From 925c70f52bff3a488208ab12263a73d5513975fa Mon Sep 17 00:00:00 2001
From: Pagwin <pgp2024@pagwin.xyz>
Date: Fri, 9 May 2025 20:32:43 -0400
Subject: [PATCH] realizing I should just use cmark and come back if I still
 don't like something later

---
 app/Markdown/Data.hs   | 39 ++++++++++++++++++++++++++-------------
 app/Markdown/Parser.hs | 42 +++++++++++-------------------------------
 2 files changed, 37 insertions(+), 44 deletions(-)

diff --git a/app/Markdown/Data.hs b/app/Markdown/Data.hs
index 91168dd..8be03a3 100644
--- a/app/Markdown/Data.hs
+++ b/app/Markdown/Data.hs
@@ -1,30 +1,43 @@
-{-# LANGUAGE RankNTypes, FlexibleContexts #-}
+{-# LANGUAGE FlexibleContexts #-}
+{-# LANGUAGE RankNTypes #-}
 
 module Markdown.Data
-  ( MarkDownElement,
-    MListItem,
+  ( MListItem,
     Parser,
+    MarkdownElement
+      ( MHeading,
+        MParagraph,
+        MBold,
+        MItalic,
+        MBoldItalic,
+        MLink,
+        MLine,
+        MUnorderedList,
+        MOrderedList,
+        Only
+      ),
   )
 where
 
 import Data.Text
 import Text.Parsec
 
---  modified version of https://github.com/tusharad/markdown-parser
-data MarkDownElement
-  = MHeading Int MarkDownElement
-  | MParagraph MarkDownElement
-  | MBold MarkDownElement
-  | MItalic MarkDownElement
-  | MBoldItalic MarkDownElement
-  | MLink MarkDownElement Text
-  | MLine [MarkDownElement]
+-- very slightly modified version of https://github.com/tusharad/markdown-parser
+data MarkdownElement
+  = MHeading Int MarkdownElement
+  | MParagraph MarkdownElement
+  | MBold MarkdownElement
+  | MItalic MarkdownElement
+  | MBoldItalic MarkdownElement
+  | MLink MarkdownElement Text
+  | MLine [MarkdownElement]
   | MUnorderedList [MListItem]
   | MOrderedList [MListItem]
+  | Raw Text -- HTML contained in the markdown
   | Only Text -- Bottom of all types
   deriving (Eq, Show)
 
-data MListItem = MListItem MarkDownElement [MarkDownElement]
+data MListItem = MListItem MarkdownElement [MarkdownElement]
   deriving (Eq, Show)
 
 type Parser v = forall s u m. (Monad m, Stream s m Char) => ParsecT s u m v
diff --git a/app/Markdown/Parser.hs b/app/Markdown/Parser.hs
index 552075e..28d9722 100644
--- a/app/Markdown/Parser.hs
+++ b/app/Markdown/Parser.hs
@@ -1,44 +1,24 @@
 {-# LANGUAGE FlexibleContexts #-}
+{-# LANGUAGE OverloadedStrings #-}
 
 module Markdown.Parser
-  ( Markdown.Parser.lines,
+  ( heading,
   )
 where
 
-import Data.Functor (void)
-import Data.Text (Text, pack)
+import Markdown.Data
 import Text.Parsec
 
 -- https://spec.commonmark.org/0.31.2/
 -- https://hackage.haskell.org/package/parsec
 
-linebreak ::
-  (Monad m, Stream s m Char) =>
-  ParsecT s u m ()
--- 2 newlines due to mark
-linebreak = void (newline *> newline)
-
-emptyParse ::
-  (Monad m, Stream s m Char) =>
-  ParsecT s u m String
-emptyParse = "" <$ notFollowedBy anyChar
-
-line ::
-  (Monad m, Stream s m Char) =>
-  ParsecT s u m Text
-line = fmap pack $ many $ notFollowedBy linebreak *> anyChar
-
-lines ::
-  (Monad m, Stream s m Char) =>
-  ParsecT s u m [Text]
-lines = line `sepBy` linebreak
-
-heading ::
-  (Monad m, Stream s m Char) =>
-  ParsecT s u m (Int, Text)
+-- only ATX headings
+heading :: Parser MarkdownElement
 heading = do
-  level <- fmap length $ many1 $ char '#'
-  text <- line
-  pure (level, text)
+  -- technically this can lead to illegal heading levels but
+  -- all the input is written by me so who cares
+  level <- fmap length $ try $ many1 $ char '#'
 
--- TODO: blockquote, single backticks, triple backticks, links, arb HTML
+  return $ MHeading level $ Only "TODO"
+
+rawHTML