Added implementation and tests for subscript and superscript; fixed nested list parsing

built-in one instead

app/Main.hs

@@ -23,5 +23,8 @@ main = do
   fileContents <- case args of
     [] -> getContents
     x : _ -> readFile x
-  let res = fst $ leftmostLongestParse parseDocument fileContents
+  let res = leftmostLongestParse parseDocument fileContents
-  print res
+  let toPrint = prettyPrint res
+  case reverse toPrint of
+    '\n' : _ -> putStr toPrint
+    _ -> putStrLn toPrint

md-to-html.cabal

@@ -58,9 +58,14 @@ library
    exposed-modules: MdToHTML
    other-modules: MdToHtmlTest
    build-depends:    base ^>=4.19.1.0,
-                      HUnit
+                      HUnit,
+                      megaparsec,
+                      parser-combinators,
+                      text,
+                      MissingH,
+                      word-wrap
 
-executable md-to-html-runner
+executable mdtoh
     -- Import common warning flags.
     import:           warnings
 

src/MdToHTML.hs

@@ -4,24 +4,31 @@
 
 module MdToHTML where
 
-import Control.Applicative
+import Control.Applicative hiding (many, some)
 import Control.Monad
+import Control.Monad.Combinators (count)
 import Data.Char
 import Data.List
 import Data.Ord (comparing)
+import Data.String.Utils
+import qualified Data.Text as T
+import Data.Void
 import Debug.Trace
-import Text.ParserCombinators.ReadP
+import Text.Megaparsec
+import Text.Megaparsec.Char
 import Text.Printf
+import Text.Wrap
+
+type Parser = Parsec Void T.Text
 
 type HeaderLevel = Int
 
+type CssClass = String
+
 newtype URL = URL {getUrl :: String} deriving (Eq)
 
 newtype ImgPath = ImgPath {getPath :: String} deriving (Eq)
 
-parseMany :: ReadP a -> ReadP [a]
-parseMany = Text.ParserCombinators.ReadP.many
-
 data MdToken
   = Document [MdToken]
   | Header HeaderLevel MdToken
@@ -34,12 +41,15 @@ data MdToken
   | UnordList [MdToken]
   | OrdList [MdToken]
   | Code MdToken
-  | Codeblock String
+  | Table [[MdToken]]
+  | Codeblock MdToken
   | Link MdToken URL
-  | Image MdToken ImgPath
+  | Image MdToken URL (Maybe [CssClass])
-  | Figure MdToken ImgPath
+  | Figure MdToken URL (Maybe [CssClass])
   | Bold MdToken
   | Italic MdToken
+  | Superscript MdToken
+  | Subscript MdToken
   | Strikethrough MdToken
   | Unit String
   deriving (Eq)
@@ -50,22 +60,37 @@ instance Show MdToken where
   show (Header level token) = "<h" ++ show level ++ ">" ++ show token ++ "</h" ++ show level ++ ">"
   show (Para token) = "<p>" ++ show token ++ "</p>"
   show (Line tokens) = concatMap show tokens
-  show Linebreak = "<br>"
+  show Linebreak = "<br />"
   show SingleNewline = " "
   show HorizontalRule = "<hr>"
   show (Blockquote tokens) = "<blockquote>" ++ concatMap show tokens ++ "</blockquote>"
   show (UnordList tokens) = "<ul>" ++ concatMap (prepend "<li>" . append "</li>" . show) tokens ++ "</ul>"
   show (OrdList tokens) = "<ol>" ++ concatMap (prepend "<li>" . append "</li>" . show) tokens ++ "</ol>"
-  show (Code code) = "<code>" ++ show code ++ "</code>"
+  show (Code code) = "<code>" ++ strip (show code) ++ "</code>"
-  show (Codeblock code) = show code
+  show (Table (thead : tokenGrid)) = "<table><thead><tr>" ++ concatMap (\x -> "<th>" ++ rstrip (show x) ++ "</th>") thead ++ "</tr></thead>" ++ "<tbody>" ++ concatMap (\x -> "<tr>" ++ concatMap (\y -> "<td>" ++ rstrip (show y) ++ "</td>") x ++ "</tr>") tokenGrid ++ "</tbody></table>"
+  show (Codeblock code) = "<pre><code>" ++ show code ++ "</code></pre>"
   show (Link txt url) = "<a href=\"" ++ getUrl url ++ "\">" ++ show txt ++ "</a>"
-  show (Image txt imgPath) = "<img src=\"" ++ getPath imgPath ++ "\"" ++ " alt=\"" ++ show txt ++ "\" />"
+  show (Image txt url cssClasses) = "<img src=\"" ++ getUrl url ++ "\"" ++ " alt=\"" ++ show txt ++ "\"" ++ maybe "" (\classes -> " class=\"" ++ unwords classes ++ "\"") cssClasses ++ "/>"
-  show (Figure txt imgPath) = "<figure><img src=\"" ++ getPath imgPath ++ "\" alt=\"" ++ show txt ++ "\"/><figcaption aria-hidden=\"true\">" ++ show txt ++ "</figcaption></figure>"
+  show (Figure txt url cssClasses) = "<figure><img src=\"" ++ getUrl url ++ "\" alt=\"" ++ show txt ++ "\"" ++ maybe "" (\classes -> " class=\"" ++ unwords classes ++ "\"") cssClasses ++ "/><figcaption aria-hidden=\"true\">" ++ show txt ++ "</figcaption></figure>"
   show (Bold token) = "<b>" ++ show token ++ "</b>"
   show (Italic token) = "<i>" ++ show token ++ "</i>"
+  show (Superscript token) = "<sup>" ++ show token ++ "</sup>"
+  show (Subscript token) = "<sub>" ++ show token ++ "</sub>"
   show (Strikethrough token) = "<s>" ++ show token ++ "</s>"
   show (Unit unit) = printf "%s" unit
 
