Add prettyPrint definition for horizontal rule

built-in one instead

app/Main.hs

@@ -1,8 +1,30 @@
 module Main where
 
 import MdToHTML
+import System.Environment
+import System.IO
+
+readLinesHelper :: [String] -> IO [String]
+readLinesHelper xs = do
+  done <- isEOF
+  if done
+    then return xs
+    else do
+      line <- getLine
+      let xs' = line : xs
+      readLinesHelper xs'
+
+readLines :: IO [String]
+readLines = reverse <$> readLinesHelper []
 
 main :: IO ()
 main = do
-   let res = fst $ leftmostLongestParse parseDocument "# _Hello_\n"
+  args <- getArgs
-   putStrLn (show res)
+  fileContents <- case args of
+    [] -> getContents
+    x : _ -> readFile x
+  let res = leftmostLongestParse parseDocument fileContents
+  let toPrint = prettyPrint res
+  case reverse toPrint of
+    '\n' : _ -> putStr toPrint
+    _ -> putStrLn toPrint

md-to-html.cabal

@@ -56,18 +56,22 @@ common warnings
 library
    hs-source-dirs:   src
    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
 
     -- .hs or .lhs file containing the Main module.
     main-is:          Main.hs
 
-    -- Modules included in this executable, other than Main.
-    -- other-modules:
 
     -- LANGUAGE extensions used by modules in this package.
     -- other-extensions:

src/MdToHTML.hs

@@ -1,177 +1,528 @@
+{-# OPTIONS_GHC -Wno-unrecognised-pragmas #-}
+
+{-# HLINT ignore "Use lambda-case" #-}
+
 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
 
-newtype URL = URL {getUrl :: String}
+type CssClass = String
 
-newtype ImgPath = ImgPath {getPath :: String}
+newtype URL = URL {getUrl :: String} deriving (Eq)
 
-parseMany :: ReadP a -> ReadP [a]
+newtype ImgPath = ImgPath {getPath :: String} deriving (Eq)
-parseMany = Text.ParserCombinators.ReadP.many
 
 data MdToken
   = Document [MdToken]
   | Header HeaderLevel MdToken
   | Para MdToken
   | Line [MdToken]
+  | SingleNewline -- A single newline is rendered as a space.
   | Linebreak
   | HorizontalRule
-  | Blockquote MdToken
+  | Blockquote [MdToken]
   | UnordList [MdToken]
   | OrdList [MdToken]
-  | Code String
+  | Code MdToken
-  | Codeblock String
+  | Table [[MdToken]]
+  | Codeblock MdToken
   | Link MdToken URL
-  | Image MdToken ImgPath
+  | Image MdToken URL (Maybe [CssClass])
+  | Figure MdToken URL (Maybe [CssClass])
   | Bold MdToken
   | Italic MdToken
   | Strikethrough MdToken
   | Unit String
+  deriving (Eq)
 
