From 2843cc352023b07667c183463256d15bbfd41bf1 Mon Sep 17 00:00:00 2001
From: BentiGorlich <benjamin_g@posteo.de>
Date: Mon, 9 Dec 2024 13:21:22 +0100
Subject: [PATCH] Add strikethrough support Add a new optional converter
 `StrikethroughConverter`, mention it in the readme and add tests for it

---
 README.md                                | 17 ++++++-
 src/Converter/StrikethroughConverter.php | 58 ++++++++++++++++++++++++
 tests/HtmlConverterTest.php              | 12 +++++
 3 files changed, 86 insertions(+), 1 deletion(-)
 create mode 100644 src/Converter/StrikethroughConverter.php

diff --git a/README.md b/README.md
index 6bce00a..16bc88c 100644
--- a/README.md
+++ b/README.md
@@ -61,7 +61,7 @@ The included `demo` directory contains an HTML->Markdown conversion form to try
 
 ### Conversion options
 
-> [!CAUTION]  
+> [!CAUTION]
 > By default, this library preserves HTML tags without Markdown equivalents, like `<span>`, `<div>`, `<iframe>`, `<script>`, etc. If you will be parsing untrusted input from users, **please consider setting the `strip_tags` and/or `remove_nodes` options** documented below, and also using a library (like [HTML Purifier](https://github.com/ezyang/htmlpurifier)) to provide additional HTML filtering.
 
 To strip HTML tags that don't have a Markdown equivalent while preserving the content inside them, set `strip_tags` to true, like this:
@@ -197,6 +197,21 @@ $html = "<table><tr><th>A</th></tr><tr><td>a</td></tr></table>";
 $markdown = $converter->convert($html);
 ```
 
+### Strikethrough support
+
+Support for Markdown strikethrough is not enabled by default, because it is not part of the original Markdown syntax. To use strikethrough add the converter explicitly:
+
+```php
+use League\HTMLToMarkdown\HtmlConverter;
+use League\HTMLToMarkdown\Converter\StrikethroughConverter;
+
+$converter = new HtmlConverter();
+$converter->getEnvironment()->addConverter(new StrikethroughConverter());
+
+$html = "<del>Some Text</del> <strike>Some Text2<strike>";
+$markdown = $converter->convert($html);
+```
+
 ### Limitations
 
 - Markdown Extra, MultiMarkdown and other variants aren't supported – just Markdown.
diff --git a/src/Converter/StrikethroughConverter.php b/src/Converter/StrikethroughConverter.php
new file mode 100644
index 0000000..a987425
--- /dev/null
+++ b/src/Converter/StrikethroughConverter.php
@@ -0,0 +1,58 @@
+<?php
+
+declare(strict_types=1);
+
+namespace League\HTMLToMarkdown\Converter;
+
+use League\HTMLToMarkdown\Configuration;
+use League\HTMLToMarkdown\ConfigurationAwareInterface;
+use League\HTMLToMarkdown\ElementInterface;
+
+class StrikethroughConverter implements ConverterInterface, ConfigurationAwareInterface
+{
+    /** @var Configuration */
+    protected $config;
+
+    /**
+     * @return string[]
+     */
+    public function getSupportedTags(): array
+    {
+        return ['del', 'strike'];
+    }
+
+    public function setConfig(Configuration $config): void
+    {
+        $this->config = $config;
+    }
+
+    public function convert(ElementInterface $element): string
+    {
+        $value = $element->getValue();
+        if (! \trim($value)) {
+            return $value;
+        }
+
+        $prefix = \ltrim($value) !== $value ? ' ' : '';
+        $suffix = \rtrim($value) !== $value ? ' ' : '';
+
+        $previousTagName = null;
+        if (($previous = $element->getPreviousSibling()) !== null) {
+            $previousTagName = $previous->getTagName();
+        }
+
+        $nextTagName = null;
+        if (($next = $element->getNextSibling()) !== null) {
+            $nextTagName = $next->getTagName();
+        }
+
+        /* If this node is immediately preceded or followed by one of the same type don't emit
+         * the start or end $style, respectively. This prevents <del>foo</del><del>bar</del> from
+         * being converted to ~~foo~~~~bar~~ which is incorrect. We want ~~foobar~~ instead.
+         */
+        $preStyle  = \in_array($previousTagName, $this->getSupportedTags(), true) ? '' : '~~';
+        $postStyle = \in_array($nextTagName, $this->getSupportedTags(), true) ? '' : '~~';
+
+        return $prefix . $preStyle . \trim($value) . $postStyle . $suffix;
+    }
+}
diff --git a/tests/HtmlConverterTest.php b/tests/HtmlConverterTest.php
index 4b71832..353e126 100644
--- a/tests/HtmlConverterTest.php
+++ b/tests/HtmlConverterTest.php
@@ -5,6 +5,7 @@
 namespace League\HTMLToMarkdown\Test;
 
 use League\HTMLToMarkdown\Converter\ConverterInterface;
+use League\HTMLToMarkdown\Converter\StrikethroughConverter;
 use League\HTMLToMarkdown\Converter\TableConverter;
 use League\HTMLToMarkdown\Environment;
 use League\HTMLToMarkdown\HtmlConverter;
@@ -422,4 +423,15 @@ public function testInstatiationWithEnvironment(): void
         $result = $markdown->convert($htmlH4);
         $this->assertEquals($htmlH4, $result);
     }
+
+    public function testStrikethrough(): void
+    {
+        $opt  = [];
+        $conv = [new StrikethroughConverter()];
+        $this->assertHtmlGivesMarkdown('<strike>Some Text</strike>', '~~Some Text~~', $opt, $conv);
+        $this->assertHtmlGivesMarkdown('<del>Some Text</del>', '~~Some Text~~', $opt, $conv);
+        $this->assertHtmlGivesMarkdown('<del>Some</del><del> Text</del>', '~~Some Text~~', $opt, $conv);
+        $this->assertHtmlGivesMarkdown('<del>Some</del><strike> Text</strike>', '~~Some Text~~', $opt, $conv);
+        $this->assertHtmlGivesMarkdown('<del>Some</del> <strike>Text</strike>', '~~Some~~ ~~Text~~', $opt, $conv);
+    }
 }