diff options
Diffstat (limited to 'haddock-library/src/Documentation/Haddock/Parser.hs')
-rw-r--r-- | haddock-library/src/Documentation/Haddock/Parser.hs | 26 |
1 files changed, 20 insertions, 6 deletions
diff --git a/haddock-library/src/Documentation/Haddock/Parser.hs b/haddock-library/src/Documentation/Haddock/Parser.hs index de336d45..19c92721 100644 --- a/haddock-library/src/Documentation/Haddock/Parser.hs +++ b/haddock-library/src/Documentation/Haddock/Parser.hs @@ -78,7 +78,7 @@ overIdentifier f d = g d g (DocMonospaced x) = DocMonospaced $ g x g (DocBold x) = DocBold $ g x g (DocUnorderedList x) = DocUnorderedList $ fmap g x - g (DocOrderedList x) = DocOrderedList $ fmap g x + g (DocOrderedList x) = DocOrderedList $ fmap (\(index, a) -> (index, g a)) x g (DocDefList x) = DocDefList $ fmap (\(y, z) -> (g y, g z)) x g (DocCodeBlock x) = DocCodeBlock $ g x g (DocHyperlink (Hyperlink u x)) = DocHyperlink (Hyperlink u (fmap g x)) @@ -173,11 +173,11 @@ encodedChar = "&#" *> c <* ";" -- Once we have checked for any of these and tried to parse the -- relevant markup, we can assume they are used as regular text. specialChar :: [Char] -specialChar = "_/<@\"&'`# " +specialChar = "_/<@\"&'`#[ " -- | Plain, regular parser for text. Called as one of the last parsers -- to ensure that we have already given a chance to more meaningful parsers --- before capturing their characers. +-- before capturing their characters. string' :: Parser (DocH mod a) string' = DocString . unescape . T.unpack <$> takeWhile1_ (`notElem` specialChar) where @@ -361,7 +361,7 @@ table = do firstRow <- parseFirstRow let len = T.length firstRow - -- then we parse all consequtive rows starting and ending with + or |, + -- then we parse all consecutive rows starting and ending with + or |, -- of the width `len`. restRows <- many (try (parseRestRows len)) @@ -577,10 +577,24 @@ unorderedList indent = DocUnorderedList <$> p orderedList :: Text -> Parser (DocH mod Identifier) orderedList indent = DocOrderedList <$> p where - p = (paren <|> dot) *> innerList indent p + p = do + index <- paren <|> dot + innerList' indent p index dot = (decimal :: Parser Int) <* "." paren = "(" *> decimal <* ")" +-- | Like 'innerList' but takes the parsed index of the list item +innerList' :: Text -> Parser [(Int, DocH mod Identifier)] + -> Int + -> Parser [(Int, DocH mod Identifier)] +innerList' indent item index = do + c <- takeLine + (cs, items) <- more indent item + let contents = docParagraph . parseText . dropNLs . T.unlines $ c : cs + return $ case items of + Left p -> [(index, contents `docAppend` p)] + Right i -> (index, contents) : i + -- | Generic function collecting any further lines belonging to the -- list entry and recursively collecting any further lists in the -- same paragraph. Usually used as @@ -710,7 +724,7 @@ stripSpace = fromMaybe <*> mapM strip' Just (' ',t') -> Just t' _ -> Nothing --- | Parses examples. Examples are a paragraph level entitity (separated by an empty line). +-- | Parses examples. Examples are a paragraph level entity (separated by an empty line). -- Consecutive examples are accepted. examples :: Parser (DocH mod a) examples = DocExamples <$> (many (try (skipHorizontalSpace *> "\n")) *> go) |