bans plugin: fix badword regular expression creation

[user/henk/code/ruby/rbot.git] / data / rbot / plugins / search.rb

diff --git a/data/rbot/plugins/search.rb b/data/rbot/plugins/search.rb
index 3e1066f784fcad18fdcb2879ea1b83b0b9099c68..becfafc2368eb9717ee87ba60f5eec6d80d11a25 100644 (file)
--- a/data/rbot/plugins/search.rb
+++ b/data/rbot/plugins/search.rb
@@ -1,50 +1,13 @@
+# vim: set sw=2 et:
+#
+# TODO: use lr=lang_<code> or whatever is most appropriate to let google know
+# it shouldn't use the bot's location to find the preferred language
 require 'uri'
 
 Net::HTTP.version_1_2
 
 GOOGLE_WAP_LINK = /<a accesskey="(\d)" href=".*?u=(.*?)">(.*?)<\/a>/im
 
-class ::String
-  def omissis_after(len)
-    if self.length > len
-      return self[0...len].sub(/\s+\S*$/,"...")
-    else
-      return self
-    end
-  end
-
-  def ircify_html
-    txt = self
-
-    # bold and strong -> bold
-    txt.gsub!(/<\/?(?:b|strong)\s*>/, "#{Bold}")
-
-    # italic, emphasis and underline -> underline
-    txt.gsub!(/<\/?(?:i|em|u)\s*>/, "#{Underline}")
-
-    ## This would be a nice addition, but the results are horrible
-    ## Maybe make it configurable?
-    # txt.gsub!(/<\/?a( [^>]*)?>/, "#{Reverse}")
-
-    # Paragraph and br tags are converted to whitespace.
-    txt.gsub!(/<\/?(p|br)\s*\/?\s*>/, ' ')
-    txt.gsub!("\n", ' ')
-
-    # All other tags are just removed
-    txt.gsub!(/<[^>]+>/, '')
-
-    # Remove double formatting options, since they only waste bytes
-    txt.gsub!(/#{Bold}\s*#{Bold}/,"")
-    txt.gsub!(/#{Underline}\s*#{Underline}/,"")
-
-    # And finally whitespace is squeezed
-    txt.gsub!(/\s+/, ' ')
-
-    # Decode entities and strip whitespace
-    return Utils.decode_html_entities(txt).strip!
-  end
-end
-
 class SearchPlugin < Plugin
   BotConfig.register BotConfigIntegerValue.new('google.hits',
     :default => 3,
@@ -107,55 +70,14 @@ class SearchPlugin < Plugin
       "#{n}. #{Bold}#{t}#{Bold}: #{u}"
     }.join(" | ")
 
-    m.reply "Results for #{what}: #{results}"
+    m.reply "Results for #{what}: #{results}", :split_at => /\s+\|\s+/
 
     first_pars = params[:firstpar] || @bot.config['google.first_par']
 
-    idx = 0
-    while first_pars > 0 and urls.length > 0
-      url.replace(urls.shift)
-      idx += 1
-
-      # FIXME what happens if some big file is returned? We should share
-      # code with the url plugin to only retrieve partial file content!
-      xml = @bot.httputil.get_cached(url)
-      if xml.nil?
-        debug "Unable to retrieve #{url}"
-        next
-      end
-      # We get the first par after the first main heading, if possible
-      header_found = xml.match(/<h1(?:\s+[^>]*)?>(.*?)<\/h1>/im)
-      txt = String.new
-      if header_found
-        debug "Found header: #{header_found[1].inspect}"
-        while txt.empty? 
-          header_found = $'
-          candidate = header_found[/<p(?:\s+[^>]*)?>.*?<\/p>/im]
-          break unless candidate
-          txt.replace candidate.ircify_html
-        end
-      end
-      # If we haven't found a first par yet, try to get it from the whole
-      # document
-      if txt.empty?
-       header_found = xml
-        while txt.empty? 
-          candidate = header_found[/<p(?:\s+[^>]*)?>.*?<\/p>/im]
-          break unless candidate
-          txt.replace candidate.ircify_html
-          header_found = $'
-        end
-      end
-      # Nothing yet, try title
-      if txt.empty?
-        debug "No first par found\n#{xml}"
-       # FIXME only do this if the 'url' plugin is loaded
-       txt.replace @bot.plugins['url'].get_title_from_html(xml)
-        next if txt.empty?
-      end
-      m.reply "[#{idx}] #{txt}".omissis_after(400)
-      first_pars -=1
-    end
+    return unless first_pars > 0
+
+    Utils.get_first_pars urls, first_pars, :http_util => @bot.httputil, :message => m
+
   end
 
   def wikipedia(m, params)