Facebook providers based on the mobile site, fuck you

1 files changed, 107 insertions, 0 deletions

diff --git a/providers/Facebook.php b/providers/Facebook.php
new file mode 100644
index 0000000..6514cfe
--- /dev/null
+++ b/providers/Facebook.php
@@ -0,0 +1,107 @@
+<?php
+
+namespace Providers;
+
+require_once('Provider.php');
+require_once('Item.php');
+require_once('../lib/querypath/src/qp.php');
+
+class Facebook extends \Providers\Provider {
+
+    protected $_cacheTimeout = '5 minutes';
+    //protected $_cacheTimeout = '1 second';
+    private $_encoding;
+
+    public function __construct($feed, $options=[]) {
+        parent::__construct($feed, $options);
+    }
+
+    protected function _getCachePath() {
+        return '../cache/facebook.%s';
+    }
+
+    private function __getWWWContent() {
+        $header = array();
+        $header[] = 'Accept: text/xml,application/xml,application/xhtml+xml,text/html;q=0.9,text/plain;q=0.8,image/png,*/*;q=0.5';
+        $header[] = 'Cache-Control: max-age=0';
+        $header[] = 'Connection: keep-alive';
+        $header[] = 'Keep-Alive: 300';
+        $header[] = 'Accept-Charset: ISO-8859-1,utf-8;q=0.7,*;q=0.7';
+        $header[] = 'Pragma: ';
+        $ch = curl_init(sprintf('https://m.facebook.com/%s/', $this->_feed));
+        curl_setopt($ch, CURLOPT_USERAGENT, 'Mozilla/5.0 (Windows; U; Windows NT 6.0; en-US; rv:1.9.0.11) Gecko/2009060215 Firefox/3.0.11 (.NET CLR 3.5.30729)');
+        curl_setopt($ch, CURLOPT_HTTPHEADER, $header);
+        curl_setopt($ch, CURLOPT_AUTOREFERER, TRUE);
+        curl_setopt($ch, CURLOPT_RETURNTRANSFER, 1);
+        curl_setopt($ch, CURLOPT_FOLLOWLOCATION, 1);
+        curl_setopt($ch, CURLOPT_ENCODING, '');
+        curl_setopt($ch, CURLOPT_TIMEOUT, 20);
+        $page = curl_exec($ch);
+        curl_close ($ch);
+        $this->_encoding = mb_detect_encoding($page);
+        return $page;
+    }
+
+    protected function _fetchItems() {
+        $items = [];
+        $page = $this->__getWWWContent();
+        $tree = htmlqp($page, NULL, ['convert_from_encoding' => $this->_encoding, 'convert_to_encoding' => $this->_encoding]);
+        $parents = [];
+        foreach ($tree->find('div[data-ft]') as $div) {
+            $data = json_decode($div->attr()['data-ft'], TRUE);
+            if (isset($data['mf_story_key'])) {
+                $pIns = $data['page_insights'][$data['page_id']];
+                $key = $data['mf_story_key'];
+                $texts = [];
+                foreach ($div->find('h3') as $h3) {
+                    if (!$h3->find(sprintf('a[href^="/%s/"]', $this->_feed))->count() || $h3->find('a')->count() > 1) {
+                        $texts[] = trim($h3->text());
+                    }
+                }
+                foreach ($div->find('div>span p') as $p) {
+                    $texts[] = trim(strip_tags($p->html()));
+                }
+                $items[$key] = [
+                    'id' => $key,
+                    'content' => $div->html(),
+                    'texts' => $texts,
+                    'time' => $pIns['post_context']['publish_time']
+                ];
+            }
+        }
+        return array_values($items);
+    }
+
+    protected function _spamFilter($items) {
+      return $items;
+    }
+
+    protected function _mapItems($content) {
+        return array_map(
+            function ($obj) {
+                $item = new Item();
+                $item->ID = $obj['id'];
+                $item->Link = sprintf(
+                    'https://facebook.com/%s',
+                    $obj['id']
+                );
+                $item->Title = $obj['texts'][0];
+                $item->Text = implode('<br />', $obj['texts']);
+                $item->Time = $obj['time'];
+                return $item;
+            },
+            $content
+        );
+    }
+
+    protected function _sortContent($content) {
+        return $content;
+    }
+
+    public function title() {
+        return sprintf("%s's Facebook page posts", $this->_feed);
+    }
+
+}
+
+?>