+-- Pretty print the given token into a string.
+-- This is the same as calling 'show' for most tokens, but is different for paragraphs and tables,
+-- which have newlines inserted into them.
+prettyPrint :: MdToken -> String
+prettyPrint (Para token) = "<p>" ++ T.unpack (wrapText defaultWrapSettings 70 (T.pack $ prettyPrint token)) ++ "</p>\n"
+prettyPrint (Table (thead : tokenGrid)) = "<table>\n<thead>\n<tr>\n" ++ concatMap (\x -> "<th>" ++ rstrip (prettyPrint x) ++ "</th>\n") thead ++ "</tr>\n</thead>\n" ++ "<tbody>\n" ++ concatMap (\x -> "<tr>\n" ++ concatMap (\y -> "<td>" ++ rstrip (prettyPrint y) ++ "</td>\n") x ++ "</tr>\n") tokenGrid ++ "</tbody>\n</table>\n"
+prettyPrint Linebreak = "<br />\n"
+prettyPrint HorizontalRule = "<hr>\n"
+prettyPrint (Line tokens) = concatMap prettyPrint tokens
+prettyPrint (Document tokens) = concatMap prettyPrint tokens
+prettyPrint token = show token
+
 instance Semigroup MdToken where
   a <> b = Document [a, b]
 
@@ -83,27 +108,26 @@ leftmostLongest xs =
         (x : xs) -> Just x
 
 -- Get the first parse returned by readP_to_S that consumed the most input
-leftmostLongestParse :: (Monoid a) => ReadP a -> String -> (a, String)
+leftmostLongestParse :: (Monoid a) => Parser a -> String -> a
 leftmostLongestParse parser input =
-  let res = leftmostLongest $ readP_to_S parser input
+  case runParser parser "input" (T.pack input) of
-   in case res of
+    (Left a) -> mempty
-        Nothing -> (mempty, mempty)
+    (Right a) -> a
-        Just x -> x
 
-specialChars = "\n\\`*_{}[]()<>#+|"
+specialChars = ">\n\\`*_{}[]#+|"
 
-escapableChars = "-~!." ++ specialChars
+escapableChars = "-~!.$()" ++ specialChars
 
 -- Makes a parser greedy. Instead of returning all possible parses, only the longest one is returned.
-greedyParse :: ReadP a -> ReadP [a]
+greedyParse :: Parser a -> Parser [a]
 greedyParse parser = do
-  greedyParse1 parser <++ return []
+  greedyParse1 parser <|> return []
 
 -- Like greedyParse, but the parser must succeed atleast once.
-greedyParse1 :: ReadP a -> ReadP [a]
+greedyParse1 :: Parser a -> Parser [a]
 greedyParse1 parser = do
   parsed1 <- parser
-  parsed2 <- greedyParse1 parser <++ return []
+  parsed2 <- greedyParse1 parser <|> return []
   return (parsed1 : parsed2)
 
 prepend :: [a] -> [a] -> [a]
@@ -113,124 +137,204 @@ append :: [a] -> [a] -> [a]
 append x1 x2 = x2 ++ x1
 
 -- Parse until EOL or EOF
-parseTillEol :: ReadP String
+parseTillEol :: Parser String
-parseTillEol = manyTill get (void (char '\n') <++ eof)
+parseTillEol = manyTill anySingle (void (char '\n') <|> eof)
 
 -- Takes a list of parsers. Returns a parser that will try them in
 -- order, moving to the next one only if the current one fails.
-fallthroughParser :: [ReadP a] -> ReadP a
+fallthroughParser :: [Parser a] -> Parser a
 fallthroughParser [x] = x
-fallthroughParser (x : xs) = x <++ fallthroughParser xs
+fallthroughParser (x : xs) = try x <|> fallthroughParser xs
+
+escapeChar :: Char -> String
+escapeChar '>' = "&gt;"
+escapeChar '<' = "&lt;"
+escapeChar '&' = "&amp;"
+escapeChar x = [x]
+
+htmlEscapeChars :: T.Text -> T.Text
+htmlEscapeChars = T.concatMap (T.pack . escapeChar)
+
+-- -- Wraps a list of words after (at most) the given number of characters, trying to prevent word-breaks
+-- wordwrap :: Int -> String -> String
+-- wordwrap wraplength str = if (length str) < wraplength
+--     then str
+--     else
+--         let spaceIndex = lastgtSpaceIndex 0 (takeRev (length str) - wraplength str)
+--
+--     where
+--         takeRev n = (reverse . take n . reverse)
+--         lastSpaceIndex counter str = case str of
+--             [] -> counter
+--             x:xs -> if (isSpace x) counter else lastSpaceIndex counter+1 xs
 
