Use league/commonmark Markdown renderer (#163)

Closes #155

composer.json

@@ -19,6 +19,7 @@
         "doctrine/doctrine-migrations-bundle": "^3.0",
         "doctrine/orm": "^2.7",
         "endroid/qr-code": "^4.2",
+        "league/commonmark": "^1.6",
         "league/flysystem-aws-s3-v3": "^1.0",
         "samwilson/phpflickr": "^4.14",
         "sensio/framework-extra-bundle": "^5.6",

src/Markdown.php

@@ -2,55 +2,34 @@
 
 namespace App;
 
+use League\CommonMark\Environment;
+use League\CommonMark\Extension\CommonMarkCoreExtension;
+use League\CommonMark\Extension\Autolink\AutolinkExtension;
+use League\CommonMark\MarkdownConverter;
+use League\CommonMark\Extension\SmartPunct\SmartPunctExtension;
+
 /**
- * This class formats *some parts* of Markdown, or rather CommonMark. It doesn't support everything (notably not HTML),
- * and also adds a few extra things that are good only for this application.
- *
- * @link https://commonmark.org/help/
+ * Wrapper for rendering Markdown with the league/commonmark package.
  */
 class Markdown
 {
 
     /**
-     * Convert Twyne-flavoured Markdown to HTML.
+     * Convert Markdown to HTML.
      *
      * @param string $in The Markdown source.
      * @return string The (safe) HTML.
      */
     public function toHtml(string $in): string
     {
-        // Platform-independent newlines.
-        $out = preg_replace("/(\r\n|\r)/", "\n", trim($in));
-        // Prevent HTML.
-        $out = str_replace(['&', '<', '"', "'"], ['&amp;', '&lt;', '&quot;', '&#039;'], $out);
-        $out = preg_replace("|([^\n])>|", "$1&gt;", $out);
-        // Links.
-        $out = preg_replace("|(https?://\S+)|", "<a href=\"$1\">$1</a>", $out);
-        // Paragraphs.
-        $out = "<p>$out</p>";
-        $out = preg_replace("|\n+\s*\n+|", "</p>\n\n<p>", $out);
-        // Remove paragraphs if they contain nothing (including only whitespace).
-        $out = preg_replace('|<p>\s*</p>|', '', $out);
-        // Blockquotes
-        $out = preg_replace('|<p>>\s*(.*)|', "<blockquote>\n> $1", $out);
-        $out = preg_replace('|\n>\s*(.*)</p>|', "\n$1\n</blockquote>", $out);
-        $out = preg_replace('|\n>\s(.*)|', "\n$1", $out);
-        // Separator.
-        $out = preg_replace("|<p>[-*]{3}</p>|", "<hr />", $out);
-        // Lists.
-        $out = preg_replace("|<p>\d\.(.*)|", "<ol>\n#$1", $out); // begin ordered
-        $out = preg_replace("|\n\d\.(.*)</p>|", "\n#$1\n</ol>", $out); // end ordered
-        $out = preg_replace("|<p>\*(.*)|", "<ul>\n*$1", $out); // begin unordered
-        $out = preg_replace("|\n\*(.*)</p>|", "\n*$1\n</ul>", $out); // end unordered
-        $out = preg_replace("|\n[*#]\s*(.*)|", "\n<li>$1</li>", $out); // list items
-        // Headings.
-        $out = preg_replace("|<p>#{4}\s*(.*)</p>|", "\n<h4>$1</h4>\n", $out);
-        $out = preg_replace("|<p>#{1,3}\s*(.*)</p>|", "\n<h3>$1</h3>\n", $out);
-        // Emphasis.
-        $out = preg_replace("|\*(.*?)\*|", "<em>$1</em>", $out);
-        // Monospacing.
-        $out = preg_replace("|`(.*?)`|", "<code>$1</code>", $out);
-
-        return trim($out);
+        $environment = new Environment([
+            'html_input' => 'escape',
+            'allow_unsafe_links' => true,
+        ]);
+        $environment->addExtension(new CommonMarkCoreExtension());
+        $environment->addExtension(new AutolinkExtension());
+        $environment->addExtension(new SmartPunctExtension());
+        $converter = new MarkdownConverter($environment);
+        return trim($converter->convertToHtml($in));
     }
 }

