Custom WP feed for MSN News

class-msn-news.php

<?php

// Prevent direct file access.
defined( 'ABSPATH' ) || die;

/**
 * Add MSN News feed.
 * Flush permalinks after adding this feed.
 *
 * @return void
 */
// add_action( 'after_setup_theme', function() {
// 	if ( ! class_exists( 'Mai_MSN_Feed' ) ) {
// 		return;
// 	}
//
// 	new Mai_MSN_Feed(
// 		[
// 			'slug'       => 'msn-news', // The feed name.
// 			'credits'    => true,       // Show "This post originally appeared on" credits.
// 			'query_args' => [
// 				'posts_per_page' => 36,
// 				'meta_query'     => [
// 					[
// 						'relation' => 'OR',
// 						[
// 							'key'     => 'tvn_source_url',
// 							'compare' => 'NOT EXISTS',
// 						],
// 						[
// 							'key'     => 'tvn_source_url',
// 							'value'   => '',
// 							'compare' => '=',
// 						],
// 					],
// 				],
// 			],
// 		]
// 	);
// });

/**
 * Custom MSN RSS feed class.
 *
 * @version 1.0.0
 *
 * @link (specs)         https://helpcenter.microsoftstart.com/kb/feed-specifications
 * @link (old validator) https://feeds.msn.com/evaluation
 * @link (new validator) https://www.msn.com/en-us/partnerhub/management/feed/list
 */
class Mai_MSN_Feed {
	protected $args;

	/**
	 * Construct the class.
	 */
	function __construct( $args ) {
		$this->args = wp_parse_args( $args,
			[
				'slug'       => 'msn-news', // The feed name.
				'credits'    => true,       // Add credits to the feed.
				'duration'   => 'hourly',   // Default 'hourly'. Accepts 'hourly', 'daily', 'weekly', 'monthly', 'yearly'.
				'frequency'  => 1,          // Default '1'. The frequency of RSS updates within the update period.
				'ttl'        => 15,         // Time to live in minutes.
			]
		);

		$this->args['query_args'] = wp_parse_args( $this->args['query_args'],
			[
				'post_type'      => 'post',
				'post_status'    => 'publish',
				'no_found_rows'  => true,
				'update_post_meta_cache' => false,
				'update_post_term_cache' => false,
			]
		);

		// Run hooks.
		$this->hooks();
	}

	/**
	 * Add hooks.
	 *
	 * @since 1.0.0
	 *
	 * @return void
	 */
	function hooks() {
		add_feed( $this->args['slug'], [ $this, 'render_feed' ] );

	}

	/**
	 * Custom RSS feed callback.
	 *
	 * @since 1.0.0
	 *
	 * @return void
	 */
	function render_feed() {
		/**
		 * Feed header.
		 */
		header( 'Content-Type: ' . feed_content_type( 'rss-http' ) . '; charset=' . get_option( 'blog_charset' ), true );

		/**
		 * Start RSS feed.
		 */
		echo '<?xml version="1.0" encoding="' . get_option( 'blog_charset' ) . '"?' . '>';
		?>
		<rss version="2.0"
			xmlns:atom="http://www.w3.org/2005/Atom"
			xmlns:content="http://purl.org/rss/1.0/modules/content/"
			xmlns:dc="http://purl.org/dc/elements/1.1/"
			xmlns:dcterms="http://purl.org/dc/terms/"
			xmlns:media="http://search.yahoo.com/mrss/"
			xmlns:slash="http://purl.org/rss/1.0/modules/slash/"
			xmlns:sy="http://purl.org/rss/1.0/modules/syndication/"
			xmlns:wfw="http://wellformedweb.org/CommentAPI/"
		>
			<channel>
				<title><?php bloginfo_rss( 'name' ); ?></title>
				<link><?php bloginfo_rss( 'url' ); ?></link>
				<description><?php bloginfo_rss( 'description' ); ?></description>
				<lastBuildDate><?php echo mysql2date( 'D, d M Y H:i:s +0000', get_lastpostmodified( 'GMT' ), false ); ?></lastBuildDate>
				<language><?php bloginfo_rss( 'language' ); ?></language>
				<copyright><?php echo get_bloginfo( 'name' ); ?></copyright>
				<ttl><?php echo $this->args['ttl']; ?></ttl>
				<sy:updatePeriod><?php echo apply_filters( 'rss_update_period', $this->args['duration'] ); ?></sy:updatePeriod>
				<sy:updateFrequency><?php echo apply_filters( 'rss_update_frequency', $this->args['frequency'] ); ?></sy:updateFrequency>
				<atom:link href="<?php self_link(); ?>" rel="self" type="application/rss+xml" />
				<?php do_action( 'rss2_head' ); ?>

