module Text.Highlighting.Kate.Syntax.Sgml ( highlight, parseExpression, syntaxName, syntaxExtensions ) where
import Text.Highlighting.Kate.Definitions
import Text.Highlighting.Kate.Common
import Text.ParserCombinators.Parsec
import Control.Monad (when)
import Data.Map (fromList)
import Data.Maybe (fromMaybe, maybeToList)
syntaxName :: String
syntaxName = "SGML"
syntaxExtensions :: String
syntaxExtensions = "*.sgml"
highlight :: String -> Either String [SourceLine]
highlight input =
case runParser parseSource startingState "source" input of
Left err -> Left $ show err
Right result -> Right result
parseExpression :: GenParser Char SyntaxState LabeledSource
parseExpression = do
st <- getState
let oldLang = synStLanguage st
setState $ st { synStLanguage = "SGML" }
context <- currentContext <|> (pushContext "Normal Text" >> currentContext)
result <- parseRules context
updateState $ \st -> st { synStLanguage = oldLang }
return result
parseSource = do
lineContents <- lookAhead wholeLine
updateState $ \st -> st { synStCurrentLine = lineContents }
result <- manyTill parseSourceLine eof
return $ map normalizeHighlighting result
startingState = SyntaxState {synStContexts = fromList [("SGML",["Normal Text"])], synStLanguage = "SGML", synStCurrentLine = "", synStCharsParsedInLine = 0, synStPrevChar = '\n', synStCaseSensitive = True, synStKeywordCaseSensitive = False, synStCaptures = []}
parseSourceLine = manyTill parseExpressionInternal pEndLine
pEndLine = do
lookAhead $ newline <|> (eof >> return '\n')
context <- currentContext
case context of
"Normal Text" -> return () >> pHandleEndLine
"Attribute" -> return () >> pHandleEndLine
"Value" -> return () >> pHandleEndLine
"Value 2" -> return () >> pHandleEndLine
"Comment" -> return () >> pHandleEndLine
_ -> pHandleEndLine
withAttribute attr txt = do
when (null txt) $ fail "Parser matched no text"
let labs = attr : maybeToList (lookup attr styles)
st <- getState
let oldCharsParsed = synStCharsParsedInLine st
let prevchar = if null txt then '\n' else last txt
updateState $ \st -> st { synStCharsParsedInLine = oldCharsParsed + length txt, synStPrevChar = prevchar }
return (labs, txt)
styles = [("Tag","kw"),("Attribute Name","ot"),("Attribute Value","dt"),("Comment","co")]
parseExpressionInternal = do
context <- currentContext
parseRules context <|> (pDefault >>= withAttribute (fromMaybe "" $ lookup context defaultAttributes))
regex_'3c'5cs'2a'5c'2f'3f'5cs'2a'5ba'2dzA'2dZ'5f'3a'5d'5ba'2dzA'2dZ0'2d9'2e'5f'3a'2d'5d'2a = compileRegex "<\\s*\\/?\\s*[a-zA-Z_:][a-zA-Z0-9._:-]*"
regex_'5cs'2a'3d'5cs'2a = compileRegex "\\s*=\\s*"
defaultAttributes = [("Normal Text","Normal Text"),("Attribute","Attribute Name"),("Value","Attribute Value"),("Value 2","Attribute Value"),("Comment","Comment")]
parseRules "Normal Text" =
do (attr, result) <- (((pString False "<!--" >>= withAttribute "Comment") >>~ pushContext "Comment")
<|>
((pRegExpr regex_'3c'5cs'2a'5c'2f'3f'5cs'2a'5ba'2dzA'2dZ'5f'3a'5d'5ba'2dzA'2dZ0'2d9'2e'5f'3a'2d'5d'2a >>= withAttribute "Tag") >>~ pushContext "Attribute"))
return (attr, result)
parseRules "Attribute" =
do (attr, result) <- (((pDetect2Chars False '/' '>' >>= withAttribute "Tag") >>~ (popContext))
<|>
((pDetectChar False '>' >>= withAttribute "Tag") >>~ (popContext))
<|>
((pRegExpr regex_'5cs'2a'3d'5cs'2a >>= withAttribute "Normal Text") >>~ pushContext "Value"))
return (attr, result)
parseRules "Value" =
do (attr, result) <- (((pDetect2Chars False '/' '>' >>= withAttribute "Tag") >>~ (popContext >> popContext))
<|>
((pDetectChar False '>' >>= withAttribute "Tag") >>~ (popContext >> popContext))
<|>
((pDetectChar False '"' >>= withAttribute "Attribute Value") >>~ pushContext "Value 2"))
return (attr, result)
parseRules "Value 2" =
do (attr, result) <- ((pDetectChar False '"' >>= withAttribute "Attribute Value") >>~ (popContext >> popContext))
return (attr, result)
parseRules "Comment" =
do (attr, result) <- ((pString False "-->" >>= withAttribute "Comment") >>~ (popContext))
return (attr, result)
parseRules x = fail $ "Unknown context" ++ x