Add support for PureContent

baraja-core · Mar 27, 2021 · 607fd47 · 607fd47
1 parent a0491ea
commit 607fd47
Show file tree

Hide file tree

Showing 2 changed files with 13 additions and 3 deletions.
diff --git a/src/ContentManager.php b/src/ContentManager.php
@@ -23,17 +23,20 @@ function (array $match) use (&$items): string {
 			$html,
 		);
 		$title = null;
-		if (preg_match('/<h1>([^<]+)<\/h1>/', $html, $htmlTitleParser)) {
-			$title = strip_tags(trim($htmlTitleParser[1] ?? ''));
+		$pureContent = null;
+		if (preg_match('/^((?:.|\n)*)<h1>([^<]+)<\/h1>((?:.|\n)*)$/', $content, $htmlTitleParser)) {
+			$title = strip_tags(trim($htmlTitleParser[2] ?? ''));
+			$pureContent = trim($htmlTitleParser[1] ?? '') . trim($htmlTitleParser[3] ?? '');
 		}
 		$perex = null;
-		if (preg_match('/<p>([^<]+)<\/p>/', $html, $htmlPerexParser)) {
+		if (preg_match('/<p>([^<]+)<\/p>/', $content, $htmlPerexParser)) {
 			$perex = strip_tags(trim($htmlPerexParser[1] ?? ''));
 		}
 
 		return new Response(
 			$html,
 			$content,
+			$pureContent ?? $content,
 			$title,
 			$perex,
 			$items,

diff --git a/src/Response.php b/src/Response.php
@@ -13,6 +13,7 @@ final class Response
 	public function __construct(
 		private string $original,
 		private string $content,
+		private string $pureContent,
 		private ?string $title,
 		private ?string $perex,
 		private array $items,
@@ -41,6 +42,12 @@ public function getContent(): string
 	}
 
 
+	public function getPureContent(): string
+	{
+		return $this->pureContent;
+	}
+
+
 	public function getTitle(): ?string
 	{
 		return $this->title;