 -- Deriving Show for MdToken
 instance Show MdToken where
-  show (Document tokens) = concat (map show tokens)
+  show (Document tokens) = concatMap show tokens
   show (Header level token) = "<h" ++ show level ++ ">" ++ show token ++ "</h" ++ show level ++ ">"
   show (Para token) = "<p>" ++ show token ++ "</p>"
-  show (Line tokens) = concat (map show tokens)
+  show (Line tokens) = concatMap show tokens
-  show Linebreak = "<br>"
+  show Linebreak = "<br />"
-  show HorizontalRule = "---------"
+  show SingleNewline = " "
-  show (Blockquote token) = "BLOCK" ++ show token
+  show HorizontalRule = "<hr>"
-  show (UnordList tokens) = "UNORD" ++ concat (map show tokens)
+  show (Blockquote tokens) = "<blockquote>" ++ concatMap show tokens ++ "</blockquote>"
-  show (OrdList tokens) = "ORD" ++ concat (map show tokens)
+  show (UnordList tokens) = "<ul>" ++ concatMap (prepend "<li>" . append "</li>" . show) tokens ++ "</ul>"
-  show (Code code) = show code
+  show (OrdList tokens) = "<ol>" ++ concatMap (prepend "<li>" . append "</li>" . show) tokens ++ "</ol>"
-  show (Codeblock code) = show code
+  show (Code code) = "<code>" ++ strip (show code) ++ "</code>"
-  show (Link txt url) = "<a href=" ++ (getUrl url) ++ ">" ++ show txt ++ "</a>"
+  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 (Image txt imgPath) = "<img src=" ++ (getPath imgPath) ++ ">" ++ show txt ++ "</img>"
+  show (Codeblock code) = "<pre><code>" ++ show code ++ "</code></pre>"
+  show (Link txt url) = "<a href=\"" ++ getUrl url ++ "\">" ++ show txt ++ "</a>"
+  show (Image txt url cssClasses) = "<img src=\"" ++ getUrl url ++ "\"" ++ " alt=\"" ++ show txt ++ "\"" ++ maybe "" (\classes -> " class=\"" ++ unwords classes ++ "\"") cssClasses ++ "/>"
+  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 (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]
+
+instance Monoid MdToken where
+  mempty = Unit ""
+
 -- ---------------
 -- Helpers
-mustBeHash :: ReadP Char
+leftmostLongest :: (Foldable t) => [(a, t b)] -> Maybe (a, t b)
-mustBeHash = satisfy (\x -> x == '#')
-
-leftmostLongest :: (Foldable t) => [(a, t b)] -> (a, t b)
 leftmostLongest xs =
-  let lastElem = (last xs)
+  let lastElem = last xs
-      filteredLst = (filter (\val -> (length $ snd val) == (length $ snd lastElem)) xs)
+      filteredLst = filter (\val -> length (snd val) == length (snd lastElem)) xs
-   in head filteredLst
+   in case filteredLst of
+        [] -> Nothing
+        (x : xs) -> Just x
 
 -- Get the first parse returned by readP_to_S that consumed the most input
-leftmostLongestParse :: ReadP a -> String -> (a, String)
+leftmostLongestParse :: (Monoid a) => Parser a -> String -> a
-leftmostLongestParse parser input = leftmostLongest $ readP_to_S parser input
+leftmostLongestParse parser input =
+  case runParser parser "input" (T.pack input) of
+    (Left a) -> mempty
+    (Right a) -> a
 
--- Parse if the string that's left matches the string comparator function
+specialChars = ">\n\\`*_{}[]#+|"
-lookaheadParse :: (String -> Bool) -> ReadP Char
-lookaheadParse stringCmp = do
-  lookahead <- look
-  case stringCmp lookahead of
-    True -> get
-    False -> pfail
 
-lineToList :: MdToken -> [MdToken]
+escapableChars = "-~!.$()" ++ specialChars
-lineToList (Line tokens) = tokens
+
+-- Makes a parser greedy. Instead of returning all possible parses, only the longest one is returned.
+greedyParse :: Parser a -> Parser [a]
+greedyParse parser = do
+  greedyParse1 parser <|> return []
+
+-- Like greedyParse, but the parser must succeed atleast once.
+greedyParse1 :: Parser a -> Parser [a]
+greedyParse1 parser = do
+  parsed1 <- parser
+  parsed2 <- greedyParse1 parser <|> return []
+  return (parsed1 : parsed2)
+
+prepend :: [a] -> [a] -> [a]
+prepend x1 x2 = x1 ++ x2
+
+append :: [a] -> [a] -> [a]
+append x1 x2 = x2 ++ x1
+
+-- Parse until EOL or EOF
+parseTillEol :: Parser String
+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 :: [Parser a] -> Parser a
+fallthroughParser [x] = x
+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 <- many1 mustBeHash
+  headers <- greedyParse1 (char '#')
   when
-    ((length headers) > 6)
+    (length headers > 6)
-    pfail
+    empty
-  _ <- string " "
+  space
---  text <- manyTill (get) ((string "\n") <|> (eof >> return ""))-- Parse until EOL or EOF
+  parsedText <- manyTill parseLineToken (void (char '\n') <|> eof)
-  text <- munch1 (/= '\n')
+  greedyParse (char '\n')
-  Text.ParserCombinators.ReadP.optional (char '\n')
+  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 = do
+parseBold = parseBoldWith asteriskBold <|> parseBoldWith underscoreBold
-  text <-
+  where
-    choice
+    parseBoldWith delim = do
-      [ between (string "__") (string "__") (many1 (lookaheadParse (/= "__"))),
+      string delim
-        between (string "**") (string "**") (many1 (lookaheadParse (/= "**")))
+      inside <- someTill parseLineToken $ string delim
-      ]
+      return (Bold (Line inside))
-  let parsedText = fst $ leftmostLongestParse parseLine text
-  return (Bold parsedText)
 
