From e8f3b0c8d000f48981b50a2ef95b32748bd17017 Mon Sep 17 00:00:00 2001
From: Trevor Buckner <calculuschild@gmail.com>
Date: Sat, 2 May 2020 15:03:18 -0400
Subject: [PATCH] Fix Tables with uneven columns

---
 shared/naturalcrit/markdown.js | 135 +++++++++++++++++++++++++--------
 1 file changed, 103 insertions(+), 32 deletions(-)

diff --git a/shared/naturalcrit/markdown.js b/shared/naturalcrit/markdown.js
index 266230d5b..e6e6be292 100644
--- a/shared/naturalcrit/markdown.js
+++ b/shared/naturalcrit/markdown.js
@@ -1,40 +1,112 @@
 const _ = require('lodash');
 const Markdown = require('marked');
-const renderer = new Markdown.Renderer();
-const lexer    = new Markdown.Lexer();
 
-//Processes the markdown within an HTML block if it's just a class-wrapper
-renderer.html = function (html) {
-	if(_.startsWith(_.trim(html), '<div')){
-		let openTag = html.substring(0, html.indexOf('>')+1);
-		let closeTag = '';
-		html = html.substring(html.indexOf('>')+1);
-		if(_.endsWith(_.trim(html), '</div>')){
-			closeTag = '</div>';
-			html = html.substring(0,html.lastIndexOf('</div'));
+// Copied directly from Marked helpers.js source with permission
+function splitCells(tableRow, count) {
+  // ensure that every cell-delimiting pipe has a space
+  // before it to distinguish it from an escaped pipe
+  const row = tableRow.replace(/\|/g, (match, offset, str) => {
+      let escaped = false,
+        curr = offset;
+      while (--curr >= 0 && str[curr] === '\\') escaped = !escaped;
+      if (escaped) {
+        // odd number of slashes means | is escaped
+        // so we leave it alone
+        return '|';
+      } else {
+        // add space before unescaped |
+        return ' |';
+      }
+    }),
+    cells = row.split(/ \|/);
+  let i = 0;
+
+  if (cells.length > count) {
+    cells.splice(count);
+  } else {
+    while (cells.length < count) cells.push('');
+  }
+
+  for (; i < cells.length; i++) {
+    // leading or trailing whitespace is ignored per the gfm spec
+    cells[i] = cells[i].trim().replace(/\\\|/g, '|');
+  }
+  return cells;
+}
+
+const renderer = {
+	// Adjust the way html is handled
+	html(html) {
+		// Processes the markdown within an HTML block if it's just a class-wrapper
+		if(_.startsWith(_.trim(html), '<div')){
+			let openTag = html.substring(0, html.indexOf('>')+1);
+			let closeTag = '';
+			html = html.substring(html.indexOf('>')+1);
+			if(_.endsWith(_.trim(html), '</div>')){
+				closeTag = '</div>';
+				html = html.substring(0,html.lastIndexOf('</div'));
+			}
+			return `${openTag} ${Markdown(html)} ${closeTag}`;
 		}
-		return `${openTag} ${Markdown(html)} ${closeTag}`;
-	}
 
-// Allow raw HTML tags to end without a blank line if markdown is right after
-	if(html.includes('\n')){
-		let openTag = html.substring(0, html.indexOf('>')+1);
-		if(!_.endsWith(_.trim(html), '>')){	// If there is no closing tag, parse markdown directly after
-			let remainder = html.substring(html.indexOf('>')+1);
-			return `${openTag} ${Markdown(remainder)}`;
+		// Allow raw HTML tags to end without a blank line if markdown is right after
+		if(html.includes('\n')){
+			let openTag = html.substring(0, html.indexOf('>')+1);
+			if(!_.endsWith(_.trim(html), '>')){	// If there is no closing tag, parse markdown directly after
+				let remainder = html.substring(html.indexOf('>')+1);
+				return `${openTag} ${Markdown(remainder)}`;
+			}
 		}
+
+		// Above may work better if we just force <script, <pre and <style
+		// tags to return directly instead of working around <divs>
+			/*if(_.startsWith(_.trim(html), '<script')){
+				return html;
+			}*/
+
+		return html;
 	}
-
-	// Above may work better if we just explicitly allow <script, <pre and <style
-	// tags to return directly
-		/*if(_.startsWith(_.trim(html), '<script')){
-			return html;
-		}*/
-
-	return html;
 };
 
-console.log(lexer.rules);
+const tokenizer = {
+  //Adjust tables to work even if columns are uneven
+	table(src) {
+		const cap = this.rules.block.table.exec(src);
+		if (cap) {
+			const item = {
+				type: 'table',
+				header: splitCells(cap[1].replace(/^ *| *\| *$/g, '')),
+				align: cap[2].replace(/^ *|\| *$/g, '').split(/ *\| */),
+				cells: cap[3] ? cap[3].replace(/\n$/, '').split('\n') : []
+			};
+
+			item.raw = cap[0];
+
+			let l = item.align.length;
+			let i;
+			for (i = 0; i < l; i++) {
+				if (/^ *-+: *$/.test(item.align[i])) {
+					item.align[i] = 'right';
+				} else if (/^ *:-+: *$/.test(item.align[i])) {
+					item.align[i] = 'center';
+				} else if (/^ *:-+ *$/.test(item.align[i])) {
+					item.align[i] = 'left';
+				} else {
+					item.align[i] = null;
+				}
+			}
+
+			l = item.cells.length;
+			for (i = 0; i < l; i++) {
+				item.cells[i] = splitCells(
+					item.cells[i].replace(/^ *\| *| *\| *$/g, ''),
+					item.header.length);
+			}
+
+			return item;
+		}
+	}
+};
 
 /*renderer.code = function (code, infostring, escaped) {
 	if(code == ''){
@@ -43,6 +115,8 @@ console.log(lexer.rules);
 	return code;
 }*/
 
+Markdown.use({ renderer, tokenizer });
+
 const sanatizeScriptTags = (content)=>{
 	return content
 		.replace(/<script/ig, '&lt;script')
@@ -60,10 +134,7 @@ module.exports = {
 	marked : Markdown,
 	render : (rawBrewText)=>{
 		return Markdown(
-			sanatizeScriptTags(rawBrewText),
-			{ renderer: renderer,
-				lexer: lexer }
-		);
+			sanatizeScriptTags(rawBrewText));
 	},
 
 	validate : (rawBrewText)=>{