 -- ---------------
 
 -- Parse a markdown header, denoted by 1-6 #'s followed by some text, followed by EOL.
-parseHeader :: ReadP MdToken
+parseHeader :: Parser MdToken
 parseHeader = do
-  skipSpaces
+  space
-  headers <- munch1 (== '#')
+  headers <- greedyParse1 (char '#')
   when
     (length headers > 6)
-    pfail
+    empty
-  skipSpaces
+  space
-  text <- munch1 (/= '\n')
+  parsedText <- manyTill parseLineToken (void (char '\n') <|> eof)
-  --  Text.ParserCombinators.ReadP.optional (char '\n')
+  greedyParse (char '\n')
-  skipSpaces
+  return (Header (length headers) (Line parsedText))
-  let parsedText = fst $ leftmostLongestParse parseLine text
+
-  return (Header (length headers) parsedText)
+asteriskBold = T.pack "**"
+
+underscoreBold = T.pack "__"
 
 -- Parse bold text
-parseBold :: ReadP MdToken
+parseBold :: Parser MdToken
-parseBold = parseBoldWith "**" <|> parseBoldWith "__"
+parseBold = parseBoldWith asteriskBold <|> parseBoldWith underscoreBold
   where
     parseBoldWith delim = do
       string delim
-      inside <- greedyParse1 parseLineToken
+      inside <- someTill parseLineToken $ string delim
-      string delim
       return (Bold (Line inside))
 
 -- Parse italic text
-parseItalic :: ReadP MdToken
+parseItalic :: Parser MdToken
-parseItalic = parseItalicWith "*" <|> parseItalicWith "_"
+parseItalic = parseItalicWith '*' <|> parseItalicWith '_'
   where
     parseItalicWith delim = do
-      string delim
+      char delim
-      inside <- greedyParse1 parseLineToken
+      inside <- someTill parseLineToken (char delim)
-      string delim
       return (Italic (Line inside))
 
+-- Parse subscript
+parseSubscript :: Parser MdToken
+parseSubscript = do
+  char '~'
+  inside <- someTill parseLineToken (char '~')
+  return (Subscript (Line inside))
+
+-- Parse superscript
+parseSuperscript :: Parser MdToken
+parseSuperscript = do
+  char '^'
+  inside <- someTill parseLineToken (char '^')
+  return (Superscript (Line inside))
+
 -- Parse strikethrough text
-parseStrikethrough :: ReadP MdToken
+parseStrikethrough :: Parser MdToken
 parseStrikethrough = do
-  string "~~"
+  string (T.pack "~~")
-  inside <- many1 parseLineToken
+  inside <- someTill parseLineToken $ string (T.pack "~~")
-  string "~~"
   return (Strikethrough (Line inside))
 
 -- Parse code
-parseCode :: ReadP MdToken
+parseCode :: Parser MdToken
 parseCode = do
-  string "`"
+  opening <- some $ char '`'
-  inside <- many1 get
+  inside <- someTill (satisfy (/= '\n')) (char '`')
-  string "`"
+  closing <- count (length opening - 1) (char '`')
-  return (Code (Unit inside))
+  return (Code (Unit (concatMap escapeChar inside)))
 
 -- Parse a link
-parseLink :: ReadP MdToken
+parseLink :: Parser MdToken
 parseLink = do
-  linkText <- between (string "[") (string "]") (many1 get)
+  char '['
-  linkURL <- between (string "(") (string ")") (many1 get)
+  linkText <- manyTill parseLineToken (char ']')
-  let parsedLinkText = fst $ leftmostLongestParse parseLine linkText
+  char '('
-  return $ Link parsedLinkText (URL linkURL)
+  linkURL <- manyTill anySingle (char ')')
+  return $ Link (Line linkText) (URL linkURL)
 
 -- Parse a linebreak character
-parseLinebreak :: ReadP MdToken
+parseLinebreak :: Parser MdToken
-parseLinebreak = do
+parseLinebreak = parseLinebreakSpace <|> parseLinebreakBackslash
-  char ' '
+  where
-  many1 (char ' ')
+    parseLinebreakSpace = do
-  char '\n'
+      char ' '
-  return Linebreak
+      some (char ' ')
+      char '\n'
+      return Linebreak
+    parseLinebreakBackslash = try $ do
+      char '\\'
+      char '\n'
+      return Linebreak
 