 -- Parse italic text
-parseItalic :: ReadP MdToken
+parseItalic :: Parser MdToken
-parseItalic = do
+parseItalic = parseItalicWith '*' <|> parseItalicWith '_'
-  text <-
+  where
-    choice
+    parseItalicWith delim = do
-      [ (between (string "_") (string "_") (munch1 (/= '_'))),
+      char delim
-        (between (string "*") (string "*") (munch1 (/= '*')))
+      inside <- someTill parseLineToken (char delim)
-      ]
+      return (Italic (Line inside))
-  let parsedText = fst $ leftmostLongestParse parseLine text
+
-  return (Italic parsedText)
+-- Parse strikethrough text
+parseStrikethrough :: Parser MdToken
+parseStrikethrough = do
+  string (T.pack "~~")
+  inside <- someTill parseLineToken $ string (T.pack "~~")
+  return (Strikethrough (Line inside))
+
+-- Parse code
+parseCode :: Parser MdToken
+parseCode = do
+  opening <- some $ char '`'
+  inside <- someTill (satisfy (/= '\n')) (char '`')
+  closing <- count (length opening - 1) (char '`')
+  return (Code (Unit (concatMap escapeChar inside)))
+
+-- Parse a link
+parseLink :: Parser MdToken
+parseLink = do
+  char '['
+  linkText <- manyTill parseLineToken (char ']')
+  char '('
+  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
+  where
+    parseLinebreakSpace = do
       char ' '
-  many1 (char ' ')
+      some (char ' ')
+      char '\n'
+      return Linebreak
+    parseLinebreakBackslash = try $ do
+      char '\\'
       char '\n'
       return Linebreak
 
--- Parse a regular string as a Unit.
+parseTableRow :: Parser [MdToken]
-parseString :: ReadP MdToken
+parseTableRow = do
-parseString = do
+  char '|'
-  firstChar <- get -- Must parse at least one character here
+  row <- some (many (satisfy (\x -> x == ' ' || x == '\t')) *> someTill parseListLineToken (char '|'))
-  text <- munch (\x -> not (elem x "#*_[\n "))
+  return (map Line row)
-  return (Unit (firstChar : text))
 
-lineParsers :: [ReadP MdToken]
+parseTable :: Parser MdToken
-lineParsers = [parseLinebreak, parseBold, parseItalic, parseString] -- A 'line' doesn't include a 'header'
+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 <- getInput
+  case T.unpack remaining of
+    [] -> return $ Unit ""
+    _ -> return SingleNewline
+
+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 '!'
+  link <- parseLink
+  cssClasses <- optional $ try parseCssClasses
+  case link of
+    Link text path -> return $ Image text path cssClasses
+    _ -> empty -- This should never be reached
+
+parseFigure = do
+  img <- parseImage
+  void (string doubleNewlineText) <|> eof
+  case img of
+    Image text path cssClasses -> return $ Figure text path cssClasses
+    _ -> return img
+
+-- Parse an escaped character
+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 :: Parser MdToken
+parseUnit = do
+  -- text <- satisfy (`notElem` specialChars)
+  text <- anySingle
+  return (Unit [text])
+
+-- 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,
+    parseEscapedChar,
+    parseCode,
+    parseImage,
+    parseBold,
+    parseItalic,
+    parseStrikethrough,
+    parseLink,
+    parseUnit
+  ] -- A 'line' doesn't include a 'header'
+
+lineParsersWithoutNewline :: [Parser MdToken]
+lineParsersWithoutNewline =
+  [ parseEscapedChar,
+    parseCode,
+    parseImage,
+    parseBold,
+    parseItalic,
+    parseStrikethrough,
+    parseLink,
+    parseUnitExceptNewline
+  ] -- A list line cannot contain newlines.
 
