23ab5ff · typeable/pandoc

git.iko.soy/typeable/pandoc

Implemented Ext_header_attributes.

This allows explicit attributes to be put on headers, using a syntax like that for code blocks: {#id .class .class k=v k=v}

commit23ab5ffcb0b708e4f07cafd7a877926ce0dc532b authorJohn MacFarlane <fiddlosopher@gmail.com> dateThu 10 Jan 2013, 20:22 (14 years ago) parent1a7686a

3 files changed+46−15

compare with master

unifiedsplit

README+24−3 src/Text/Pandoc/Options.hs+2 src/Text/Pandoc/Readers/Markdown.hs+20−12

README+24−3

@@ -872,10 +872,31 @@ wrapping). Consider, for example:

872872

873873### Header identifiers in HTML, LaTeX, and ConTeXt ###

874874

875−**Extension: `header_identifiers`**

875+**Extension: `header_attributes`**

876876

877−Each header element in pandoc's HTML and ConTeXt output is given a

878−unique identifier. This identifier is based on the text of the header.

877+Headers can be assigned attributes using this syntax at the end

878+of the line containing the header text:

879+

880+ {#identifier .class .class key=value key=value}

881+

882+Although this syntax allows assignment of classes and key/value attributes,

883+only identifiers currently have any affect in the writers (and only in some

884+writers: HTML, LaTeX, ConTeXt, Textile, AsciiDoc). Thus, for example,

885+the following headers will all be assigned the identifier `foo`:

886+

887+ # My header {#foo}

888+

889+ ## My header ## {#foo}

890+

891+ My other header {#foo}

892+ ---------------

893+

894+(This syntax is compatible with [PHP Markdown Extra].)

895+

896+**Extension: `auto_identifiers`**

897+

898+A header without an explicitly specified identifier will be

899+automatically assigned a unique identifier based on the header text.

879900To derive the identifier from the header text,

880901

881902 - Remove all formatting, links, etc.

src/Text/Pandoc/Options.hs+2

@@ -91,6 +91,7 @@ data Extension =

9191 | Ext_literate_haskell -- ^ Enable literate Haskell conventions

9292 | Ext_abbreviations -- ^ PHP markdown extra abbreviation definitions

9393 | Ext_auto_identifiers -- ^ Automatic identifiers for headers

94+ | Ext_header_attributes -- ^ Explicit header attributes {#id .class k=v}

9495 | Ext_implicit_header_references -- ^ Implicit reference links for headers

9596 deriving (Show, Read, Enum, Eq, Ord, Bounded)

9697

@@ -129,6 +130,7 @@ pandocExtensions = Set.fromList

129130 , Ext_superscript

130131 , Ext_subscript

131132 , Ext_auto_identifiers

133+ , Ext_header_attributes

132134 , Ext_implicit_header_references

133135 ]

134136

src/Text/Pandoc/Readers/Markdown.hs+20−12

@@ -373,40 +373,48 @@ header :: MarkdownParser (F Blocks)

373373header = setextHeader <|> atxHeader <?> "header"

374374

375375-- returns unique identifier

376−addToHeaderList :: F Inlines -> MarkdownParser String

377−addToHeaderList text = do

376+addToHeaderList :: Attr -> F Inlines -> MarkdownParser Attr

377+addToHeaderList (ident,classes,kvs) text = do

378378 let headerList = B.toList $ runF text defaultParserState

379379 updateState $ \st -> st{ stateHeaders = headerList : stateHeaders st }

380380 (do guardEnabled Ext_auto_identifiers

381381 ids <- stateIdentifiers `fmap` getState

382− let id' = uniqueIdent headerList ids

382+ let id' = if null ident

383+ then uniqueIdent headerList ids

384+ else ident

383385 updateState $ \st -> st{ stateIdentifiers = id' : ids }

384− return id') <|> return ""

386+ return (id',classes,kvs)) <|> return ("",classes,kvs)

385387

386388atxHeader :: MarkdownParser (F Blocks)

387389atxHeader = try $ do

388390 level <- many1 (char '#') >>= return . length

389391 notFollowedBy (char '.' <|> char ')') -- this would be a list

390392 skipSpaces

391− text <- trimInlinesF . mconcat <$> manyTill inline atxClosing

392− id' <- addToHeaderList text

393− return $ B.headerWith (id',[],[]) level <$> text

393+ text <- trimInlinesF . mconcat <$> many (notFollowedBy atxClosing >> inline)

394+ attr <- atxClosing

395+ attr' <- addToHeaderList attr text

396+ return $ B.headerWith attr' level <$> text

394397

395−atxClosing :: Parser [Char] st String

396−atxClosing = try $ skipMany (char '#') >> blanklines

398+atxClosing :: Parser [Char] st Attr

399+atxClosing =

400+ try $ skipMany (char '#') >> skipSpaces >> option nullAttr attributes <* blanklines

401+

402+setextHeaderEnd :: MarkdownParser Attr

403+setextHeaderEnd = try $ option nullAttr attributes <* blankline

397404

398405setextHeader :: MarkdownParser (F Blocks)

399406setextHeader = try $ do

400407 -- This lookahead prevents us from wasting time parsing Inlines

401408 -- unless necessary -- it gives a significant performance boost.

402409 lookAhead $ anyLine >> many1 (oneOf setextHChars) >> blankline

403− text <- trimInlinesF . mconcat <$> many1Till inline newline

410+ text <- trimInlinesF . mconcat <$> many1 (notFollowedBy setextHeaderEnd >> inline)

411+ attr <- setextHeaderEnd

404412 underlineChar <- oneOf setextHChars

405413 many (char underlineChar)

406414 blanklines

407415 let level = (fromMaybe 0 $ findIndex (== underlineChar) setextHChars) + 1

408− id' <- addToHeaderList text

409− return $ B.headerWith (id',[],[]) level <$> text

416+ attr' <- addToHeaderList attr text

417+ return $ B.headerWith attr' level <$> text

410418

411419--

412420-- hrule block