1bfe1b8 · typeable/pandoc

git.iko.soy/typeable/pandoc

Added support for Cite to Markdown reader, and conditional support for citeproc module.

+ The citeproc cabal configuration option sets the _CITEPROC macro, which conditionally includes code for handling citations. + Added Text.Pandoc.Biblio module. + Made highlighting option default to False. git-svn-id: https://pandoc.googlecode.com/svn/trunk@1376 788f1e2b-df1e-0410-8736-df70ead52e1b

commit1bfe1b84a8692c5e2ea8a036208f61612b48e9fb authorfiddlosopher <fiddlosopher@788f1e2b-df1e-0410-8736-df70ead52e1b> dateMon 4 Aug 2008, 03:15 (18 years ago) parent4719c78

7 files changed+197−16

compare with master

unifiedsplit

Main.hs+58−13 Text/Pandoc/Biblio.hs+76 Text/Pandoc/Definition.hs+2 Text/Pandoc/Readers/Markdown.hs+41−2 Text/Pandoc/Shared.hs+6 Text/Pandoc/Writers/MediaWiki.hs+2 pandoc.cabal+12−1

Main.hs+58−13

@@ -47,6 +47,10 @@ import System.IO ( stdout, stderr )

4747#else

4848import System.IO

4949#endif

50+#ifdef _CITEPROC

51+import Text.CSL

52+import Text.Pandoc.Biblio

53+#endif

5054

5155copyrightMessage :: String

5559 "warranty, not even for merchantability or fitness for a particular purpose."

5660

5761compileInfo :: String

58−compileInfo = "Compiled" ++

62+compileInfo =

5963#ifdef _UTF8

60− " with UTF-8 support" ++

64+ " +utf8" ++

65+#else

66+ " -utf8" ++

67+#endif

68+#ifdef _CITEPROC

69+ " +citeproc" ++

70+#else

71+ " -citeproc" ++

72+#endif

73+#ifdef _HIGHLIGHTING

74+ " +highlighting" ++

6175#else

62− " without UTF-8 support" ++

76+ " -highlighting" ++

6377#endif

6478 if null languages

65− then " and without syntax highlighting support."

66− else " and with syntax highlighting support for:\n" ++

79+ then "\n"

80+ else "\nCompiled with syntax highlighting support for:\n" ++

6781 (unlines $ map unwords $ chunk 5 $ map (\s -> s ++ replicate (15 - length s) ' ') languages)

6882

6983-- | Splits a list into groups of at most n.

@@ -137,6 +151,10 @@ data Opt = Opt

137151 , optReferenceLinks :: Bool -- ^ Use reference links in writing markdown, rst

138152 , optWrapText :: Bool -- ^ Wrap text

139153 , optSanitizeHTML :: Bool -- ^ Sanitize HTML

154+#ifdef _CITEPROC

155+ , optModsFile :: String

156+ , optCslFile :: String

157+#endif

140158 }

141159

142160-- | Defaults for command-line options.

@@ -166,6 +184,10 @@ defaultOpts = Opt

166184 , optReferenceLinks = False

167185 , optWrapText = True

168186 , optSanitizeHTML = False

187+#ifdef _CITEPROC

188+ , optModsFile = []

189+ , optCslFile = []

190+#endif

169191 }

170192

171193-- | A list of functions, each transforming the options data structure

@@ -333,13 +355,24 @@ options =

333355 exitWith ExitSuccess)

334356 "FORMAT")

335357 "" -- "Print default header for FORMAT"

336−

358+#ifdef _CITEPROC

359+ , Option "" ["mods"]

360+ (ReqArg

361+ (\arg opt -> return opt { optModsFile = arg} )

362+ "FILENAME")

363+ ""

364+ , Option "" ["csl"]

365+ (ReqArg

366+ (\arg opt -> return opt { optCslFile = arg} )

367+ "FILENAME")

368+ ""

369+#endif

337370 , Option "" ["dump-args"]

338371 (NoArg

339372 (\opt -> return opt { optDumpArgs = True }))

340373 "" -- "Print output filename and arguments to stdout."