 -- List of all parsers
-allParsers :: [ReadP MdToken]
+allParsers :: [Parser MdToken]
 allParsers = parseHeader : lineParsers
 
--- Parse any of the above tokens.
+-- Parse any of the line tokens.
-parseLineToken :: ReadP MdToken
+parseLineToken :: Parser MdToken
-parseLineToken = choice lineParsers
+parseLineToken = fallthroughParser lineParsers
+
+-- Parse any of the list line tokens.
+parseListLineToken :: Parser MdToken
+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.
-  remaining <- look
+  parsed <- manyTill parseLineToken eof
-  when (null remaining) pfail
-  parsed <- parseMany parseLineToken
-  --         traceM $ show parsed
   return (Line parsed)
 
 -- Parse a paragraph, which is a 'Line' (can span multiple actual lines), separated by double-newlines.
--- As a weird special case, a 'Paragraph' can also be a 'Header'.
+parsePara :: Parser MdToken
-parsePara :: ReadP 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 (parseHeader <|> parseLine) text -- Parse either a line or a header.
+  return (Para (Line parsedText))
-  -- If the paragraph is a header, return a Header token. Otheriwse return a Para token.
+  where
-  case parsedText of
+    paraEnding = void (char '\n' *> (char '\n' <|> lookAhead (char '>'))) <|> eof
-    Header level token -> return (Header level token)
+
-    _ -> return (Para parsedText)
+-- Parse a line starting with '>', return the line except for the '>'.
+parseQuotedLine :: Parser String
+parseQuotedLine = do
+  char '>'
+  many (char ' ' <|> char '\t')
+  restOfLine <- many (satisfy (/= '\n'))
+  void (char '\n') <|> eof
+  return restOfLine
+
+-- Parse many 'quoted lines' until I see a non-quoted line.
+parseQuotedLines :: Parser [String]
+parseQuotedLines = some parseQuotedLine
+
+-- some $ do
+--   getInput >>= \line ->
+--     case T.unpack line of
+--       ('>' : _) -> parseQuotedLine
+--       _ -> empty
+
+-- Parse a blockquote, which is a greater-than sign followed by a paragraph.
+parseBlockquote :: Parser MdToken
+parseBlockquote = do
+  quotedLines <- parseQuotedLines
+  -- remaining <- look
+  -- let quotedLines = fst $ leftmostLongestParse parseQuotedLines remaining
+  -- string (init $ unlines quotedLines)
+  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 :: Parser MdToken
+parseListNested = do
+  let firstCharParser = string (T.pack "    ") <|> string (T.pack "\t")
+  let restOfLineParser = manyTill anySingle (void (char '\n') <|> eof)
+  lines <- greedyParse1 (firstCharParser *> restOfLineParser)
+  let linesParsed = leftmostLongestParse (parseUnorderedList <|> parseOrderedList) (init $ unlines lines)
+  when (null (show linesParsed)) empty
+  return linesParsed
+
+-- Parse an unordered list line item.
+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 :: Parser MdToken
+parseOListLineItem = do
+  num <- greedyParse1 (satisfy isDigit)
+  char '.'
+  char ' ' -- At least one space between list indicator and list text.
+  parseListLineItemCommon
+
+-- Common code for parsing list line items
+parseListLineItemCommon :: Parser MdToken
+parseListLineItemCommon = do
+  space
+  restOfLine <- manyTill parseListLineToken (void (char '\n') <|> eof)
+  nestedList <- try parseListNested <|> return (Unit "")
+  return $ Line [Line restOfLine, nestedList]
+
+-- Parse an unordered list paragraph item.
+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 :: Parser MdToken
+parseOListParaItem = do
+  firstLine <- parseOListLineItem
+  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.
+
+-- Common code for parsing list paragraph items.
+-- 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 :: Parser [MdToken]
+parseListParaItemCommon = do
+  char '\n'
+  lines <- greedyParse1 ((string (T.pack "    ") <|> string (T.pack "\t")) *> parseTillEol)
+  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 :: Parser MdToken
+parseUListItem = space *> (try parseUListParaItem <|> parseUListLineItem)
+
+-- Parse an unordered list.
+parseUnorderedList :: Parser MdToken
+parseUnorderedList = do
+  lineItems <- some $ try parseUListItem
+  void (char '\n') <|> eof -- A list must end in an extra newline or eof
+  return $ UnordList lineItems
+
+-- --------
+
+parseOListItem :: Parser MdToken
+parseOListItem = space *> (try parseOListParaItem <|> parseOListLineItem)
+
+-- Parses the first element of an ordered list, which must start with '1.'
+parseFirstOListItem :: Parser MdToken
+parseFirstOListItem = do
+  space
+  remaining <- getInput
+  when (take 2 (T.unpack remaining) /= "1.") empty
+  parseOListLineItem
+
+parseOrderedList :: Parser MdToken
+parseOrderedList = do
+  firstLine <- try parseFirstOListItem
+  lineItems <- many $ try parseOListItem
+  void (char '\n') <|> eof
+  return $ OrdList (firstLine : lineItems)
+
+horizontalRuleText :: T.Text
+horizontalRuleText = T.pack "---"
+
+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,
+    parseOrderedList,
+    parseFigure,
+    parsePara
+  ]
 
