[Mime] Add a way to control the HTML to text conversion

fabpot · fabpot · commit 3058b6561393 · 2022-08-07T15:41:23.000+02:00
diff --git a/HtmlToTextConverter/DefaultHtmlToTextConverter.php b/HtmlToTextConverter/DefaultHtmlToTextConverter.php
@@ -0,0 +1,23 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Mime\HtmlToTextConverter;
+
+/**
+ * @author Fabien Potencier <fabien@symfony.com>
+ */
+class DefaultHtmlToTextConverter implements HtmlToTextConverterInterface
+{
+    public function convert(string $html, string $charset): string
+    {
+        return strip_tags(preg_replace('{<(head|style)\b.*?</\1>}is', '', $html));
+    }
+}
diff --git a/HtmlToTextConverter/HtmlToTextConverterInterface.php b/HtmlToTextConverter/HtmlToTextConverterInterface.php
@@ -0,0 +1,25 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Mime\HtmlToTextConverter;
+
+/**
+ * @author Fabien Potencier <fabien@symfony.com>
+ */
+interface HtmlToTextConverterInterface
+{
+    /**
+     * Converts and HTML representation of a Message to a text representation.
+     *
+     * The output must use the same charset as the HTML one.
+     */
+    public function convert(string $html, string $charset): string;
+}
diff --git a/HtmlToTextConverter/LeagueHtmlToMarkdownConverter.php b/HtmlToTextConverter/LeagueHtmlToMarkdownConverter.php
@@ -0,0 +1,35 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Mime\HtmlToTextConverter;
+
+use League\HTMLToMarkdown\HtmlConverter;
+use League\HTMLToMarkdown\HtmlConverterInterface;
+
+/**
+ * @author Fabien Potencier <fabien@symfony.com>
+ */
+class LeagueHtmlToMarkdownConverter implements HtmlToTextConverterInterface
+{
+    public function __construct(
+        private HtmlConverterInterface $converter = new HtmlConverter([
+            'hard_break' => true,
+            'strip_tags' => true,
+            'remove_nodes' => 'head style',
+        ]),
+    ) {
+    }
+
+    public function convert(string $html, string $charset): string
+    {
+        return $this->converter->convert($html);
+    }
+}
diff --git a/Tests/HtmlToTextConverter/DefaultHtmlToTextConverterTest.php b/Tests/HtmlToTextConverter/DefaultHtmlToTextConverterTest.php
@@ -0,0 +1,24 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Mime\Tests\HtmlToTextConverter;
+
+use PHPUnit\Framework\TestCase;
+use Symfony\Component\Mime\HtmlToTextConverter\DefaultHtmlToTextConverter;
+
+class DefaultHtmlToTextConverterTest extends TestCase
+{
+    public function testConvert()
+    {
+        $converter = new DefaultHtmlToTextConverter();
+        $this->assertSame('HTML', $converter->convert('<head><meta charset="utf-8"></head><b>HTML</b><style>css</style>', 'UTF-8'));
+    }
+}
diff --git a/Tests/HtmlToTextConverter/LeagueHtmlToMarkdownConverterTest.php b/Tests/HtmlToTextConverter/LeagueHtmlToMarkdownConverterTest.php
@@ -0,0 +1,24 @@
+<?php
+
+/*
+ * This file is part of the Symfony package.
+ *
+ * (c) Fabien Potencier <fabien@symfony.com>
+ *
+ * For the full copyright and license information, please view the LICENSE
+ * file that was distributed with this source code.
+ */
+
+namespace Symfony\Component\Mime\Tests\HtmlToTextConverter;
+
+use PHPUnit\Framework\TestCase;
+use Symfony\Component\Mime\HtmlToTextConverter\LeagueHtmlToMarkdownConverter;
+
+class LeagueHtmlToMarkdownConverterTest extends TestCase
+{
+    public function testConvert()
+    {
+        $converter = new LeagueHtmlToMarkdownConverter();
+        $this->assertSame('**HTML**', $converter->convert('<head><meta charset="utf-8"></head><b>HTML</b><style>css</style>', 'UTF-8'));
+    }
+}
diff --git a/composer.json b/composer.json
@@ -22,6 +22,7 @@
     },
     "require-dev": {
         "egulias/email-validator": "^2.1.10|^3.1",
+        "league/html-to-markdown": "^5.0",
         "phpdocumentor/reflection-docblock": "^3.0|^4.0|^5.0",
         "symfony/dependency-injection": "^5.4|^6.0",
         "symfony/property-access": "^5.4|^6.0",