341374

342− , Option "" ["ignore-args"]

375+ , Option "" ["ignore-args"]

343376 (NoArg

344377 (\opt -> return opt { optIgnoreArgs = True }))

345378 "" -- "Ignore command-line arguments."

@@ -348,7 +381,7 @@ options =

348381 (NoArg

349382 (\_ -> do

350383 prg <- getProgName

351− hPutStrLn stderr (prg ++ " " ++ pandocVersion ++ "\n" ++ compileInfo ++

384+ hPutStrLn stderr (prg ++ " " ++ pandocVersion ++ compileInfo ++

352385 copyrightMessage)

353386 exitWith $ ExitFailure 4))

354387 "" -- "Print version"

@@ -464,6 +497,10 @@ main = do

464497 , optReferenceLinks = referenceLinks

465498 , optWrapText = wrap

466499 , optSanitizeHTML = sanitize

500+#ifdef _CITEPROC

501+ , optModsFile = modsFile

502+ , optCslFile = cslFile

503+#endif

467504 } = opts

468505

469506 if dumpArgs

@@ -513,11 +550,18 @@ main = do

513550

514551 let standalone' = (standalone && not strict) || writerName' == "odt"

515552

553+#ifdef _CITEPROC

554+ refs <- if null modsFile then return [] else readModsColletionFile modsFile

555+#endif

556+

516557 let startParserState =

517558 defaultParserState { stateParseRaw = parseRaw,

518559 stateTabStop = tabStop,

519560 stateSanitizeHTML = sanitize,

520561 stateStandalone = standalone',

562+#ifdef _CITEPROC

563+ stateCitations = map citeKey refs,

564+#endif

521565 stateSmart = smart || writerName' `elem`

522566 ["latex", "context"],

523567 stateColumns = columns,

@@ -564,11 +608,12 @@ main = do

564608 then putStrLn

565609 else writeFile outputFile . (++ "\n")

566610

567− (readSources sources) >>= writeOutput .

568− writer writerOptions .

569− reader startParserState .

570− tabFilter tabStop .

571− joinWithSep "\n"

611+ fmap (reader startParserState . tabFilter tabStop . joinWithSep "\n")

612+ (readSources sources) >>=

613+#ifdef _CITEPROC

614+ processBiblio cslFile refs >>=

615+#endif

616+ writeOutput . writer writerOptions

572617

573618 where

574619 readSources [] = mapM readSource ["-"]

Text/Pandoc/Biblio.hsnew file+76

@@ -0,0 +1,76 @@

1+{-# LANGUAGE PatternGuards, CPP #-}

2+{-

5+This program is free software; you can redistribute it and/or modify

6+it under the terms of the GNU General Public License as published by

7+the Free Software Foundation; either version 2 of the License, or

8+(at your option) any later version.

10+This program is distributed in the hope that it will be useful,

11+but WITHOUT ANY WARRANTY; without even the implied warranty of

12+MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE. See the

13+GNU General Public License for more details.

14+

15+You should have received a copy of the GNU General Public License

16+along with this program; if not, write to the Free Software

17+Foundation, Inc., 59 Temple Place, Suite 330, Boston, MA 02111-1307 USA

18+-}

19+

20+{- |

21+ Module : Text.Pandoc.Biblio

23+ License : GNU GPL, version 2 or above

24+

25+ Maintainer : Andrea Rossato <andrea.rossato@ing.unitn.it>

26+ Stability : alpha

27+ Portability : portable

28+-}

29+

30+#ifdef _CITEPROC

31+module Text.Pandoc.Biblio ( processBiblio ) where

32+

33+import Control.Monad ( when )

34+import Data.List

35+import Text.CSL

36+import Text.Pandoc.Definition

37+

38+#else

39+module Text.Pandoc.Biblio () where

40+#endif

41+

42+#ifdef _CITEPROC

43+processBiblio :: String -> [Reference] -> Pandoc -> IO Pandoc

44+processBiblio cf r p

45+ = if null r then return p

46+ else do

47+ when (null cf) $ error "Missing the needed citation style file"

