]> git.wh0rd.org - tt-rss.git/commitdiff
readability: increase maximum source document size, reorganize the reddit plugin...
authorAndrew Dolgov <noreply@madoka.volgo-balt.ru>
Tue, 2 Aug 2016 09:25:54 +0000 (12:25 +0300)
committerAndrew Dolgov <noreply@madoka.volgo-balt.ru>
Tue, 2 Aug 2016 09:25:54 +0000 (12:25 +0300)
plugins/af_readability/init.php
plugins/af_redditimgur/init.php

index f5b764d9c1220dca5e5c92496dfcc781fc7b45f2..f89576e3be10fcc13c0558bc6891e9baec0d59ac 100755 (executable)
@@ -159,7 +159,7 @@ class Af_Readability extends Plugin {
 
                $tmp = fetch_file_contents($url);
 
-               if ($tmp && mb_strlen($tmp) < 65535 * 4) {
+               if ($tmp && mb_strlen($tmp) < 1024 * 500) {
                        $tmpdoc = new DOMDocument("1.0", "UTF-8");
 
                        if (!$tmpdoc->loadHTML($tmp))
index e6e2a1709585f3e2706da620921d89cfdb917f5e..e19220bad4c7190aa5808fc7088787afb8fa1727 100755 (executable)
@@ -348,75 +348,12 @@ class Af_RedditImgur extends Plugin {
 
                        $found = $this->inline_stuff($article, $doc, $xpath);
 
-                       if (!defined('NO_CURL') && function_exists("curl_init") && !$found && $this->host->get($this, "enable_readability") &&
-                               mb_strlen(strip_tags($article["content"])) <= 150) {
-
-                               if (!class_exists("Readability")) require_once(dirname(dirname(__DIR__)). "/lib/readability/Readability.php");
-
-                               if ($content_link &&
-                                       strpos($content_link->getAttribute("href"), "twitter.com") === FALSE &&
-                                       strpos($content_link->getAttribute("href"), "youtube.com") === FALSE &&
-                                       strpos($content_link->getAttribute("href"), "reddit.com") === FALSE) {
-
-                                       /* link may lead to a huge video file or whatever, we need to check content type before trying to
-                                       parse it which p much requires curl */
-
-                                       $useragent_compat = "Mozilla/5.0 (compatible; MSIE 10.0; Windows NT 6.1; WOW64; Trident/6.0)";
-
-                                       $content_type = $this->get_content_type($content_link->getAttribute("href"), $useragent_compat);
-
-                                       if ($content_type && strpos($content_type, "text/html") !== FALSE) {
-
-                                               $tmp = fetch_file_contents(array("url" => $content_link->getAttribute("href"),
-                                                       "useragent" => $useragent_compat));
-
-                                               //_debug("tmplen: " . mb_strlen($tmp));
-
-                                               if ($tmp && mb_strlen($tmp) < 1024 * 250) {
-
-                                                       $r = new Readability($tmp, $content_link->getAttribute("href"));
-
-                                                       if ($r->init()) {
-
-                                                               $tmpxpath = new DOMXPath($r->dom);
-
-                                                               $entries = $tmpxpath->query('(//a[@href]|//img[@src])');
-
-                                                               foreach ($entries as $entry) {
-                                                                       if ($entry->hasAttribute("href")) {
-                                                                               $entry->setAttribute("href",
-                                                                                       rewrite_relative_url($content_link->getAttribute("href"), $entry->getAttribute("href")));
-
-                                                                       }
-
-                                                                       if ($entry->hasAttribute("src")) {
-                                                                               $entry->setAttribute("src",
-                                                                                       rewrite_relative_url($content_link->getAttribute("href"), $entry->getAttribute("src")));
-
-                                                                       }
-
-                                                               }
-
-                                                               $article["content"] = $r->articleContent->innerHTML . "<hr/>" . $article["content"];
-
-                                                               // prob not a very good idea (breaks wikipedia pages, etc) -
-                                                               // inliner currently is not really fit for any random web content
-
-                                                               //$doc = new DOMDocument();
-                                                               //@$doc->loadHTML($article["content"]);
-                                                               //$xpath = new DOMXPath($doc);
-                                                               //$found = $this->inline_stuff($article, $doc, $xpath);
-                                                       }
-                                               }
-                                       }
-                               }
-
-                       }
-
                        $node = $doc->getElementsByTagName('body')->item(0);
 