-parseSingleNewline :: ReadP MdToken
+parseTableRow :: Parser [MdToken]
+parseTableRow = do
+  char '|'
+  row <- some (many (satisfy (\x -> x == ' ' || x == '\t')) *> someTill parseListLineToken (char '|'))
+  return (map Line row)
+
+parseTable :: Parser MdToken
+parseTable = do
+  tableHead <- parseTableRow
+  char '\n'
+  char '|'
+  sepEndBy1 (some (char '-')) (char '|') *> char '\n'
+  tableBody <- sepEndBy parseTableRow (char '\n')
+  many (char '\n') -- Parse trailing newlines, if any
+  return $ Table (tableHead : tableBody)
+
+parseSingleNewline :: Parser MdToken
 parseSingleNewline = do
   char '\n'
-  remaining <- look
+  remaining <- getInput
-  case remaining of
+  case T.unpack remaining of
     [] -> return $ Unit ""
     _ -> return SingleNewline
 
-parseImage :: ReadP MdToken
+parseCssClasses :: Parser [CssClass]
+parseCssClasses = do
+  char '{'
+  classes <- some parseCssClass
+  char '}'
+  return classes
+  where
+    parseCssClass :: Parser CssClass
+    parseCssClass = do
+      char '.'
+      let firstLetterParser = char '_' <|> char '-' <|> label "letter" (satisfy isAlpha)
+      cssClassFirstLetter <- firstLetterParser
+      cssClass <- many (firstLetterParser <|> label "digit" (satisfy isDigit))
+      space
+      return (cssClassFirstLetter : cssClass)
+
+parseImage :: Parser MdToken
 parseImage = do
   char '!'
-  char '['
+  link <- parseLink
-  altText <- many1 (parseEscapedChar <++ parseUnit)
+  cssClasses <- optional $ try parseCssClasses
-  char ']'
+  case link of
-  char '('
+    Link text path -> return $ Image text path cssClasses
-  path <- many1 get
+    _ -> empty -- This should never be reached
-  char ')'
-  return $ Image (Line altText) (ImgPath path)
 
 parseFigure = do
   img <- parseImage
-  void (string "\n\n") <++ eof
+  void (string doubleNewlineText) <|> eof
   case img of
-    Image text path -> return $ Figure text path
+    Image text path cssClasses -> return $ Figure text path cssClasses
     _ -> return img
 
 -- Parse an escaped character
-parseEscapedChar :: ReadP MdToken
+parseEscapedChar :: Parser MdToken
 parseEscapedChar = do
   char '\\'
   escapedChar <- choice (map char escapableChars) -- Parse any of the special chars.
   return (Unit [escapedChar])
 
 -- Parse a character as a Unit.
-parseUnit :: ReadP MdToken
+parseUnit :: Parser MdToken
 parseUnit = do
-  text <- satisfy (`notElem` specialChars)
+  -- text <- satisfy (`notElem` specialChars)
+  text <- anySingle
   return (Unit [text])
 
-lineParsers :: [ReadP MdToken]
+-- Parse any character except a newline
+parseUnitExceptNewline :: Parser MdToken
+parseUnitExceptNewline = do
+  -- text <- satisfy (`notElem` specialChars)
+  text <- satisfy (/= '\n')
+  return (Unit [text])
+
+lineParsers :: [Parser MdToken]
 lineParsers =
   [ parseLinebreak,
     parseSingleNewline,
@@ -240,102 +344,112 @@ lineParsers =
     parseBold,
     parseItalic,
     parseStrikethrough,
+    parseSubscript,
+    parseSuperscript,
     parseLink,
     parseUnit
   ] -- A 'line' doesn't include a 'header'
 
-listLineParsers :: [ReadP MdToken]
+lineParsersWithoutNewline :: [Parser MdToken]
-listLineParsers =
+lineParsersWithoutNewline =
-  [ parseLinebreak,
+  [ parseEscapedChar,
-    parseEscapedChar,
     parseCode,
     parseImage,
     parseBold,
     parseItalic,
     parseStrikethrough,
+    parseSubscript,
+    parseSuperscript,
     parseLink,
-    parseUnit
+    parseUnitExceptNewline
   ] -- A list line cannot contain newlines.
 
 -- List of all parsers
-allParsers :: [ReadP MdToken]
+allParsers :: [Parser MdToken]
 allParsers = parseHeader : lineParsers
 
 -- Parse any of the line tokens.
-parseLineToken :: ReadP MdToken
+parseLineToken :: Parser MdToken
 parseLineToken = fallthroughParser lineParsers
 
 -- Parse any of the list line tokens.
-parseListLineToken :: ReadP MdToken
+parseListLineToken :: Parser MdToken
-parseListLineToken = fallthroughParser listLineParsers
+parseListLineToken = fallthroughParser lineParsersWithoutNewline
 
 -- Parse a line, consisting of one or more tokens.
-parseLine :: ReadP MdToken
+parseLine :: Parser MdToken
 parseLine = do
-  skipSpaces
+  space
   -- Fail if we have reached the end of the document.
   parsed <- manyTill parseLineToken eof
   return (Line parsed)
 