 -- Parse a document, which is multiple paragraphs.
-parseDocument :: ReadP MdToken
+parseDocument :: Parser MdToken
-parseDocument = (many1 parsePara) >>= (\res -> return (Document (res)))
+parseDocument = do
+  res <- manyTill (fallthroughParser documentParsers) eof
+  return (Document res)

src/MdToHtmlTest.hs

@@ -0,0 +1,177 @@
+module MdToHtmlTest where
+
+import MdToHTML
+import Test.HUnit
+
+check_equal :: String -> String -> String -> Test
+check_equal desc expected actual = TestCase (assertEqual desc expected actual)
+
+convert :: String -> String
+convert md = show $ leftmostLongestParse parseDocument md
+
+headerTests =
+  TestList
+    [ check_equal "Should convert H1 heading" "<h1>Hello</h1>" (convert "# Hello"),
+      check_equal "Should convert H2 heading" "<h2>Hello</h2>" (convert "## Hello"),
+      check_equal "Should convert H3 heading" "<h3>Hello</h3>" (convert "### Hello"),
+      check_equal "Should convert H4 heading" "<h4>Hello</h4>" (convert "#### Hello"),
+      check_equal "Should convert H5 heading" "<h5>Hello</h5>" (convert "##### Hello"),
+      check_equal "Should convert H6 heading" "<h6>Hello</h6>" (convert "###### Hello")
+    ]
+
+boldTests =
+  TestList
+    [ 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><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.")
+    ]
+
+linkTests =
+  TestList
+    [ check_equal "Should convert normal link" "<p><a href=\"https://example.com\">This is an example link.</a></p>" (convert "[This is an example link.](https://example.com)"),
+      check_equal "Should convert styled link" "<p><a href=\"https://example.com\"><b>Fancy</b>!!!</a></p>" (convert "[__Fancy__!!!](https://example.com)")
+    ]
+
+escapedCharTests =
+  TestList
+    [ check_equal "Should print literal underscore" "<p>This is an underscore - _</p>" (convert "This is an underscore - \\_"),
+      check_equal "Should print literal asterisk" "<p>This is an asterisk - *</p>" (convert "This is an asterisk - \\*"),
+      check_equal "Should print literal asterisk in bold" "<p>This is a bolded asterisk - <b>*</b></p>" (convert "This is a bolded asterisk - **\\***")
+    ]
+
+blockquoteTests =
+  TestList
+    [ check_equal "Should wrap para in blockquote" "<blockquote><p>What a <b>truly</b> <i>lovely</i> day!!!</p></blockquote>" (convert "> What a __truly__ _lovely_ day!!!"),
+      check_equal "Simple nested blockquotes" "<blockquote><p>Hello</p><blockquote><p>World</p></blockquote></blockquote>" (convert "> Hello\n>\n>> World"),
+      check_equal
+        "Nested blockquotes"
+        "<blockquote><p>Dorothy followed her through many \
+        \of the beautiful rooms in her castle.</p><blockquote><p>The Witch \
+        \bade her clean the pots and kettles and sweep the floor and keep the fire \
+        \fed with wood.</p></blockquote></blockquote>"
+        ( convert
+            "> Dorothy followed her through many of the \
+            \beautiful rooms in her castle.\n> \n>> The Witch bade her \
+            \clean the pots and kettles and sweep the floor and keep the fire fed with wood."
+        )
+    ]
+
+unorderedListTests =
+  TestList
+    [ check_equal "Basic unordered list" "<ul><li>Item 1</li><li>Item 2</li><li>Item 3</li></ul>" (convert "* Item 1\n* Item 2\n* Item 3"),
+      check_equal "Mixing list indicators" "<ul><li>Item 1</li><li>Item 2</li><li>Item 3</li></ul>" (convert "* Item 1\n+ Item 2\n- Item 3"),
+      check_equal "Formatted lists" "<ul><li><b>Item 1</b></li><li><i>Item 2</i></li><li><b><i>Item 3</i></b></li></ul>" (convert "* __Item 1__\n+ _Item 2_\n- ***Item 3***"),
