From: Mischa POSLAWSKY <perl@shiar.org>
Date: Sun, 29 Nov 2020 03:07:33 +0000 (+0100)
Subject: page: ignore asides in article body (cq title, teaser)
X-Git-Tag: v4.5~15
X-Git-Url: http://git.shiar.nl/minimedit.git/commitdiff_plain/a5cbc9ddb6ad65b17f83d2931e3318d94e498a0e?ds=sidebyside

page: ignore asides in article body (cq title, teaser)

Prepare significant page contents in body attribute, regardless of raw.
Assume it will be used for summary text, without stripped parts in preface.

Fixes preceding aside becoming page metadata on lijtweg.nl/doc/regels.
---

diff --git a/article.inc.php b/article.inc.php
index 98cd475..6c3ca93 100644
--- a/article.inc.php
+++ b/article.inc.php
@@ -16,7 +16,7 @@ function showdate($parts)
 
 class ArchiveArticle
 {
-	public $raw, $preface, $title, $body;
+	public $raw, $title, $body; # file contents
 	public $meta = [];
 
 	function __construct($path)
@@ -41,8 +41,11 @@ class ArchiveArticle
 			$this->meta = array_combine($meta[1], $meta[2]); # [property => content]
 		}
 
-		@list ($this->preface, $this->title, $this->body) =
-			preg_split('{<h2>(.*?)</h2>\s*}s', $this->raw, 2, PREG_SPLIT_DELIM_CAPTURE);
+		// find significant contents
+		$this->body = preg_replace('{<aside\b.*?</aside>}s', '', $this->raw);
+		if (preg_match('{<h2>(.*?)</h2>\s*(.*)}s', $this->body, $titlematch)) {
+			list (, $this->title, $this->body) = $titlematch;
+		}
 	}
 
 	function __get($col)
@@ -129,11 +132,9 @@ class ArchiveArticle
 		}
 
 		# paragraph contents following the page header if any
-		$offset = strpos($this->raw, '</h2>');
-		$offset = $offset ? $offset + 5 : 0;
 		if (preg_match('{
-			\G (?> \s+ | <aside\b.*?</aside> | <div [^>]*> | \[\[[^]]*\]\] )* <p> \s* (.*?) </p>
-		}sx', $this->raw, $bodyp, 0, $offset)) {
+			\G (?> \s+ | <div [^>]*> | \[\[[^]]*\]\] )* <p> \s* (.*?) </p>
+		}sx', $this->body, $bodyp, 0)) {
 			return $bodyp[1];
 		}
 	}