText ++ [CSkip (build s2)])
pDataDef :: Parser [Chunk]
pDataDef = try $ do
s1 <- text "data"
let limiter = lookAhead $ try $
choice [eof
,do _ <- newline
_ <- lookAhead (satisfy (not . isSpace))
notFollowedBy (string "--")]
typeText <- manyTill anyChar limiter
s2 <- (eof >> return mempty) <|> (B.singleton <$> newline)
pat <- ask
return (CSkip (build s1) : parseTypeText pat typeText ++ [CSkip (build s2)])
pInstanceDef :: Parser [Chunk]
pInstanceDef = try $ do
s1 <- text "instance"
let limiter = lookAhead $ try $
choice [eof
,lookAhead $ try $ do
_ <- satisfy prewordc
_ <- string "where"
eof <|> (satisfy prewordc >> return ())]
typeText <- manyTill anyChar limiter
pat <- ask
return (CSkip (build s1) : parseTypeText pat typeText)
parseTypeText :: Pattern -> String -> [Chunk]
parseTypeText (Pattern patname _) inputText =
fromRight [CSkip (T.pack inputText)] $
parse pTypeText "" inputText
where
pTypeText :: Parser' [Chunk]
pTypeText = do
s1 <- manyTill anyChar $
try (choice [eof
,lookAhead $ try $ do
_ <- satisfy prewordc
_ <- string (T.unpack patname)
return ()])
choice [eof >> return [CSkip (T.pack s1)]
,do s2 <- B.singleton <$> satisfy prewordc
sName <- build <$> text (T.unpack patname)
-- Try to parse some arguments; if that fails wholesale, just ignore this occurrence
choice [do args <- try (many pArgument)
rest <- pTypeText
return (CSkip (build (B.fromString s1 <> s2))
:COcc (Occurrence sName args)
:rest)
,do rest <- pTypeText
return (CSkip (build (B.fromString s1 <> s2))
:CSkip sName
:rest)]]
-- Returns preceding spaces and the argument itself. Requires non-zero spacing before.
pArgument :: Parser' (Text, Text)
pArgument = try $ do
s1 <- build <$> horSpaces1
s2 <- build <$> pExprAtom
return (s1, s2)
pExprAtom :: Parser' Builder
pExprAtom = pName <|> pParens
where
pExprApp :: Parser' Builder
pExprApp = pExprAtom +++ (mconcat <$> many (try (horSpaces1 +++ pExprAtom)))
pParens :: Parser' Builder
pParens = try $
combine [text "("
,mconcat . intersperse (B.singleton ',') <$>
(horSpaces +++ pExprApp) `sepBy` char ','
,horSpaces
,text ")"]
prewordc :: Char -> Bool
prewordc c = not (isAlpha c) && not (isDigit c) && c /= '_'
pIdentifier :: Parser Builder
pIdentifier = pOperator <|> pName
where
pOperator :: Parser Builder
pOperator = try $ combine
[text "("
,do s <- many1 (satisfy (\c -> isSymbol c && c `notElem` "()"))
guard (not (all (== '-') s))
return (B.fromString s)
,text ")"]
pName :: Monad m => ParsecT String u m Builder
pName = try $ do
c1 <- satisfy (\c -> isAlpha c || c == '_')
cs <- many $ satisfy (\c -> isAlpha c || isDigit c || c `elem` "_'")
guard (not (reservedName (c1 : cs)))
return (B.fromString (c1 : cs))
where
reservedName :: String -> Bool
reservedName s = s `elem`
["case", "class", "data", "default", "deriving", "do", "else"
,"foreign", "if", "import", "in", "infix", "infixl", "infixr"
,"instance", "let", "module", "newtype", "of", "then", "type"
,"where", "_"]
horSpaces :: Monad m => ParsecT String u m Builder
horSpaces = mconcat . map B.singleton <$> many (satisfy (== ' '))
horSpaces1 :: Monad m => ParsecT String u m Builder
horSpaces1 = mconcat . map B.singleton <$> many1 (satisfy (== ' '))
text :: Monad m => String -> ParsecT String u m Builder
text s = B.fromString <$> string s
atmost :: Int -> ParsecT s u m a -> ParsecT s u m [a]
atmost 0 _ = return []
atmost n p = ((:) <$> p <*> atmost (n-1) p) <|> return []
(+++) :: ParsecT s u m Builder -> ParsecT s u m Builder -> ParsecT s u m Builder
p +++ q = (<>) <$> p <*> q
combine :: (Monoid m, Applicative f) => [f m] -> f m
combine = fmap mconcat . sequenceA
build :: Builder -> Text
build = B.toLazyText
|