symfony.lock

@@ -163,6 +163,15 @@
     "laminas/laminas-zendframework-bridge": {
         "version": "1.3.0"
     },
+    "league/commonmark": {
+        "version": "0.0",
+        "recipe": {
+            "repo": "github.com/symfony/recipes",
+            "branch": "master",
+            "version": "0.0",
+            "ref": ""
+        }
+    },
     "league/flysystem": {
         "version": "1.1.4"
     },

tests/MarkdownTest.php

@@ -25,29 +25,29 @@ public function provideToHtml()
             // Paragraphs.
             ['foo', '<p>foo</p>'],
             ["foo\nbar", "<p>foo\nbar</p>"],
-            ["foo\n\nbar", "<p>foo</p>\n\n<p>bar</p>"],
+            ["foo\n\nbar", "<p>foo</p>\n<p>bar</p>"],
             // Lists.
-            ["auf\n\n* *emph* word\n* bar", "<p>auf</p>\n\n<ul>\n<li><em>emph</em> word</li>\n<li>bar</li>\n</ul>"],
+            ["auf\n\n* *emph* word\n* bar", "<p>auf</p>\n<ul>\n<li>\n<em>emph</em> word</li>\n<li>bar</li>\n</ul>"],
             ["1. foo\n2. bar\n", "<ol>\n<li>foo</li>\n<li>bar</li>\n</ol>"],
-            ["1. foo\n2. bar\n\nbaz", "<ol>\n<li>foo</li>\n<li>bar</li>\n</ol>\n\n<p>baz</p>"],
+            ["1. foo\n2. bar\n\nbaz", "<ol>\n<li>foo</li>\n<li>bar</li>\n</ol>\n<p>baz</p>"],
             ['Not a list 10.20 just numbers.', '<p>Not a list 10.20 just numbers.</p>'],
             // Emphasis and code.
             ['the *foo* `bar` baz', '<p>the <em>foo</em> <code>bar</code> baz</p>'],
             // Blockquotes.
-            ["foo\n\n> bar\n> baz\n>last", "<p>foo</p>\n\n<blockquote>\nbar\nbaz\nlast\n</blockquote>"],
-            ["> foo\n\nbar", "<blockquote>\nfoo\n</blockquote>\n\n<p>bar</p>"],
+            ["foo\n\n> bar\n> baz\n>last", "<p>foo</p>\n<blockquote>\n<p>bar\nbaz\nlast</p>\n</blockquote>"],
+            ["> foo\n\nbar", "<blockquote>\n<p>foo</p>\n</blockquote>\n<p>bar</p>"],
             // Separator.
-            ["foo\n\n---\n\nbar", "<p>foo</p>\n\n<hr />\n\n<p>bar</p>"],
+            ["foo\n\n---\n\nbar", "<p>foo</p>\n<hr />\n<p>bar</p>"],
             // Header.
-            ["# Foo", "<h3>Foo</h3>"],
-            ["##Foo", "<h3>Foo</h3>"],
-            ["###Foo", "<h3>Foo</h3>"],
+            ["# Foo", "<h1>Foo</h1>"],
+            ["## Foo", "<h2>Foo</h2>"],
+            ["### Foo", "<h3>Foo</h3>"],
             ["#### Foo", "<h4>Foo</h4>"],
             // Links.
             ['foo https://example.org bar', '<p>foo <a href="https://example.org">https://example.org</a> bar</p>'],
             ['https://x.net/foo_bar', '<p><a href="https://x.net/foo_bar">https://x.net/foo_bar</a></p>'],
             // No HTML.
-            ['<p>T & <stuff x="y"></p>', '<p>&lt;p&gt;T &amp; &lt;stuff x=&quot;y&quot;&gt;&lt;/p&gt;</p>']
+            ['a <p>T & <stuff x="y"></p>', '<p>a &lt;p&gt;T &amp; &lt;stuff x="y"&gt;&lt;/p&gt;</p>']
         ];
     }
 }