httputil: try to guess content-type from extension if it's not defined

[user/henk/code/ruby/rbot.git] / lib / rbot / core / utils / extends.rb

diff --git a/lib/rbot/core/utils/extends.rb b/lib/rbot/core/utils/extends.rb
index 0b07257a1b58a42b2fe4d98b6da40ba43beb63f7..7b733994b74d323b010db0bc2ee6d66dad1f598d 100644 (file)
--- a/lib/rbot/core/utils/extends.rb
+++ b/lib/rbot/core/utils/extends.rb
@@ -137,6 +137,10 @@ class ::String
     txt.gsub!(/<sub>(.*?)<\/sub>/, '_{\1}')
     txt.gsub!(/(^|_)\{(.)\}/, '\1\2')
 
+    # List items are converted to *). We don't have special support for
+    # nested or ordered lists.
+    txt.gsub!(/<li>/, ' *) ')
+
     # All other tags are just removed
     txt.gsub!(/<[^>]+>/, '')
 
@@ -144,6 +148,14 @@ class ::String
     # such as &nbsp;
     txt = Utils.decode_html_entities(txt)
 
+    # Keep unbreakable spaces or conver them to plain spaces?
+    case val = opts[:nbsp]
+    when :space, ' '
+      txt.gsub!([160].pack('U'), ' ')
+    else
+      warning "unknown :nbsp option #{val} passed to ircify_html" if val
+    end
+
     # Remove double formatting options, since they only waste bytes
     txt.gsub!(/#{Bold}(\s*)#{Bold}/, '\1')
     txt.gsub!(/#{Underline}(\s*)#{Underline}/, '\1')
@@ -182,15 +194,18 @@ class ::String
   # This method tries to find an HTML title in the string,
   # and returns it if found
   def get_html_title
-    return unless Irc::Utils::TITLE_REGEX.match(self)
-    $1
+    if defined? ::Hpricot
+      Hpricot(self).at("title").inner_html
+    else
+      return unless Irc::Utils::TITLE_REGEX.match(self)
+      $1
+    end
   end
 
   # This method returns the IRC-formatted version of an
   # HTML title found in the string
   def ircify_html_title
-    return unless Irc::Utils::TITLE_REGEX.match(self)
-    $1.ircify_html
+    self.get_html_title.ircify_html rescue nil
   end
 end