				<?php
				$query = new WP_Query( $this->args['query_args'] );

				if ( $query->have_posts() ) {
					while ( $query->have_posts() ) : $query->the_post();
						?>
						<item>
							<title><?php the_title_rss(); ?></title>
							<link><?php the_permalink_rss(); ?></link>
							<guid isPermaLink="false"><?php the_guid(); ?></guid>
							<dc:creator><?php the_author(); ?></dc:creator>
							<pubDate><?php echo mysql2date( 'D, d M Y H:i:s +0000', get_post_time( 'Y-m-d H:i:s', true, get_the_ID() ), false ); ?></pubDate>
							<dcterms:modified><?php echo mysql2date( 'D, d M Y H:i:s +0000', get_post_modified_time( 'Y-m-d H:i:s', true, get_the_ID() ), false ); ?></dcterms:modified>
							<?php
							echo $this->get_categories();
							echo $this->get_thumbnail();
							?>
							<description><![CDATA[
								<?php
								ob_start();
								the_excerpt_rss();
								$excerpt = ob_get_clean();
								echo $this->get_content( $excerpt );
								echo $this->get_credits();
								?>
							]]></description>
							<content:encoded><![CDATA[
								<?php
								echo $this->get_content( get_the_content_feed( 'msn-news' ) );
								echo $this->get_credits();
								?>
							]]></content:encoded>
						</item>
						<?php
					endwhile;
				}
				wp_reset_postdata();
				?>
			</channel>
		</rss>
		<?php
	}

	/**
	 * Get categories for the feed.
	 * Comma-separated list of categories.
	 *
	 * @since 1.0.0
	 *
	 * @return string
	 */
	function get_categories() {
		$categories = [];
		$cats       = get_the_terms( get_the_ID(), 'category' );

		// Add categories.
		if ( $cats && ! is_wp_error( $cats ) ) {
			foreach ( $cats  as $cat ) {
				$categories[] = $cat->name;
			}
		}

		return $categories ? sprintf( '<category>%s</category>', implode( ',', $categories ) ) : '';
	}

	/**
	 * Get the post thumbnail for the feed.
	 *
	 * @since 1.0.0
	 *
	 * @return string
	 */
	function get_thumbnail() {
		$thumbnail = '';
		$image_id  = get_post_thumbnail_id();
		$image     = $image_id ? image_get_intermediate_size( $image_id, 'landscape-md' ) : '';
		$thumbnail = $image ? sprintf( '<p><img src="%s" class="type:primaryImage" /></p>', $image['url'] ) : '';

		return $thumbnail;
	}

	/**
	 * Handles content cleanup for MSN feeds.
	 *
	 * @since 1.0.0
	 *
	 * @param string $content The HTML content.
	 * @param string $feed    The feed name.
	 *
	 * @return string
	 */
	function get_content( $content ) {
		if ( ! $content || ! function_exists( 'mai_get_dom_document' ) || ! function_exists( 'mai_get_dom_html' ) ) {
			return $content;
		}

		// Set up DOMDocument.
		$dom   = mai_get_dom_document( $content );
		$xpath = new DOMXPath( $dom );

		// Get all paragraphs.
		$elements = $xpath->query( '/p' );

		// If we have paragraphs.
		if ( $elements->length ) {
			// Loop through each paragraph.
			foreach ( $elements as $node ) {
				// Loop through each child node.
				foreach ( $node->childNodes as $child ) {
					// Skip if not a link.
					if ( 'a' !== $child->nodeName ) {
						continue;
					}

					// Replace link with span.
					$new = $dom->createElement( 'span', $child->textContent );
					$new->setAttribute( 'class', 'link-removed' );
					$child->parentNode->replaceChild( $new, $child );
				}
			}
		}

		// Get figure/div wrappers to remove from twitter embeds.
		$elements = $xpath->query( '//figure[contains(concat(" ", normalize-space(@class), " "), " wp-block-embed-twitter ")]' );

		// If we have elements.
		if ( $elements->length ) {
			// Loop through each element.
			foreach ( $elements as $node ) {
				// Get the blockquote.
				$blockquote = $node->getElementsByTagName( 'blockquote' )->item(0);

				// If we have a blockquote, replace the node with it.
				if ( $blockquote ) {
					$node->parentNode->replaceChild( $blockquote, $node );
				}
			}
		}

		// Get the HTML content.
		$content = mai_get_dom_html( $dom );

		return $content;
	}

	/**
	 * Get the credits for the feed.
	 *
	 * @since 1.0.0
	 *
	 * @return string
	 */
	function get_credits() {
		// Bail if credits are disabled.
		if ( ! $this->args['credits'] ) {
			return '';
		}

		return sprintf( '<p>The post <a href="%s">%s</a> originally published on <a href="%s">%s</a>.</p>', get_permalink(), get_the_title(), home_url(), get_bloginfo( 'name' ) );
	}
}