                        if ($node && $found) {
                                $article["content"] = $doc->saveXML($node);
+                       } else if ($content_link) {
+                               $article = $this->readability($article, $content_link->getAttribute("href"), $doc, $xpath);
                        }
                }
 
@@ -466,12 +403,21 @@ class Af_RedditImgur extends Plugin {
                @$doc->loadHTML("<html><body><a href=\"$url\">[link]</a></body>");
                $xpath = new DOMXPath($doc);
 
-               print "Inline result: " . $this->inline_stuff([], $doc, $xpath, true) . "\n";
+               $found = $this->inline_stuff([], $doc, $xpath, true);
+
+               print "Inline result: $found\n";
+
+               if (!$found) {
+                       print "\nReadability result:\n";
 
-               print "\nResulting HTML:\n";
+                       $article = $this->readability([], $url, $doc, $xpath, true);
 
-               print $doc->saveHTML();
+                       print_r($article);
+               } else {
+                       print "\nResulting HTML:\n";
 
+                       print $doc->saveHTML();
+               }
        }
 
        private function get_content_type($url, $useragent = SELF_USER_AGENT) {
@@ -492,5 +438,66 @@ class Af_RedditImgur extends Plugin {
 
                return $content_type;
        }
+
+       private function readability($article, $url, $doc, $xpath, $debug = false) {
+
+               if (!defined('NO_CURL') && function_exists("curl_init") && $this->host->get($this, "enable_readability") &&
+                       mb_strlen(strip_tags($article["content"])) <= 150) {
+
+                       if (!class_exists("Readability")) require_once(dirname(dirname(__DIR__)). "/lib/readability/Readability.php");
+
+                       if ($url &&
+                               strpos($url, "twitter.com") === FALSE &&
+                               strpos($url, "youtube.com") === FALSE &&
+                               strpos($url, "reddit.com") === FALSE) {
+
+                               /* link may lead to a huge video file or whatever, we need to check content type before trying to
+                               parse it which p much requires curl */
+
+                               $useragent_compat = "Mozilla/5.0 (compatible; MSIE 10.0; Windows NT 6.1; WOW64; Trident/6.0)";
+
+                               $content_type = $this->get_content_type($url, $useragent_compat);
+
+                               if ($content_type && strpos($content_type, "text/html") !== FALSE) {
+
+                                       $tmp = fetch_file_contents(array("url" => $url,
+                                               "useragent" => $useragent_compat));
+
+                                       if ($debug) _debug("tmplen: " . mb_strlen($tmp));
+
+                                       if ($tmp && mb_strlen($tmp) < 1024 * 500) {
+
+                                               $r = new Readability($tmp, $url);
+
+                                               if ($r->init()) {
+
+                                                       $tmpxpath = new DOMXPath($r->dom);
+
+                                                       $entries = $tmpxpath->query('(//a[@href]|//img[@src])');
+
+                                                       foreach ($entries as $entry) {
+                                                               if ($entry->hasAttribute("href")) {
+                                                                       $entry->setAttribute("href",
+                                                                               rewrite_relative_url($url, $entry->getAttribute("href")));
+
+                                                               }
+
+                                                               if ($entry->hasAttribute("src")) {
+                                                                       $entry->setAttribute("src",
+                                                                               rewrite_relative_url($url, $entry->getAttribute("src")));
+
+                                                               }
+
+                                                       }
+
+                                                       $article["content"] = $r->articleContent->innerHTML . "<hr/>" . $article["content"];
+                                               }
+                                       }
+                               }
+                       }
+               }
+
+               return $article;
+       }
 }
 ?>