 -- Parse a paragraph, which is a 'Line' (can span multiple actual lines), separated by double-newlines.
-parsePara :: ReadP MdToken
+parsePara :: Parser MdToken
 parsePara = do
-  parseMany (char '\n')
+  space
   --  text <- many1 (lookaheadParse (\x -> ((length x) < 2) || (take 2 x) /= "\n\n")) -- Parse until a double-newline.
   --  string "\n\n" <|> (eof >> return "") -- Consume the next double-newline or EOF.
-  text <- manyTill get (string "\n\n" <|> (eof >> return ""))
+  parsedText <- someTill parseLineToken (try paraEnding)
-  when (null text) pfail
+  many (char '\n')
-  let parsedText = fst $ leftmostLongestParse parseLine text -- Parse a line
+  return (Para (Line parsedText))
-  parseMany (char '\n')
+  where
-  return (Para parsedText)
+    paraEnding = void (char '\n' *> (char '\n' <|> lookAhead (char '>'))) <|> eof
 
 -- Parse a line starting with '>', return the line except for the '>'.
-parseQuotedLine :: ReadP String
+parseQuotedLine :: Parser String
 parseQuotedLine = do
   char '>'
-  greedyParse (char ' ' +++ char '\t')
+  many (char ' ' <|> char '\t')
-  restOfLine <- munch (/= '\n')
+  restOfLine <- many (satisfy (/= '\n'))
-  Text.ParserCombinators.ReadP.optional (char '\n') >> return ""
+  void (char '\n') <|> eof
   return restOfLine
 
 -- Parse many 'quoted lines' until I see a non-quoted line.
-parseQuotedLines :: ReadP [String]
+parseQuotedLines :: Parser [String]
-parseQuotedLines =
+parseQuotedLines = some parseQuotedLine
-  greedyParse1 $ do
+
-    look >>= \line ->
+-- some $ do
-      case line of
+--   getInput >>= \line ->
-        ('>' : _) -> parseQuotedLine
+--     case T.unpack line of
-        _ -> pfail
+--       ('>' : _) -> parseQuotedLine
+--       _ -> empty
 
 -- Parse a blockquote, which is a greater-than sign followed by a paragraph.
-parseBlockquote :: ReadP MdToken
+parseBlockquote :: Parser MdToken
 parseBlockquote = do
   quotedLines <- parseQuotedLines
   -- remaining <- look
   -- let quotedLines = fst $ leftmostLongestParse parseQuotedLines remaining
   -- string (init $ unlines quotedLines)
-  let parsedQuotedLines = fst $ leftmostLongestParse (many1 (parseBlockquote <++ parsePara)) (init $ unlines quotedLines) -- unlines joins the lines together with a newline, and adds a trailing newline. init removes the trailing newline.
+  let parsedQuotedLines = leftmostLongestParse (some (parseBlockquote <|> parsePara)) (init $ unlines quotedLines) -- unlines joins the lines together with a newline, and adds a trailing newline. init removes the trailing newline.
   return (Blockquote parsedQuotedLines)
 
 -- Parse a nested list item.
-parseListNested :: ReadP MdToken
+parseListNested :: Parser MdToken
 parseListNested = do
-  let firstCharParser = string "    " <++ string "\t"
+  let firstCharParser = (<>) <$> (string (T.pack "  ") <|> string (T.pack "\t")) <*> (T.pack <$> many (char ' '))
-  let restOfLineParser = manyTill get (void (char '\n') <++ eof)
+  let restOfLineParser = manyTill anySingle (void (char '\n') <|> eof)
-  lines <- greedyParse1 (firstCharParser *> restOfLineParser)
+  -- For the first line, I manually run firstCharParser and restOfLineParser. The
-  let linesParsed = fst $ leftmostLongestParse (parseUnorderedList <++ parseOrderedList) (init $ unlines lines)
+  -- result of firstCharParser is saved. For every subsequent line, I parse exactly
-  when (null (show linesParsed)) pfail
+  -- the same string as firstCharParser.
+  firstLineSpaces <- firstCharParser
+  firstLine <- restOfLineParser
+  lines <- greedyParse (string firstLineSpaces *> restOfLineParser)
+  let allLines = firstLine : lines
+  let linesParsed = leftmostLongestParse (parseUnorderedList <|> parseOrderedList) (init $ unlines allLines)
+  when (null (show linesParsed)) empty
   return linesParsed
 
 -- Parse an unordered list line item.
-parseUListLineItem :: ReadP MdToken
+parseUListLineItem :: Parser MdToken
 parseUListLineItem = do
   firstChar <- choice (map char ['*', '+', '-'])
   char ' ' -- At least one space between list indicator and list text.
   parseListLineItemCommon
 
 -- Parse an ordered list line item.
-parseOListLineItem :: ReadP MdToken
+parseOListLineItem :: Parser MdToken
 parseOListLineItem = do
   num <- greedyParse1 (satisfy isDigit)
   char '.'
@@ -343,23 +457,22 @@ parseOListLineItem = do
   parseListLineItemCommon
 
 -- Common code for parsing list line items
-parseListLineItemCommon :: ReadP MdToken
+parseListLineItemCommon :: Parser MdToken
 parseListLineItemCommon = do
-  skipSpaces
+  space
-  restOfLine <- many1 parseListLineToken
+  restOfLine <- manyTill parseListLineToken (void (char '\n') <|> eof)
-  void (char '\n') <++ eof
+  nestedList <- try parseListNested <|> return (Unit "")
-  nestedList <- parseListNested <++ return (Unit "")
   return $ Line [Line restOfLine, nestedList]
 
 -- Parse an unordered list paragraph item.
-parseUListParaItem :: ReadP MdToken
+parseUListParaItem :: Parser MdToken
 parseUListParaItem = do
   firstLine <- parseUListLineItem
   res <- parseListParaItemCommon
   return $ Document (Para firstLine : res) -- I only wrap this in a document because I want some way of converting [MdToken] to MdToken, without any overhead. There is no other reason to wrap it in a Document.
 
