[linterna-magica-commit] [129] Changes for tasks #11216.

[Ivaylo Valkov]

Revision: 129
          
http://svn.sv.gnu.org/viewvc/?view=rev&root=linterna-magica&revision=129
Author:   valkov
Date:     2011-07-13 09:26:07 +0000 (Wed, 13 Jul 2011)
Log Message:
-----------
Changes for tasks #11216. Added rule to extract object data from scripts. Sites 
that need the change use it.

Ticket Links:
------------
    http://savannah.gnu.org/task/?11216

Modified Paths:
--------------
    branches/task-11216/src/lm_extract_js_scripts.js
    branches/task-11216/src/lm_site_theonion.js
    branches/task-11216/src/lm_site_vimeo.js
    branches/task-11216/src/lm_site_youtube.js
    branches/task-11216/src/lm_sites.js

Modified: branches/task-11216/src/lm_extract_js_scripts.js
===================================================================
--- branches/task-11216/src/lm_extract_js_scripts.js    2011-07-12 19:22:47 UTC 
(rev 128)
+++ branches/task-11216/src/lm_extract_js_scripts.js    2011-07-13 09:26:07 UTC 
(rev 129)
@@ -73,44 +73,27 @@
            continue;
        }
 
-       if (/theonion\.com/i.test(window.location.hostname))
-       {
-           object_data = this.extract_object_from_script_theonion();
+       var self = this;
+       var val = this.call_site_function_at_position.apply(self,[
+           "extract_object_from_script",
+           window.location.hostname]);
 
-           if (!object_data)
-           {
-               // No other method of extraction is useful. Skip to
-               // next script.
-               continue;
-           }
-       }
 
-       if (/youtube\.com/i.test(window.location.hostname) ||
-           (/youtube-nocookie\.com/i.test(window.location.hostname)))
+       if (this.sites[window.location.hostname] && !val)
        {
-           object_data =
-               this.extract_object_from_script_youtube();
+           // Site specific code is used but no results were
+           // returned. Can't extract object information. General
+           // purpose extraction might not worк, so it is useless.
+           this.log("LinternaMagica.extract_objects_from_scripts:\n"+
+                    "Site specific code did not return object data. Skipping"+
+                    " general purpose extraction",6);
 
-           // Fix bloating in FF (mainly). Optimizes script
-           // processing. No need to check the other swf constructors
-           // in YouTube.
-           if (!object_data)
-           {
-               continue;
-           }
+           continue;
        }
 
-       if (/vimeo\.com/i.test(window.location.hostname))
+       if (val && typeof(val) != "boolean")
        {
-           object_data =
-               this.extract_object_from_script_vimeo();
-
-           // Optimizes script processing. No need to check the other
-           // swf constructors in Vimeo.
-           if (!object_data)
-           {
-               continue;
-           }
+           object_data = val;
        }
 
        if (!object_data)

Modified: branches/task-11216/src/lm_site_theonion.js
===================================================================
--- branches/task-11216/src/lm_site_theonion.js 2011-07-12 19:22:47 UTC (rev 
128)
+++ branches/task-11216/src/lm_site_theonion.js 2011-07-13 09:26:07 UTC (rev 
129)
@@ -28,8 +28,14 @@
 
 // Support for The Onion dot com
 
+LinternaMagica.prototype.sites["theonion.com"] = new Object();
+
+// Reference
+LinternaMagica.prototype.sites["www.theonion.com"] = "theonion.com";
+
 // Extracts data for the flash object in The Onion dot com from a script
-LinternaMagica.prototype.extract_object_from_script_theonion = function()
+LinternaMagica.prototype.sites["theonion.com"].extract_object_from_script =
+function()
 {
     var player_container = document.getElementById("player_container");
 
@@ -141,11 +147,6 @@
     }
 }
 
-LinternaMagica.prototype.sites["theonion.com"] = new Object();
-
-// Reference
-LinternaMagica.prototype.sites["www.theonion.com"] = "theonion.com";
-
 LinternaMagica.prototype.sites["theonion.com"].flash_plugin_installed =
 function()
 {

