module Text.Highlighting.Kate.Syntax.Sgml
(highlight, parseExpression, syntaxName, syntaxExtensions)
where
import Text.Highlighting.Kate.Types
import Text.Highlighting.Kate.Common
import Text.ParserCombinators.Parsec hiding (State)
import Data.Map (fromList)
import Control.Monad.State
import Data.Char (isSpace)
import Data.Maybe (fromMaybe)
syntaxName :: String
syntaxName = "SGML"
syntaxExtensions :: String
syntaxExtensions = "*.sgml"
highlight :: String -> [SourceLine]
highlight input = evalState (mapM parseSourceLine $ lines input) startingState
parseSourceLine :: String -> State SyntaxState SourceLine
parseSourceLine = mkParseSourceLine parseExpressionInternal pEndLine
parseExpression :: KateParser Token
parseExpression = do
st <- getState
let oldLang = synStLanguage st
setState $ st { synStLanguage = "SGML" }
context <- currentContext <|> (pushContext "Normal Text" >> currentContext)
result <- parseRules context
optional $ eof >> pEndLine
updateState $ \st -> st { synStLanguage = oldLang }
return result
startingState = SyntaxState {synStContexts = fromList [("SGML",["Normal Text"])], synStLanguage = "SGML", synStLineNumber = 0, synStPrevChar = '\n', synStPrevNonspace = False, synStCaseSensitive = True, synStKeywordCaseSensitive = False, synStCaptures = []}
pEndLine = do
updateState $ \st -> st{ synStPrevNonspace = False }
context <- currentContext
case context of
"Normal Text" -> return ()
"Attribute" -> return ()
"Value" -> return ()
"Value 2" -> return ()
"Comment" -> return ()
_ -> return ()
withAttribute attr txt = do
when (null txt) $ fail "Parser matched no text"
updateState $ \st -> st { synStPrevChar = last txt
, synStPrevNonspace = synStPrevNonspace st || not (all isSpace txt) }
return (attr, txt)
parseExpressionInternal = do
context <- currentContext
parseRules context <|> (pDefault >>= withAttribute (fromMaybe NormalTok $ 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",NormalTok),("Attribute",OtherTok),("Value",DataTypeTok),("Value 2",DataTypeTok),("Comment",CommentTok)]
parseRules "Normal Text" =
(((pString False "<!--" >>= withAttribute CommentTok) >>~ 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 KeywordTok) >>~ pushContext "Attribute"))
parseRules "Attribute" =
(((pDetect2Chars False '/' '>' >>= withAttribute KeywordTok) >>~ (popContext))
<|>
((pDetectChar False '>' >>= withAttribute KeywordTok) >>~ (popContext))
<|>
((pRegExpr regex_'5cs'2a'3d'5cs'2a >>= withAttribute NormalTok) >>~ pushContext "Value"))
parseRules "Value" =
(((pDetect2Chars False '/' '>' >>= withAttribute KeywordTok) >>~ (popContext >> popContext))
<|>
((pDetectChar False '>' >>= withAttribute KeywordTok) >>~ (popContext >> popContext))
<|>
((pDetectChar False '"' >>= withAttribute DataTypeTok) >>~ pushContext "Value 2"))
parseRules "Value 2" =
((pDetectChar False '"' >>= withAttribute DataTypeTok) >>~ (popContext >> popContext))
parseRules "Comment" =
((pString False "-->" >>= withAttribute CommentTok) >>~ (popContext))
parseRules "" = parseRules "Normal Text"
parseRules x = fail $ "Unknown context" ++ x