 -- Parse an unordered list paragraph item.
-parseOListParaItem :: ReadP MdToken
+parseOListParaItem :: Parser MdToken
 parseOListParaItem = do
   firstLine <- parseOListLineItem
   res <- parseListParaItemCommon
@@ -369,50 +482,65 @@ parseOListParaItem = do
 -- A list paragraph item is defined as a line item, followed by an empty line, followed by one or more
 -- lines indented by a space or tab.
 -- A list paragraph item can also be a blockquote.
-parseListParaItemCommon :: ReadP [MdToken]
+parseListParaItemCommon :: Parser [MdToken]
 parseListParaItemCommon = do
   char '\n'
-  lines <- greedyParse1 ((string "    " <|> string "\t") *> parseTillEol)
+  lines <- greedyParse1 ((string (T.pack "    ") <|> string (T.pack "\t")) *> parseTillEol)
-  let res = fst $ leftmostLongestParse (greedyParse1 parseBlockquote <++ greedyParse1 parsePara) (init $ unlines lines)
+  let res = leftmostLongestParse (greedyParse1 parseBlockquote <|> greedyParse1 parsePara) (init $ unlines lines)
   char '\n'
   return res -- I only wrap this in a document because I want some way of converting [MdToken] to MdToken, without any overhead. There is no other reason to wrap it in a Document.
 
 -- Parse an unordered list item, which can be a line item or another list.
-parseUListItem :: ReadP MdToken
+parseUListItem :: Parser MdToken
-parseUListItem = parseUListParaItem <++ parseUListLineItem
+parseUListItem = space *> (try parseUListParaItem <|> parseUListLineItem)
 
 -- Parse an unordered list.
-parseUnorderedList :: ReadP MdToken
+parseUnorderedList :: Parser MdToken
 parseUnorderedList = do
-  lineItems <- greedyParse1 parseUListItem
+  lineItems <- some $ try parseUListItem
-  void (char '\n') <++ eof -- A list must end in an extra newline or eof
+  void (char '\n') <|> eof -- A list must end in an extra newline or eof
   return $ UnordList lineItems
 
 -- --------
 
-parseOListItem :: ReadP MdToken
+parseOListItem :: Parser MdToken
-parseOListItem = parseOListParaItem <++ parseOListLineItem
+parseOListItem = space *> (try parseOListParaItem <|> parseOListLineItem)
 
 -- Parses the first element of an ordered list, which must start with '1.'
-parseFirstOListItem :: ReadP MdToken
+parseFirstOListItem :: Parser MdToken
 parseFirstOListItem = do
-  remaining <- look
+  space
-  when (take 2 remaining /= "1.") pfail
+  remaining <- getInput
+  when (take 2 (T.unpack remaining) /= "1.") empty
   parseOListLineItem
 
-parseOrderedList :: ReadP MdToken
+parseOrderedList :: Parser MdToken
 parseOrderedList = do
-  firstLine <- parseFirstOListItem
+  firstLine <- try parseFirstOListItem
-  lineItems <- greedyParse1 parseOListItem
+  lineItems <- many $ try parseOListItem
-  void (char '\n') <++ eof
+  void (char '\n') <|> eof
   return $ OrdList (firstLine : lineItems)
 
-parseHorizontalRule :: ReadP MdToken
+horizontalRuleText :: T.Text
-parseHorizontalRule = string "---" *> (void (string "\n\n") <++ eof) *> return HorizontalRule
+horizontalRuleText = T.pack "---"
 
-documentParsers :: [ReadP MdToken]
+doubleNewlineText :: T.Text
+doubleNewlineText = T.pack "\n\n"
+
+parseHorizontalRule :: Parser MdToken
+parseHorizontalRule = string horizontalRuleText *> (void (string doubleNewlineText) <|> eof) *> return HorizontalRule
+
+parseCodeblock :: Parser MdToken
+parseCodeblock = do
+  string (T.pack "```\n")
+  inside <- someTill anySingle (string (T.pack "\n```"))
+  return $ Codeblock (Unit (concatMap escapeChar inside))
+
+documentParsers :: [Parser MdToken]
 documentParsers =
   [ parseHorizontalRule,
+    parseCodeblock,
+    parseTable,
     parseHeader,
     parseBlockquote,
     parseUnorderedList,
@@ -422,7 +550,7 @@ documentParsers =
   ]
 