48+ csl <- readCSLFile cf

49+ let groups = queryPandoc getCite p

50+ citations = zip groups . processCitations csl r $ groups

51+ Pandoc m b = processPandoc (processCite citations) p

52+ return $ Pandoc m $ b ++ renderBiblio csl r p

53+

54+renderBiblio :: Style -> [Reference] -> Pandoc -> [Block]

55+renderBiblio s r p

56+ = map (read . renderPandoc') $ processBibliography s refs

57+ where cits = nub . map fst . concat . queryPandoc getCite $ p

58+ refs = getRefs r $ zip cits (repeat "")

59+

60+processCite :: [([Target],[FormattedOutput])] -> Inline -> Inline

61+processCite cs il

62+ | Cite t _ <- il = Cite t (process t)

63+ | otherwise = il

64+ where

65+ process t = case elemIndex t (map fst cs) of

66+ Just i -> read . renderPandoc $ snd (cs !! i)

67+ Nothing -> [Str ("Error processing " ++ show t)]

68+

69+getCite :: Inline -> [[(String,String)]]

70+getCite i | Cite t _ <- i = [t]

71+ | otherwise = []

72+

73+getRefs :: [Reference] -> [Target] -> [Reference]

74+getRefs r = map $ getReference r

75+

76+#endif

Text/Pandoc/Definition.hs+2

@@ -125,8 +125,10 @@ data Inline

125125 | Note [Block] -- ^ Footnote or endnote

126126 deriving (Show, Eq, Read, Typeable, Data)

127127

128+-- | Applies a transformation to matching elements in a Pandoc document.

128129processPandoc :: Typeable a => (a -> a) -> Pandoc -> Pandoc

129130processPandoc f = everywhere (mkT f)

130131

132+-- | Runs a query on matching elements in a Pandoc document.

131133queryPandoc :: Typeable a => (a -> [b]) -> Pandoc -> [b]

132134queryPandoc f = everything (++) ([] `mkQ` f)

Text/Pandoc/Readers/Markdown.hs+41−2

@@ -1,3 +1,4 @@

1+{-# LANGUAGE CPP #-}

12{-

@@ -34,7 +35,7 @@ module Text.Pandoc.Readers.Markdown (

3435import Data.List ( transpose, isPrefixOf, isSuffixOf, lookup, sortBy, findIndex )

3536import Data.Ord ( comparing )

3637import Data.Char ( isAlphaNum, isAlpha, isLower, isDigit )

37−import Data.Maybe ( fromMaybe )

38+import Data.Maybe

3839import Text.Pandoc.Definition

3940import Text.Pandoc.Shared

4041import Text.Pandoc.Readers.LaTeX ( rawLaTeXInline, rawLaTeXEnvironment )

@@ -173,7 +174,7 @@ parseMarkdown = do

173174 setPosition startPos

174175 -- now parse it for real...

175176 (title, author, date) <- option ([],[],"") titleBlock

176− blocks <- parseBlocks

177+ blocks <- parseBlocks

177178 return $ Pandoc (Meta title author date) $ filter (/= Null) blocks

178179

179180--

@@ -804,6 +805,9 @@ inlineParsers = [ abbrev

804805 , note

805806 , inlineNote

806807 , link

808+#ifdef _CITEPROC

809+ , inlineCitation

810+#endif

807811 , image

808812 , math

809813 , strikeout

@@ -1152,3 +1156,38 @@ rawHtmlInline' = do

11521156 else anyHtmlInlineTag

11531157 return $ HtmlInline result

11541158

1159+#ifdef _CITEPROC

1160+inlineCitation :: GenParser Char ParserState Inline

1161+inlineCitation = try $ do

1162+ failIfStrict

1163+ cit <- citeMarker

1164+ let citations = readWith parseCitation defaultParserState cit

1165+ mr <- mapM chkCit citations

1166+ if catMaybes mr /= []

1167+ then return $ Cite citations []

1168+ else fail "no citation found"

1169+

1170+chkCit :: Target -> GenParser Char ParserState (Maybe Target)

1171+chkCit t = do

1172+ st <- getState

1173+ case lookupKeySrc (stateKeys st) [Str $ fst t] of

1174+ Just _ -> fail "This is a link"