Modified: branches/task-11216/src/lm_site_vimeo.js
===================================================================
--- branches/task-11216/src/lm_site_vimeo.js    2011-07-12 19:22:47 UTC (rev 
128)
+++ branches/task-11216/src/lm_site_vimeo.js    2011-07-13 09:26:07 UTC (rev 
129)
@@ -27,6 +27,11 @@
 
 // END OF LICENSE HEADER
 
+LinternaMagica.prototype.sites["vimeo.com"] = new Object();
+
+// Reference 
+LinternaMagica.prototype.sites["www.vimeo.com"] = "vimeo.com";
+
 // Detect vimeo browser upgrade warning (no flash & h264) .This is
 // called withing setInterval. It is needed because when the elements
 // with the warning are inserted all our data that has been added
@@ -83,7 +88,7 @@
 
 // Extract object data in Vimeo. This makes Firefox and forks to work
 // without plugin and without HTML5 (missing H264)
-LinternaMagica.prototype.extract_object_from_script_vimeo = function()
+LinternaMagica.prototype.sites["vimeo.com"].extract_object_from_script = 
function()
 {
     var player_element_re = new RegExp(
        "player[0-9]+_[0-9]+_element\\\s*=\\\s*"+

Modified: branches/task-11216/src/lm_site_youtube.js
===================================================================
--- branches/task-11216/src/lm_site_youtube.js  2011-07-12 19:22:47 UTC (rev 
128)
+++ branches/task-11216/src/lm_site_youtube.js  2011-07-13 09:26:07 UTC (rev 
129)
@@ -167,9 +167,80 @@
     }
 }
 
+// Extract links data for youtube from fmt_url_map
+LinternaMagica.prototype.extract_youtube_fmt_url_map = function(data)
+{
+    var fmt_re = new RegExp (
+       "(\\\"|\\\'|\\\&)fmt_url_map"+
+           "(\\\"|\\\')*(\\\=|\\\:|,)\\\s*(\\\"|\\\')*"+
+           "([a-zA-Z0-9\\\-\\\_\\\%\\\=\\\/,\\\\\.\|:=&%\?]+)");
 
+    var fmt = data.match(fmt_re);
+
+    if (fmt)
+    {
+
+       // For debug level 1
+       this.log("LinternaMagica.extract_youtube_fmt_url_map:\n"+
+                "Extracted fmt_url_map.",1);
+
+       // Hash with keys fmt_ids and values video URLs
+       var map = new Object();
+
+       // There was unescape here but it broke the split by ,. How it
+       // worked it is not clear. Maybe YT changed something.
+       fmt = fmt[fmt.length-1].replace(/\\\//g, "/");
+
+       fmt = fmt.split(/,/);
+
+       for (var url=0; url<fmt.length; url++) 
+       {
+           // fmt_id|link
+           var m = fmt[url].split(/\|/);
+           map[m[0]] =  m[1];
+       }
+
+       return map;
+    }
+    else
+    {
+       this.log("LinternaMagica.extract_youtube_fmt_url_map:\n"+
+                "No fmt_url_map parameter found. ",1);
+    }
+
+    return null;
+}
+
+LinternaMagica.prototype.sites["youtube.com"] = new Object();
+
+// Reference
+LinternaMagica.prototype.sites["www.youtube.com"] = "youtube.com";
+LinternaMagica.prototype.sites["www.youtube-nocookie.com"] = "youtube.com";
+LinternaMagica.prototype.sites["youtube-nocookie.com"] = "youtube.com";
+
+LinternaMagica.prototype.sites["youtube.com"].set_cookies_domain =
+function()
+{
+    return ".youtube.com";
+}
+
+LinternaMagica.prototype.sites["youtube.com"].skip_link_extraction = function()
+{
+    // Link extraction bloats FF in youtube:
+    // LinternaMagica.extract_link_from_param_list: 
+    // Trying to extract a link from param/attribute "flashvars"
+    // at www.youtube.com time: ***14:58:59:999***
+    // LinternaMagica.extract_link: No link found. at
+    // www.youtube.com time: ***15:12:21:356***
+    this.log("LinternaMagica.sites.skip_link_extraction:\n"+
+            "Skipping link extraction in YouTube. Might bloat "+
+            "GNU IceCat and other forks and versions of Firefox.",4);
+    return false;
+}
+
 // Extracts data for the flash object in youtube from a script
-LinternaMagica.prototype.extract_object_from_script_youtube = function()
+LinternaMagica.prototype.sites["youtube.com"].extract_object_from_script =
+function()
 {
     var data = this.script_data;
     if (!data.match(/var\s*swfConfig/))
@@ -258,74 +329,3 @@
 
