X-Git-Url: http://git.shiar.nl/minimedit.git/blobdiff_plain/022d44c8ddf33f855e6ed33cae6ad9ae103e074d..ea368b7d7b7a5a853c48f6247d8dfe46107c2ced:/article.inc.php

diff --git a/article.inc.php b/article.inc.php
index 5de1492..6fcb640 100644
--- a/article.inc.php
+++ b/article.inc.php
@@ -16,10 +16,27 @@ function showdate($parts)
 
 class ArchiveArticle
 {
+	public $raw, $preface, $title, $body;
+	public $meta = [];
+
 	function __construct($path)
 	{
-		$this->page = $path;
-		$this->link = preg_replace('{(?:/index)?\.html$}', '', $path);
+		$this->page = preg_replace('{^\.(?:/|$)}', '', $path);
+		$this->link = preg_replace('{(?:/index)?\.html$}', '', $this->page);
+		if (file_exists($this->page)) {
+			$this->raw = file_get_contents($this->page);
+
+			if (preg_match_all('{
+				\G <meta \s+ property="( [^"]+ )" \s+ content="( [^"]* )" > \s*
+			}x', $this->raw, $meta)) {
+				$matchlen = array_sum(array_map('strlen', $meta[0]));
+				$this->raw = substr($this->raw, $matchlen); # delete matched contents
+				$this->meta = array_combine($meta[1], $meta[2]); # [property => content]
+			}
+
+			@list ($this->preface, $this->title, $this->body) =
+				preg_split('{<h2>(.*?)</h2>\s*}', $this->raw, 2, PREG_SPLIT_DELIM_CAPTURE);
+		}
 	}
 
 	function __get($col)
@@ -27,22 +44,10 @@ class ArchiveArticle
 		return $this->$col = $this->$col();  # run method and cache
 	}
 
-	function file()
-	{
-		if (!file_exists($this->page)) return;
-		return fopen($this->page, 'r');
-	}
-
-	function title()
-	{
-		return preg_replace('{<h2>(.*)</h2>\s*}', '\1', fgets($this->file));
-	}
-
 	function safetitle()
 	{
-		return trim(strip_tags($this->title));
+		return trim($this->meta['og:title'] ?? strip_tags($this->title));
 	}
-
 	function name()
 	{
 		return $this->safetitle ?: $this->link;
@@ -52,7 +57,6 @@ class ArchiveArticle
 	{
 		return filemtime($this->page);
 	}
-
 	function lastiso()
 	{
 		return date(DATE_ATOM, $this->last);
@@ -64,12 +68,10 @@ class ArchiveArticle
 		array_shift($ymd);
 		return $ymd;
 	}
-
 	function dateiso()
 	{
 		return implode('-', $this->dateparts()) . 'T12:00:00+02:00';
 	}
-
 	function date()
 	{
 		return showdate($this->dateparts);
@@ -77,23 +79,38 @@ class ArchiveArticle
 
 	function story()
 	{
-		$this->title;
-		$rest = fread($this->file, filesize($this->page));
 		if ( preg_match('{
 			\n (?: < (?: p | figure [^>]* ) >\s* )+ (<img\ [^>]*>) | \n <hr\ />
-		}x', $rest, $img, PREG_OFFSET_CAPTURE) ) {
+		}x', $this->body, $img, PREG_OFFSET_CAPTURE) ) {
 			# strip part after matching divider (image)
 			if (isset($img[1])) {
 				$this->img = $img[1][0];
 			}
-			return substr($rest, 0, $img[0][1]);
+			return substr($this->body, 0, $img[0][1]);
 		}
-		return $rest;
+		return $this->body;
 	}
 
 	function teaser()
 	{
-		if (preg_match('{<p>(.*?)</p>}s', $this->story, $bodyp)) {
+		if ($override = @$this->meta['og:description']) {
+			# prefer specific page description if found in metadata
+			return $override;
+		}
+
+		if (preg_match('{
+			</h2> (?: \s+ | <p\sclass="nav\b.*?</p> | <div[^>]*> )* <p> \s* (.*?) </p>
+		}sx', $this->raw, $bodyp, PREG_OFFSET_CAPTURE)) {
+			# fallback paragraph contents following the page header
+			if ($bodyp[1][1] < 256) {
+				return $bodyp[1][0];
+			}
+		}
+
+		# starting paragraph for documents without title (assumed simple/partial)
+		if (strpos($this->raw, '<h2') === FALSE and preg_match('{
+			\A <p> \s* (.*?) </p>
+		}sx', $this->raw, $bodyp)) {
 			return $bodyp[1];
 		}
 	}
@@ -104,14 +121,17 @@ class ArchiveArticle
 		$this->story;
 		return $this->img;
 	}
-
 	function image()
 	{
+		if ($override = @$this->meta['og:image']) {
+			# prefer specific page image if found in metadata
+			return $override;
+		}
+
 		if ( preg_match('/\bsrc="([^"]*)"/', $this->img, $src) ) {
 			return $src[1];
 		}
 	}
-
 	function thumb($size = '300x')
 	{
 		if (!$this->image or $this->image[0] !== '/') return;
@@ -121,3 +141,38 @@ class ArchiveArticle
 		);
 	}
 }
+
+class PageSearch
+{
+	function __construct($path = '.')
+	{
+		$this->iterator = new RecursiveCallbackFilterIterator(
+			new RecursiveDirectoryIterator($path),
+			function ($current) {
+				if ($current->getFilename()[0] === '.') {
+					# skip hidden files and directories
+					return FALSE;
+				}
+				if ($current->isLink()) {
+					# ignore symlinks, original contents only
+					return FALSE;
+				}
+				# match **/*.html
+				return $current->isDir()
+					|| preg_match('/(?<!\.inc)\.html$/', $current->getFilename());
+			}
+		);
+	}
+
+	function files()
+	{
+		# order alphabetically by link
+		$dir = iterator_to_array(new RecursiveIteratorIterator($this->iterator));
+		array_walk($dir, function (&$row, $name) {
+			# prepare values for sorting (directory index first)
+			$row = preg_replace('{/index\.html$}', '', $name);
+		});
+		asort($dir);
+		return $dir;
+	}
+}