Merge remote branch 'upstream/master'

[friendica.git] / include / items.php

diff --git a/include/items.php b/include/items.php
index a88246cd6b6b00efe6179f037cae455410c09a8c..b81208f3666bda612614011667979fb2fa119084 100755 (executable)
--- a/include/items.php
+++ b/include/items.php
@@ -374,6 +374,29 @@ function limit_body_size($body) {
                return $body;
 }}
 
+function title_is_body($title, $body) {
+
+       $title = strip_tags($title);
+       $title = trim($title);
+       $title = str_replace(array("\n", "\r", "\t", " "), array("","","",""), $title);
+
+       $body = strip_tags($body);
+       $body = trim($body);
+       $body = str_replace(array("\n", "\r", "\t", " "), array("","","",""), $body);
+
+       if (strlen($title) < strlen($body))
+               $body = substr($body, 0, strlen($title));
+
+       if (($title != $body) and (substr($title, -3) == "...")) {
+               $pos = strrpos($title, "...");
+               if ($pos > 0) {
+                       $title = substr($title, 0, $pos);
+                       $body = substr($body, 0, $pos);
+               }
+       }
+
+       return($title == $body);
+}
 
 
 
@@ -400,6 +423,11 @@ function get_atom_elements($feed,$item) {
        $res['body'] = unxmlify($item->get_content());
        $res['plink'] = unxmlify($item->get_link(0));
 
+       // removing the content of the title if its identically to the body
+       // This helps with auto generated titles e.g. from tumblr
+       if (title_is_body($res["title"], $res["body"]))
+               $res['title'] = "";
+
        if($res['plink'])
                $base_url = implode('/', array_slice(explode('/',$res['plink']),0,3));
        else
@@ -418,7 +446,7 @@ function get_atom_elements($feed,$item) {
                                        $res['author-avatar'] = unxmlify($link['attribs']['']['href']);
                        }
                }
-       }                       
+       }
 
        $rawactor = $item->get_item_tags(NAMESPACE_ACTIVITY, 'actor');
 
@@ -450,7 +478,7 @@ function get_atom_elements($feed,$item) {
                                                $res['author-avatar'] = unxmlify($link['attribs']['']['href']);
                                }
                        }
-               }                       
+               }
 
                $rawactor = $feed->get_feed_tags(NAMESPACE_ACTIVITY, 'subject');
 
@@ -475,7 +503,7 @@ function get_atom_elements($feed,$item) {
                $res['app'] = strip_tags(unxmlify($apps[0]['attribs']['']['source']));
                if($res['app'] === 'web')
                        $res['app'] = 'OStatus';
-       }                  
+       }
 
        // base64 encoded json structure representing Diaspora signature
 
@@ -618,7 +646,7 @@ function get_atom_elements($feed,$item) {
 
                foreach($base as $link) {
                        if(!x($res, 'owner-avatar') || !$res['owner-avatar']) {
-                               if($link['attribs']['']['rel'] === 'photo' || $link['attribs']['']['rel'] === 'avatar')                 
+                               if($link['attribs']['']['rel'] === 'photo' || $link['attribs']['']['rel'] === 'avatar')
                                        $res['owner-avatar'] = unxmlify($link['attribs']['']['href']);
                        }
                }
@@ -762,6 +790,12 @@ function get_atom_elements($feed,$item) {
 
        call_hooks('parse_atom', $arr);
 
+       //if (($res["title"] != "") or (strpos($res["body"], "RT @") > 0)) {
+       if (strpos($res["body"], "RT @") !== false) {
+               $debugfile = tempnam("/home/ike/log", "item-res2-");
+               file_put_contents($debugfile, serialize($arr));
+       }
+
        return $res;
 }
 
@@ -816,6 +850,12 @@ function item_store($arr,$force_parent = false) {
        if((strpos($arr['body'],'<') !== false) || (strpos($arr['body'],'>') !== false)) 
                $arr['body'] = strip_tags($arr['body']);
 
+       require_once('Text/LanguageDetect.php');
+       $naked_body = preg_replace('/\[(.+?)\]/','',$arr['body']);
+       $l = new Text_LanguageDetect;
+       $lng = $l->detectConfidence($naked_body);
+       $arr['postopts'] = (($lng['language']) ? 'lang=' . $lng['language'] . ';' . $lng['confidence'] : '');
+
 
        $arr['wall']          = ((x($arr,'wall'))          ? intval($arr['wall'])                : 0);
        $arr['uri']           = ((x($arr,'uri'))           ? notags(trim($arr['uri']))           : random_string());
@@ -884,7 +924,7 @@ function item_store($arr,$force_parent = false) {
                        if($r[0]['uri'] != $r[0]['parent-uri']) {
                                $arr['thr-parent'] = $arr['parent-uri'];
                                $arr['parent-uri'] = $r[0]['parent-uri'];
-                               $z = q("SELECT `id` FROM `item` WHERE `uri` = '%s' AND `parent-uri` = '%s' AND `uid` = %d 
+                               $z = q("SELECT * FROM `item` WHERE `uri` = '%s' AND `parent-uri` = '%s' AND `uid` = %d 
                                        ORDER BY `id` ASC LIMIT 1",
                                        dbesc($r[0]['parent-uri']),
                                        dbesc($r[0]['parent-uri']),
@@ -1709,7 +1749,7 @@ function consume_feed($xml,$importer,&$contact, &$hub, $datedir = 0, $pass = 0)
 
        // Now process the feed
 
-       if($feed->get_item_quantity()) {                
+       if($feed->get_item_quantity()) {
 
                logger('consume_feed: feed item count = ' . $feed->get_item_quantity());
 
@@ -1722,7 +1762,7 @@ function consume_feed($xml,$importer,&$contact, &$hub, $datedir = 0, $pass = 0)
 
                foreach($items as $item) {
 
-                       $is_reply = false;              
+                       $is_reply = false;
                        $item_id = $item->get_id();
                        $rawthread = $item->get_item_tags( NAMESPACE_THREAD,'in-reply-to');
                        if(isset($rawthread[0]['attribs']['']['ref'])) {
@@ -1736,11 +1776,10 @@ function consume_feed($xml,$importer,&$contact, &$hub, $datedir = 0, $pass = 0)
                                        continue;
 
                                // Have we seen it? If not, import it.
-       
+
                                $item_id  = $item->get_id();
                                $datarray = get_atom_elements($feed,$item);
 
-
                                if((! x($datarray,'author-name')) && ($contact['network'] != NETWORK_DFRN))
                                        $datarray['author-name'] = $contact['name'];
                                if((! x($datarray,'author-link')) && ($contact['network'] != NETWORK_DFRN))
@@ -3657,7 +3696,6 @@ function posted_date_widget($url,$uid,$wall) {
        return $o;
 }
 
-
 function store_diaspora_retract_sig($item, $user, $baseurl) {
        // Note that we can't add a target_author_signature
        // if the comment was deleted by a remote user. That should be ok, because if a remote user is deleting