     return object_data;
 }
-
-// Extract links data for youtube from fmt_url_map
-LinternaMagica.prototype.extract_youtube_fmt_url_map = function(data)
-{
-    var fmt_re = new RegExp (
-       "(\\\"|\\\'|\\\&)fmt_url_map"+
-           "(\\\"|\\\')*(\\\=|\\\:|,)\\\s*(\\\"|\\\')*"+
-           "([a-zA-Z0-9\\\-\\\_\\\%\\\=\\\/,\\\\\.\|:=&%\?]+)");
-
-    var fmt = data.match(fmt_re);
-
-    if (fmt)
-    {
-
-       // For debug level 1
-       this.log("LinternaMagica.extract_youtube_fmt_url_map:\n"+
-                "Extracted fmt_url_map.",1);
-
-       // Hash with keys fmt_ids and values video URLs
-       var map = new Object();
-
-       // There was unescape here but it broke the split by ,. How it
-       // worked it is not clear. Maybe YT changed something.
-       fmt = fmt[fmt.length-1].replace(/\\\//g, "/");
-
-       fmt = fmt.split(/,/);
-
-       for (var url=0; url<fmt.length; url++) 
-       {
-           // fmt_id|link
-           var m = fmt[url].split(/\|/);
-           map[m[0]] =  m[1];
-       }
-
-       return map;
-    }
-    else
-    {
-       this.log("LinternaMagica.extract_youtube_fmt_url_map:\n"+
-                "No fmt_url_map parameter found. ",1);
-    }
-
-    return null;
-}
-
-LinternaMagica.prototype.sites["youtube.com"] = new Object();
-
-// Reference
-LinternaMagica.prototype.sites["www.youtube.com"] = "youtube.com";
-LinternaMagica.prototype.sites["www.youtube-nocookie.com"] = "youtube.com";
-LinternaMagica.prototype.sites["youtube-nocookie.com"] = "youtube.com";
-
-LinternaMagica.prototype.sites["youtube.com"].set_cookies_domain =
-function()
-{
-    return ".youtube.com";
-}
-
-LinternaMagica.prototype.sites["youtube.com"].skip_link_extraction = function()
-{
-    // Link extraction bloats FF in youtube:
-    // LinternaMagica.extract_link_from_param_list: 
-    // Trying to extract a link from param/attribute "flashvars"
-    // at www.youtube.com time: ***14:58:59:999***
-    // LinternaMagica.extract_link: No link found. at
-    // www.youtube.com time: ***15:12:21:356***
-    this.log("LinternaMagica.sites.skip_link_extraction:\n"+
-            "Skipping link extraction in YouTube. Might bloat "+
-            "GNU IceCat and other forks and versions of Firefox.",4);
-    return false;
-}

Modified: branches/task-11216/src/lm_sites.js
===================================================================
--- branches/task-11216/src/lm_sites.js 2011-07-12 19:22:47 UTC (rev 128)
+++ branches/task-11216/src/lm_sites.js 2011-07-13 09:26:07 UTC (rev 129)
@@ -143,6 +143,12 @@
     return true;
 }
 
+// Extract object data from script
+LinternaMagica.prototype.sites.__extract_object_from_script = function()
+{
+    return true;
+}
+
 // LinternaMagica.prototype.sites.__extract_scripts_extract_when // Condition 
? DM /ted? 
 // LinternaMagica.prototype.sites.__extract_scripts_once // YT ?
 // LinternaMagica.prototype.sites.__extract_scripts_wait_insert // FB