 -- Parse a document, which is multiple paragraphs.
-parseDocument :: ReadP MdToken
+parseDocument :: Parser MdToken
 parseDocument = do
   res <- manyTill (fallthroughParser documentParsers) eof
   return (Document res)

src/MdToHtmlTest.hs

@@ -7,7 +7,7 @@ check_equal :: String -> String -> String -> Test
 check_equal desc expected actual = TestCase (assertEqual desc expected actual)
 
 convert :: String -> String
-convert md = show . fst $ leftmostLongestParse parseDocument md
+convert md = show $ leftmostLongestParse parseDocument md
 
 headerTests =
   TestList
@@ -24,14 +24,14 @@ boldTests =
     [ check_equal "Should convert bold" "<p><b>Hello</b></p>" (convert "__Hello__"),
       check_equal "Should convert italic" "<p><i>Hello</i></p>" (convert "_Hello_"),
       check_equal "Should convert bold and italic in a sentence" "<p>It <i>is</i> a <b>wonderful</b> day</p>" (convert "It _is_ a __wonderful__ day"),
-      check_equal "Should convert nested bold and italic" "<p><b>Bold then <i>Italic</i></b></p>" (convert "**Bold then *Italic***"),
+      check_equal "Should convert nested bold and italic" "<p><b>Bold then <i>Italic</i></b></p>" (convert "**Bold then _Italic_**"),
-      check_equal "Should convert nested bold and italic" "<p><i>Italic then <b>Bold</b></i></p>" (convert "*Italic then **Bold***")
+      check_equal "Should convert nested bold and italic" "<p><i>Italic then <b>Bold</b></i></p>" (convert "*Italic then __Bold__*")
     ]
 
 strikethroughTests =
   TestList
     [ check_equal "Should convert strikethrough" "<p><s>Hello</s></p>" (convert "~~Hello~~"),
-      check_equal "Should convert long sentence with tilde" "<p><s>The universe is ~7 days old</s>. The universe is 13 billion years old.</p>" (convert "~~The universe is ~7 days old~~. The universe is 13 billion years old.")
+      check_equal "Should convert long sentence with tilde" "<p><s>The universe is ~7 days old</s>. The universe is 13 billion years old.</p>" (convert "~~The universe is \\~7 days old~~. The universe is 13 billion years old.")
     ]
 
 linkTests =
@@ -74,6 +74,7 @@ unorderedListTests =
       check_equal "Paragraph before list" "<p>This is a list</p><ul><li>Item 1</li><li>Item 2</li></ul>" (convert "This is a list\n\n* Item 1\n* Item 2"),
       check_equal "Paragraph before list" "<h3>This is a list</h3><ul><li>Item 1</li><li>Item 2</li></ul>" (convert "### This is a list\n\n* Item 1\n* Item 2"),
       check_equal "Nested list then back" "<ul><li>Item 1</li><li>Item 2<ul><li>Item 3</li><li>Item 4</li></ul></li><li>Item 5</li></ul>" (convert "- Item 1\n- Item 2\n    - Item 3\n    - Item 4\n- Item 5"),
+      check_equal "Triply nested list" "<ul><li>Item 1</li><li>Item 2<ul><li>Item 3<ul><li>Item 4</li></ul></li></ul></li><li>Item 5</li></ul>" (convert "- Item 1\n- Item 2\n    - Item 3\n      - Item 4\n- Item 5"),
       check_equal "Blockquote in list" "<ul><li>Item 1</li><li><p>Item 2</p><blockquote><p>Quote</p></blockquote></li><li>Item 3</li></ul>" (convert "- Item 1\n- Item 2\n\n    > Quote\n\n- Item 3"),
       check_equal "Ordered list in unordered list" "<ul><li>Item 1</li><li>Item 2<ol><li>Item 1</li><li>Item 2</li></ol></li><li>Item 3</li></ul>" (convert "- Item 1\n- Item 2\n    1. Item 1\n    2. Item 2\n- Item 3")
     ]
@@ -90,19 +91,25 @@ orderedListTests =
       check_equal "Paragraph before list" "<h3>This is a list</h3><ol><li>Item 1</li><li>Item 2</li></ol>" (convert "### This is a list\n\n1. Item 1\n200. Item 2"),
       check_equal "Nested list then back" "<ol><li>Item 1</li><li>Item 2<ol><li>Item 3</li><li>Item 4</li></ol></li><li>Item 5</li></ol>" (convert "1. Item 1\n2. Item 2\n    1. Item 3\n    3. Item 4\n5. Item 5"),
       check_equal "Blockquote in list" "<ol><li>Item 1</li><li><p>Item 2</p><blockquote><p>Quote</p></blockquote></li><li>Item 3</li></ol>" (convert "1. Item 1\n2. Item 2\n\n    > Quote\n\n3. Item 3"),
-      check_equal "Unordered list in ordered list" "<ol><li>Item 1</li><li>Item 2<ul><li>Item 1</li><li>Item 2</li></ul></li><li>Item 3</li></ol>" (convert "1. Item 1\n2. Item 2\n    - Item 1\n    * Item 2\n4. Item 3")
+      check_equal "Unordered list in ordered list" "<ol><li>Item 1</li><li>Item 2<ul><li>Item 1</li><li>Item 2</li></ul></li><li>Item 3</li></ol>" (convert "1. Item 1\n2. Item 2\n    - Item 1\n    * Item 2\n4. Item 3"),
+      check_equal "List with just 1 item" "<ol><li>Item 1</li></ol>" (convert "1. Item 1")
     ]
 