+      check_equal "Nested list" "<ul><li>Item 1</li><li>Item 2</li><li>Item 3<ul><li>Subitem 1</li><li>Subitem 2</li></ul></li></ul>" (convert "* Item 1\n* Item 2\n* Item 3\n    * Subitem 1\n    * Subitem 2"),
+      check_equal "Paragraph in list" "<ul><li>Item 1</li><li><p>Item 2</p><p>More stuff</p></li><li>Item 3</li></ul>" (convert "- Item 1\n- Item 2\n\n    More stuff\n\n- Item 3"),
+      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 "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")
+    ]
+
+orderedListTests =
+  TestList
+    [ check_equal "Basic ordered list" "<ol><li>Item 1</li><li>Item 2</li><li>Item 3</li></ol>" (convert "1. Item 1\n2. Item 2\n3. Item 3"),
+      check_equal "Mixing list numbering" "<ol><li>Item 1</li><li>Item 2</li><li>Item 3</li></ol>" (convert "1. Item 1\n3. Item 2\n2. Item 3"),
+      check_equal "Should not convert list without number 1" "<p>2. Item 1 1. Item 2</p>" (convert "2. Item 1\n1. Item 2"),
+      check_equal "Formatted lists" "<ol><li><b>Item 1</b></li><li><i>Item 2</i></li><li><b><i>Item 3</i></b></li></ol>" (convert "1. __Item 1__\n2. _Item 2_\n3. ***Item 3***"),
+      check_equal "Nested list" "<ol><li>Item 1</li><li>Item 2</li><li>Item 3<ol><li>Subitem 1</li><li>Subitem 2</li></ol></li></ol>" (convert "1. Item 1\n2. Item 2\n3. Item 3\n    1. Subitem 1\n    2. Subitem 2"),
+      check_equal "Paragraph in list" "<ol><li>Item 1</li><li><p>Item 2</p><p>More stuff</p></li><li>Item 3</li></ol>" (convert "1. Item 1\n2. Item 2\n\n    More stuff\n\n1. Item 3"),
+      check_equal "Paragraph before list" "<p>This is a list</p><ol><li>Item 1</li><li>Item 2</li></ol>" (convert "This is a list\n\n1. Item 1\n1. Item 2"),
+      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 "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>`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 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 =
+  TestList
+    [ check_equal "Image by itself" "<figure><img src=\"img.png\" alt=\"Image 1\"/><figcaption aria-hidden=\"true\">Image 1</figcaption></figure>" (convert "![Image 1](img.png)")
+    ]
+
+horizontalRuleTests =
+  TestList
+    [check_equal "Horizontal Rule" "<p>a</p><hr><p>b</p>" (convert "a\n\n---\n\nb")]
+
+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_"),
+      check_equal "Integration 2" "<p><b>Hello</b> <i>World</i></p>" (convert "__Hello__\n_World_"),
+      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 7"
+        "<h1>Sample Markdown</h1><p>This is some basic, sample markdown.</p><h2>Second \
+        \Heading</h2><ul><li>Unordered lists, and:<ol><li>One</li><li>Two</li><li>\
+        \Three</li></ol></li><li>More</li></ul><blockquote><p>Blockquote</p>\
+        \</blockquote><p>And <b>bold</b>, <i>italics</i>, and even <i>italics \
+        \and later <b>bold</b></i>. Even <s>strikethrough</s>. \
+        \<a href=\"https://markdowntohtml.com\">A link</a> to somewhere.</p>"
+        ( convert
+            "# 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."
+        )
+    ]
+
+tests =
+  TestList
+    [ headerTests,
+      boldTests,
+      strikethroughTests,
+      linkTests,
+      escapedCharTests,
+      blockquoteTests,
+      unorderedListTests,
+      orderedListTests,
+      imageTests,
+      htmlTests,
+      figureTests,
+      codeTests,
+      horizontalRuleTests,
+      tableTests,
+      integrationTests
+    ]
+
+runTests = runTestTT tests

