module Text.Highlighting.Kate.Syntax.Bibtex ( 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)
import qualified Data.Set as Set
syntaxName :: String
syntaxName = "BibTeX"
syntaxExtensions :: String
syntaxExtensions = "*.bib"
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 = "BibTeX" }
context <- currentContext <|> (pushContext "Normal" >> 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 [("BibTeX",["Normal"])], synStLanguage = "BibTeX", synStCurrentLine = "", synStCharsParsedInLine = 0, synStPrevChar = '\n', synStCaseSensitive = True, synStKeywordCaseSensitive = False, synStCaptures = []}
parseSourceLine = manyTill parseExpressionInternal pEndLine
pEndLine = do
newline <|> (eof >> return '\n')
context <- currentContext
case context of
"Normal" -> return ()
"Entry" -> return ()
"String" -> return ()
_ -> return ()
lineContents <- lookAhead wholeLine
updateState $ \st -> st { synStCurrentLine = lineContents, synStCharsParsedInLine = 0, synStPrevChar = '\n' }
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 = [("Entry","kw"),("Command","fu"),("Field","dt"),("Ref Key","ot"),("String","st"),("Char","ch")]
parseExpressionInternal = do
context <- currentContext
parseRules context <|> (pDefault >>= withAttribute (fromMaybe "" $ lookup context defaultAttributes))
list_kw'5fentry = Set.fromList $ words $ "@article @book @booklet @conference @inbook @incollection @inproceedings @manual @mastersthesis @misc @phdthesis @proceedings @techreport @unpublished @collection @patent"
list_kw'5fcommand = Set.fromList $ words $ "@string @preamble @comment"
regex_'28'5ba'2dzA'2dZ'5d'2b'29'5cs'2a'3d = compileRegex "([a-zA-Z]+)\\s*="
regex_'5c'5c'28'5ba'2dzA'2dZ'5d'2b'7c'2e'29 = compileRegex "\\\\([a-zA-Z]+|.)"
defaultAttributes = [("Normal","Normal Text"),("Entry","Ref Key"),("String","String")]
parseRules "Normal" =
do (attr, result) <- (((pFirstNonSpace >> pRegExpr regex_'28'5ba'2dzA'2dZ'5d'2b'29'5cs'2a'3d >>= withAttribute "Field"))
<|>
((pKeyword " \n\t.():!+,-<=>%&*/;?[]^{|}~" list_kw'5fentry >>= withAttribute "Entry") >>~ pushContext "Entry")
<|>
((pKeyword " \n\t.():!+,-<=>%&*/;?[]^{|}~" list_kw'5fcommand >>= withAttribute "Command"))
<|>
((pDetectChar False '{' >>= withAttribute "Normal Text"))
<|>
((pDetectChar False '}' >>= withAttribute "Normal Text"))
<|>
((pRegExpr regex_'5c'5c'28'5ba'2dzA'2dZ'5d'2b'7c'2e'29 >>= withAttribute "Char"))
<|>
((pDetectChar False '"' >>= withAttribute "String") >>~ pushContext "String"))
return (attr, result)
parseRules "Entry" =
do (attr, result) <- (((pDetectChar False ',' >>= withAttribute "Normal Text") >>~ (popContext >> return ()))
<|>
((pDetectChar False '{' >>= withAttribute "Normal Text"))
<|>
((pDetectChar False '}' >>= withAttribute "Normal Text"))
<|>
((pRegExpr regex_'5c'5c'28'5ba'2dzA'2dZ'5d'2b'7c'2e'29 >>= withAttribute "Char"))
<|>
((pDetectChar False '"' >>= withAttribute "String") >>~ (popContext >> return ())))
return (attr, result)
parseRules "String" =
do (attr, result) <- (((pRegExpr regex_'5c'5c'28'5ba'2dzA'2dZ'5d'2b'7c'2e'29 >>= withAttribute "Char"))
<|>
((pDetectChar False '"' >>= withAttribute "String") >>~ (popContext >> return ())))
return (attr, result)
parseRules x = fail $ "Unknown context" ++ x