From b440a5f61db2c0dbf5266ce6885d2240b6fb02e9 Mon Sep 17 00:00:00 2001
From: Michael Orlitzky <michael@orlitzky.com>
Date: Sun, 17 Oct 2010 23:19:47 -0400
Subject: [PATCH] Add blip.tv support for URLs of the form
 http://blip.tv/play/<foo>.

---
 src/websites/bliptv.rb | 75 +++++++++++++++++++++++++++++++++++-------
 1 file changed, 63 insertions(+), 12 deletions(-)

diff --git a/src/websites/bliptv.rb b/src/websites/bliptv.rb
index b3bd8bb..116d4f7 100644
--- a/src/websites/bliptv.rb
+++ b/src/websites/bliptv.rb
@@ -21,38 +21,89 @@ require 'src/website'
 class Bliptv < Website
 
   VALID_BLIPTV_URL_REGEX = /^(http:\/\/)?([[:alnum:]\-]+\.)?blip\.tv\/file\/(\d+)(.*)?$/
+  VALID_BLIPTV_REDIR_URL_REGEX = /^(http:\/\/)?([[:alnum:]\-]+\.)?blip\.tv\/play\/[[:alnum:]]+$/
 
   def self.owns_url?(url)
-    return url =~ VALID_BLIPTV_URL_REGEX
+    return (url =~ VALID_BLIPTV_URL_REGEX ||
+            url =~ VALID_BLIPTV_REDIR_URL_REGEX)
   end
 
- 
+
   def get_video_url()
-    page_data = self.get_page_data(@url)    
+    page_data = ''
+
+    if (@url =~ VALID_BLIPTV_URL_REGEX)
+      page_data = self.get_page_data(@url)
+    else
+      # It's a redirect. Figure out the RSS page URL from the redirect.
+      redir_url = get_redirect_url
+      rss_page_url = parse_rss_url(redir_url)
+      rss_data = get_page_data(rss_page_url)
+
+      # The "real" page URL is embedded in the RSS feed. Once we get the
+      # real URL, we can proceed as if we were given that URL in the first
+      # place.
+      real_page_url = parse_page_url(rss_data)
+      page_data = self.get_page_data(real_page_url)
+    end
+
     filepath = parse_video_url(page_data)
-    
+
     return filepath
   end
 
-  
+
   protected;
 
-  
+  def parse_page_url(data)
+    # A simplified VALID_BLIPTV_URL_REGEX.
+    page_url_regex = /http:\/\/blip\.tv\/file\/\d+/
+    matches = page_url_regex.match(data)
+
+    if matches.nil?
+      raise StandardError.new("Couldn't parse the real page URL from the RSS page.")
+    end
+
+    return matches[0]
+  end
+
+  def parse_rss_url(url)
+    rss_id_regex = /\/flash\/(\d+)/
+    matches = rss_id_regex.match(url)
+
+    if matches.nil? or (matches.length < 2)
+      raise StandardError.new("Couldn't parse the video ID from the redirect URL: #{url}")
+    end
+
+    return "http://blip.tv/rss/flash/#{matches[1]}"
+  end
+
+  def get_redirect_url
+    uri = URI.parse(@url)
+
+    response = Net::HTTP.start(uri.host, uri.port) do |http|
+      http.get(uri.request_uri, {})
+    end
+
+    return response['location']
+  end
+
+
   def parse_video_url(page_data)
     # First, try to find the MOV video. The source videos are usually
     # encoded with MOV.
     video_url_regex = /"Quicktime \(\.mov\)", "attribute" : "(.*?\.mov)/i
     matches = video_url_regex.match(page_data)
 
-    if not matches.nil?
+    if not matches.nil? and (matches.length > 1)
       return matches[1]
     end
 
-    # I've seen some free software videos encoded as OGG/Vorbis, too.    
+    # I've seen some free software videos encoded as OGG/Vorbis, too.
     video_url_regex = /"Ogg Theora\/Vorbis \(\.og[gv]\)", "attribute" : "(.*?\.og[gv])/i
     matches = video_url_regex.match(page_data)
 
-    if not matches.nil?
+    if not matches.nil? and (matches.length > 1)
       return matches[1]
     end
 
@@ -61,17 +112,17 @@ class Bliptv < Website
     video_url_regex = /"Windows Media \(\.wmv\)", "attribute" : "(.*?\.wmv)/i
     matches = video_url_regex.match(page_data)
 
-    if not matches.nil?
+    if not matches.nil? and (matches.length > 1)
       return matches[1]
     end
 
-    
+
     # If neither of the source formats are present, just grab the
     # video URL from the Flash variable and be done with it.
     video_url_regex = /setPrimaryMediaUrl\("(.*?\.(flv|mov|wmv|mp4|og[gv]))/i
     matches = video_url_regex.match(page_data)
 
-    if matches.nil?
+    if matches.nil? or (matches.length < 2)
       raise StandardError.new("Couldn't parse any of the video format URLs.")
     end
 
-- 
2.44.2