src/Test.hs

@@ -1,44 +0,0 @@
-module MdToHtmlTest where
-
-import MdToHTML
-import Test.HUnit
-
-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
-
-headerTests = TestList
-      [
-         check_equal "Should convert H1 heading" "<h1>Hello</h1>" (convert "# Hello"),
-         check_equal "Should convert H2 heading" "<h2>Hello</h2>" (convert "## Hello"),
-         check_equal "Should convert H3 heading" "<h3>Hello</h3>" (convert "### Hello"),
-         check_equal "Should convert H4 heading" "<h4>Hello</h4>" (convert "#### Hello"),
-         check_equal "Should convert H5 heading" "<h5>Hello</h5>" (convert "##### Hello"),
-         check_equal "Should convert H6 heading" "<h6>Hello</h6>" (convert "###### Hello")
-      ]
-
-boldTests = TestList
-   [
-         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")
-   ]
-
-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_"),
-        check_equal "Integration 2" "<p><b>Hello</b> <i>World</i></p>" (convert "__Hello__\n_World_"),
-        check_equal "Integration 3" "<h1>Hello</h1><p>World</p>" (convert "# Hello\nWorld")
-    ]
-
-
-tests = TestList
-   [
-      headerTests,
-      boldTests,
-      integrationTests
-   ]
-
-runTests = runTestTT tests

test.hs

@@ -1,5 +0,0 @@
-leftmostLongest :: (Foldable t) => [t a] -> t a
-leftmostLongest xs = 
-        let lastElem = (last xs)
-            filteredLst = (filter (\val -> (length val) == (length lastElem)) xs)
-        in head filteredLst