+htmlTests =
+  TestList
+    [check_equal "Convert HTML element" "<p><center>a</center></p>" (convert "<center>a</center>")]
+
 codeTests =
   TestList
     [ check_equal "Code by itself" "<p><code>Hello world!</code></p>" (convert "`Hello world!`"),
       check_equal "Code in a paragraph" "<p>The following <code>text</code> is code</p>" (convert "The following `text` is code"),
-      check_equal "Code across paragraphs (shouldn't work" "<p></p><p></p>" (convert "`Incomplete\n\nCode`") -- At the moment, this is just treated as a syntax error, so nothing is rendered.
+      check_equal "Code across paragraphs (shouldn't work)" "<p>`Incomplete</p><p>Code`</p>" (convert "`Incomplete\n\nCode`") -- At the moment, this is just treated as a syntax error, so nothing is rendered.
     ]
 
 imageTests =
   TestList
-    [ check_equal "Image with text" "<p>This is an image <img src=\"img.png\" alt=\"Image 1\" /></p>" (convert "This is an image ![Image 1](img.png)")
+    [ check_equal "Image with text" "<p>This is an image <img src=\"img.png\" alt=\"Image 1\"/></p>" (convert "This is an image ![Image 1](img.png)"),
+      check_equal "Image with classes" "<p>This is an image <img src=\"img.png\" alt=\"Image 1\" class=\"new-img\"/></p>" (convert "This is an image ![Image 1](img.png){.new-img}")
     ]
 
 figureTests =
@@ -114,6 +121,25 @@ horizontalRuleTests =
   TestList
     [check_equal "Horizontal Rule" "<p>a</p><hr><p>b</p>" (convert "a\n\n---\n\nb")]
 
+subscriptTests =
+  TestList
+    [check_equal "Should convert subscript" "A<sub>b</sub>" (convert "A~b~")]
+
+superscriptTests =
+  TestList
+    [check_equal "Should convert superscript" "A<sup>b</sup>" (convert "A^b^")]
+
+tableTests =
+  TestList
+    [ check_equal
+        "Basic table"
+        "<table>\
+        \<thead><tr><th>Col 1</th><th>Col 2</th><th>Col 3</th></tr></thead>\
+        \<tbody><tr><td>Data 1</td><td>Data 2</td><td>Data 3</td></tr>\
+        \<tr><td>More Data 1</td><td>More Data 2</td><td>More Data 3</td></tr></tbody></table>"
+        (convert "| Col 1 | Col 2 | Col 3 |\n|---|---|---|\n| Data 1 | Data 2 | Data 3 |\n| More Data 1 | More Data 2 | More Data 3 |")
+    ]
+
 integrationTests =
   TestList
     [ check_equal "Integration 1" "<h1>Sample Markdown</h1><p>This is some basic, sample markdown.</p><h2><b>Second</b> <i>Heading</i></h2>" (convert "# Sample Markdown\n\n This is some basic, sample markdown.\n\n  ## __Second__ _Heading_"),
@@ -121,7 +147,7 @@ integrationTests =
       check_equal "Integration 3" "<h1>Hello</h1><p>World</p>" (convert "# Hello\nWorld"),
       check_equal "Integration 4" "<p>a b</p>" (convert "a\nb"),
       check_equal "Integration 5" "<h1>Hello</h1>" (convert "# Hello\n"),
-      check_equal "Integration 6" "<p>First line<br>Second line</p>" (convert "First line  \nSecond line"),
+      check_equal "Integration 6" "<p>First line<br />Second line</p>" (convert "First line  \nSecond line"),
       check_equal
         "Integration 7"
         "<h1>Sample Markdown</h1><p>This is some basic, sample markdown.</p><h2>Second \
@@ -134,7 +160,7 @@ integrationTests =
             "# Sample Markdown\n\nThis is some basic, sample markdown.\n\n## Second \
             \Heading\n\n- Unordered lists, and:\n    1. One\n    2. Two\n    3. Three\n\
             \- More\n\n> Blockquote\n\nAnd **bold**, *italics*, and even *italics and \
-            \later **bold***. Even ~~strikethrough~~. [A link](https://markdowntohtml.com) to somewhere."
+            \later __bold__*. Even ~~strikethrough~~. [A link](https://markdowntohtml.com) to somewhere."
         )
     ]
 
@@ -149,9 +175,11 @@ tests =
       unorderedListTests,
       orderedListTests,
       imageTests,
+      htmlTests,
       figureTests,
       codeTests,
       horizontalRuleTests,
+      tableTests,
       integrationTests
     ]