Rss-Bridge/lib/html.php

<?php
/**
 * This file is part of RSS-Bridge, a PHP project capable of generating RSS and
 * Atom feeds for websites that don't have one.
 *
 * For the full license information, please view the UNLICENSE file distributed
 * with this source code.
 *
 * @package	Core
 * @license	http://unlicense.org/ UNLICENSE
 * @link	https://github.com/rss-bridge/rss-bridge
 */

/**
 * Removes unwanted tags from a given HTML text.
 *
 * @param string $textToSanitize The HTML text to sanitize.
 * @param array $removedTags A list of tags to remove from the DOM.
 * @param array $keptAttributes A list of attributes to keep on tags (other
 * attributes are removed).
 * @param array $keptText A list of tags where the innertext replaces the tag
 * (i.e. `<p>Hello World!</p>` becomes `Hello World!`).
 * @return object A simplehtmldom object of the remaining contents.
 *
 * @todo Check if this implementation is still necessary, because simplehtmldom
 * already removes some of the tags (search for `remove_noise` in simple_html_dom.php).
 * @todo Rename parameters to make more sense. `$textToSanitize` must be HTML,
 * `$removedTags`, `$keptAttributes` and `$keptText` are past tense.
 */
function sanitize($textToSanitize,
$removedTags = array('script', 'iframe', 'input', 'form'),
$keptAttributes = array('title', 'href', 'src'),
$keptText = array()){
	$htmlContent = str_get_html($textToSanitize);

	/*
	 * Notice: simple_html_dom currently doesn't support "->find(*)", which is a
	 * known issue: https://sourceforge.net/p/simplehtmldom/bugs/157/
	 *
	 * A solution to this is to find all nodes WITHOUT a specific attribute. If
	 * the attribute is very unlikely to appear in the DOM, this is essentially
	 * returning all nodes.
	 *
	 * "*[!b38fd2b1fe7f4747d6b1c1254ccd055e]" is doing exactly that. The attrib
	 * "b38fd2b1fe7f4747d6b1c1254ccd055e" is very unlikely to appear in any DOM.
	 */
	foreach($htmlContent->find('*[!b38fd2b1fe7f4747d6b1c1254ccd055e]') as $element) {
		if(in_array($element->tag, $keptText)) {
			$element->outertext = $element->plaintext;
		} elseif(in_array($element->tag, $removedTags)) {
			$element->outertext = '';
		} else {
			foreach($element->getAllAttributes() as $attributeName => $attribute) {
				if(!in_array($attributeName, $keptAttributes))
					$element->removeAttribute($attributeName);
			}
		}
	}

	return $htmlContent;
}

/**
 * Replace background by image
 *
 * Replaces tags with styles of `backgroud-image` by `<img />` tags.
 *
 * For example:
 *
 * ```HTML
 * <html>
 *   <body style="background-image: url('bgimage.jpg');">
 *     <h1>Hello world!</h1>
 *   </body>
 * </html>
 * ```
 *
 * results in this output:
 *
 * ```HTML
 * <html>
 *   <img style="display:block;" src="bgimage.jpg" />
 * </html>
 * ```
 *
 * @param string $htmlContent The HTML content
 * @return string The HTML content with all ocurrences replaced
 */
function backgroundToImg($htmlContent) {

	$regex = '/background-image[ ]{0,}:[ ]{0,}url\([\'"]{0,}(.*?)[\'"]{0,}\)/';
	$htmlContent = str_get_html($htmlContent);

	/*
	 * Notice: simple_html_dom currently doesn't support "->find(*)", which is a
	 * known issue: https://sourceforge.net/p/simplehtmldom/bugs/157/
	 *
	 * A solution to this is to find all nodes WITHOUT a specific attribute. If
	 * the attribute is very unlikely to appear in the DOM, this is essentially
	 * returning all nodes.
	 *
	 * "*[!b38fd2b1fe7f4747d6b1c1254ccd055e]" is doing exactly that. The attrib
	 * "b38fd2b1fe7f4747d6b1c1254ccd055e" is very unlikely to appear in any DOM.
	 */
	foreach($htmlContent->find('*[!b38fd2b1fe7f4747d6b1c1254ccd055e]') as $element) {

		if(preg_match($regex, $element->style, $matches) > 0) {

			$element->outertext = '<img style="display:block;" src="' . $matches[1] . '" />';

		}

	}

	return $htmlContent;

}

/**
 * Convert relative links in HTML into absolute links
 *
 * This function is based on `php-urljoin`.
 *
 * @link https://github.com/plaidfluff/php-urljoin php-urljoin
 *
 * @param string|object $content The HTML content. Supports HTML objects or string objects
 * @param string $server Fully qualified URL to the page containing relative links
 * @return object Content with fixed URLs.
 *
 * @todo If the input type was a string, this function should return a string as
 * well. This is currently done implicitly by how the simplehtmldom object works.
 */
function defaultLinkTo($content, $server){
	$string_convert = false;
	if (is_string($content)) {
		$string_convert = true;
		$content = str_get_html($content);
	}

	foreach($content->find('img') as $image) {
		$image->src = urljoin($server, $image->src);
	}

	foreach($content->find('a') as $anchor) {
		$anchor->href = urljoin($server, $anchor->href);
	}

	if ($string_convert) {
		$content = $content->outertext;
	}

	return $content;
}

/**
 * Extract the first part of a string matching the specified start and end delimiters
 *
 * @param string $string Input string, e.g. `<div>Post author: John Doe</div>`
 * @param string $start Start delimiter, e.g. `author: `
 * @param string $end End delimiter, e.g. `<`
 * @return string|bool Extracted string, e.g. `John Doe`, or false if the
 * delimiters were not found.
 */
function extractFromDelimiters($string, $start, $end) {
	if (strpos($string, $start) !== false) {
		$section_retrieved = substr($string, strpos($string, $start) + strlen($start));
		$section_retrieved = substr($section_retrieved, 0, strpos($section_retrieved, $end));
		return $section_retrieved;
	} return false;
}