1175+ Nothing -> if elem (fst t) $ stateCitations st

1176+ then return $ Just t

1177+ else return $ Nothing

1178+

1179+citeMarker :: GenParser Char ParserState String

1180+citeMarker = string "[" >> manyTill (noneOf "\t\n") (string "]")

1181+

1182+parseCitation :: GenParser Char ParserState [(String,String)]

1183+parseCitation = try $ sepBy (parseLabel) (oneOf ";")

1184+

1185+parseLabel :: GenParser Char ParserState (String,String)

1186+parseLabel = try $ do

1187+ res <- sepBy (skipSpaces >> many1 (noneOf "@;\n\t")) (oneOf "@")

1188+ case res of

1189+ [lab,loc] -> return (lab, loc)

1190+ [lab] -> return (lab, "" )

1191+ _ -> return ("" , "" )

1192+

1193+#endif

Text/Pandoc/Shared.hs+6

@@ -629,6 +629,9 @@ data ParserState = ParserState

629629 stateQuoteContext :: QuoteContext, -- ^ Inside quoted environment?

630630 stateSanitizeHTML :: Bool, -- ^ Sanitize HTML?

631631 stateKeys :: KeyTable, -- ^ List of reference keys

632+#ifdef _CITEPROC

633+ stateCitations :: [String], -- ^ List of available citations

634+#endif

632635 stateNotes :: NoteTable, -- ^ List of notes

633636 stateTabStop :: Int, -- ^ Tab stop

634637 stateStandalone :: Bool, -- ^ Parse bibliographic info?

@@ -649,6 +652,9 @@ defaultParserState =

649652 stateQuoteContext = NoQuote,

650653 stateSanitizeHTML = False,

651654 stateKeys = [],

655+#ifdef _CITEPROC

656+ stateCitations = [],

657+#endif

652658 stateNotes = [],

653659 stateTabStop = 4,

654660 stateStandalone = False,

Text/Pandoc/Writers/MediaWiki.hs+2

@@ -342,6 +342,8 @@ inlineToMediaWiki opts (Quoted DoubleQuote lst) = do

342342 contents <- inlineListToMediaWiki opts lst

343343 return $ "“" ++ contents ++ "”"

344344

345+inlineToMediaWiki opts (Cite _ lst) = inlineListToMediaWiki opts lst

346+

345347inlineToMediaWiki _ EmDash = return "—"

346348

347349inlineToMediaWiki _ EnDash = return "–"

pandoc.cabal+12−1

@@ -57,7 +57,7 @@ Flag splitBase

5757 Default: True

5858Flag highlighting

5959 Description: Compile in support for syntax highlighting of code blocks.

60− Default: True

60+ Default: False

6161Flag executable

6262 Description: Build the pandoc executable.

6363 Default: True

@@ -67,6 +67,9 @@ Flag library

6767Flag utf8

6868 Description: Compile in support for UTF-8 input and output.

6969 Default: True

70+Flag citeproc

71+ Description: Compile in support for citeproc-hs bibliographic formatting.

72+ Default: False

7073

7174Library

7275 if flag(splitBase)

@@ -79,10 +82,14 @@ Library

7982 if flag(utf8)

8083 Build-depends: utf8-string

8184 cpp-options: -D_UTF8

85+ if flag(citeproc)

86+ Build-depends: citeproc-hs

87+ cpp-options: -D_CITEPROC

8288 Build-Depends: parsec < 3, xhtml, mtl, network, filepath, process, directory,

8389 template-haskell, bytestring

8490 Hs-Source-Dirs: .

8591 Exposed-Modules: Text.Pandoc,

92+ Text.Pandoc.Biblio,

8693 Text.Pandoc.Blocks,

8794 Text.Pandoc.Definition,

8895 Text.Pandoc.CharacterReferences,

@@ -136,7 +143,11 @@ Executable pandoc

136143 cpp-options: -D_HIGHLIGHTING

137144 if flag(utf8)

138145 cpp-options: -D_UTF8

146+ if flag(citeproc)

147+ Build-depends: citeproc-hs

148+ cpp-options: -D_CITEPROC

139149 if flag(executable)

140150 Buildable: True

141151 else

142152 Buildable: False

153+