/**
 * Remove one or more part(s) of a string using a start and end delmiters
 *
 * @param string $string Input string, e.g. `foo<script>superscript()</script>bar`
 * @param string $start Start delimiter, e.g. `<script`
 * @param string $end End delimiter, e.g. `</script>`
 * @return string Cleaned string, e.g. `foobar`
 */
function stripWithDelimiters($string, $start, $end) {
	while(strpos($string, $start) !== false) {
		$section_to_remove = substr($string, strpos($string, $start));
		$section_to_remove = substr($section_to_remove, 0, strpos($section_to_remove, $end) + strlen($end));
		$string = str_replace($section_to_remove, '', $string);
	}
	return $string;
}

/**
 * Remove HTML sections containing one or more sections using the same HTML tag
 *
 * @param string $string Input string, e.g. `foo<div class="ads"><div>ads</div>ads</div>bar`
 * @param string $tag_name Name of the HTML tag, e.g. `div`
 * @param string $tag_start Start of the HTML tag to remove, e.g. `<div class="ads">`
 * @return string Cleaned String, e.g. `foobar`
 *
 * @todo This function needs more documentation to make it maintainable.
 */
function stripRecursiveHTMLSection($string, $tag_name, $tag_start){
	$open_tag = '<' . $tag_name;
	$close_tag = '</' . $tag_name . '>';
	$close_tag_length = strlen($close_tag);
	if(strpos($tag_start, $open_tag) === 0) {
		while(strpos($string, $tag_start) !== false) {
			$max_recursion = 100;
			$section_to_remove = null;
			$section_start = strpos($string, $tag_start);
			$search_offset = $section_start;
			do {
				$max_recursion--;
				$section_end = strpos($string, $close_tag, $search_offset);
				$search_offset = $section_end + $close_tag_length;
				$section_to_remove = substr($string, $section_start, $section_end - $section_start + $close_tag_length);
				$open_tag_count = substr_count($section_to_remove, $open_tag);
				$close_tag_count = substr_count($section_to_remove, $close_tag);
			} while ($open_tag_count > $close_tag_count && $max_recursion > 0);
			$string = str_replace($section_to_remove, '', $string);
		}
	}
	return $string;
}

/**
 * Convert Markdown into HTML. Only a subset of the Markdown syntax is implemented.
 *
 * @link https://daringfireball.net/projects/markdown/ Markdown
 * @link https://github.github.com/gfm/ GitHub Flavored Markdown Spec
 *
 * @param string $string Input string in Markdown format
 * @return string output string in HTML format
 */
function markdownToHtml($string) {

	//For more details about how these regex work:
	// https://github.com/RSS-Bridge/rss-bridge/pull/802#discussion_r216138702
	// Images: https://regex101.com/r/JW9Evr/1
	// Links: https://regex101.com/r/eRGVe7/1
	// Bold: https://regex101.com/r/2p40Y0/1
	// Italic: https://regex101.com/r/xJkET9/1
	// Separator: https://regex101.com/r/ZBEqFP/1
	// Plain URL: https://regex101.com/r/2JHYwb/1
	// Site name: https://regex101.com/r/qIuKYE/1

	$string = preg_replace('/\!\[([^\]]+)\]\(([^\) ]+)(?: [^\)]+)?\)/', '<img src="$2" alt="$1" />', $string);
	$string = preg_replace('/\[([^\]]+)\]\(([^\)]+)\)/', '<a href="$2">$1</a>', $string);
	$string = preg_replace('/\*\*(.*)\*\*/U', '<b>$1</b>', $string);
	$string = preg_replace('/\*(.*)\*/U', '<i>$1</i>', $string);
	$string = preg_replace('/__(.*)__/U', '<b>$1</b>', $string);
	$string = preg_replace('/_(.*)_/U', '<i>$1</i>', $string);
	$string = preg_replace('/[-]{6,99}/', '<hr />', $string);
	$string = str_replace('&#10;', '<br />', $string);
	$string = preg_replace('/([^"])(https?:\/\/[^ "<]+)([^"])/', '$1<a href="$2">$2</a>$3', $string . ' ');
	$string = preg_replace('/([^"\/])(www\.[^ "<]+)([^"])/', '$1<a href="http://$2">$2</a>$3', $string . ' ');

	//As the regex are not perfect, we need to fix <i> and </i> that are introduced in URLs
	// Fixup regex <i>: https://regex101.com/r/NTRPf6/1
	// Fixup regex </i>: https://regex101.com/r/aNklRp/1

	$count = 1;
	while($count > 0) {
		$string = preg_replace('/ (src|href)="([^"]+)<i>([^"]+)"/U', ' $1="$2_$3"', $string, -1, $count);
	}

	$count = 1;
	while($count > 0) {
		$string = preg_replace('/ (src|href)="([^"]+)<\/i>([^"]+)"/U', ' $1="$2_$3"', $string, -1, $count);
	}

	return '<div>' . trim($string) . '</div>';
}