imdb plugin: show popular movies acted/directed when finding people

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

diff --git a/data/rbot/plugins/imdb.rb b/data/rbot/plugins/imdb.rb
index cfadd3c8568ad4bad79feeed639e39b3606825fd..5615ac00c4dd7e4cb07fc1a8b807738a8501d77d 100644 (file)
--- a/data/rbot/plugins/imdb.rb
+++ b/data/rbot/plugins/imdb.rb
@@ -9,9 +9,7 @@
 #
 # Notes by Giuseppe Bilotta:
 # TODO return more than one match (configurable)
-# TODO why do we use CGI.unescapeHTML? shall we rely on the rbot methods?
 
-require 'cgi'
 require 'uri/common'
 
 class Imdb
@@ -20,7 +18,11 @@ class Imdb
   end
 
   def search(rawstr)
-    str = URI.escape(rawstr)
+    str = URI.escape(rawstr) << ";site=aka"
+    return do_search(str)
+  end
+
+  def do_search(str)
     resp = nil
     begin
       resp = @bot.httputil.get_response("http://us.imdb.com/find?q=#{str}",
@@ -32,14 +34,18 @@ class Imdb
     end
 
     if resp.code == "200"
-      m = /<a href="(\/title\/tt[0-9]+\/?)[^"]*"(?:[^>]*)>([^<]*)<\/a>/.match(resp.body)
+      m = /<a href="(\/(?:title|name)\/(?:tt|nm)[0-9]+\/?)[^"]*"(?:[^>]*)>(?:[^<]*)<\/a>/.match(resp.body)
       if m
         url = m[1]
-        title = m[2]
         return url
       end
     elsif resp.code == "302"
-      return resp['location'].gsub(/http:\/\/us.imdb.com/, "").gsub(/\?.*/, "")
+      new_loc = resp['location'].gsub(/http:\/\/us.imdb.com/, "")
+      if new_loc.match(/\/find\?q=(.*)/)
+        return do_search($1)
+      else
+        return new_loc.gsub(/\?.*/, "")
+      end
     end
     return nil
   end
@@ -50,6 +56,22 @@ class Imdb
       debug "IMDB: search returned NIL"
       return nil
     end
+    type = sr.match(/^\/([^\/]+)\//)[1].downcase.intern rescue nil
+    case type
+    when :title
+      return info_title(sr)
+    when :name
+      return info_name(sr)
+    else
+      return "#{sr}"
+    end
+  end
+
+  def grab_info(info, body)
+    /<div class="info">\s+<h5>#{info}:<\/h5>\s+(.*?)<\/div>/mi.match(body)[1] rescue nil
+  end
+
+  def info_title(sr)
     resp = nil
     begin
       resp = @bot.httputil.get_response('http://us.imdb.com' + sr,
@@ -63,19 +85,92 @@ class Imdb
     if resp.code == "200"
       m = /<title>([^<]*)<\/title>/.match(resp.body)
       return nil if !m
-      title = CGI.unescapeHTML(m[1])
+      title = Utils.decode_html_entities(m[1])
 
       m = /<b>([0-9.]+)\/10<\/b>\n?\r?\s+<small>\(<a href="ratings">([0-9,]+) votes?<\/a>\)<\/small>/.match(resp.body)
       return nil if !m
       score = m[1]
       votes = m[2]
 
+      plot = nil
+      data = grab_info(/Plot (?:Outline|Summary)/, resp.body)
+      if data
+        plot = "Plot: #{data.ircify_html.gsub(/\s+more$/,'')}"
+      end
+
       genre = Array.new
       resp.body.scan(/<a href="\/Sections\/Genres\/[^\/]+\/">([^<]+)<\/a>/) do |gnr|
         genre << gnr
       end
-      return ["http://us.imdb.com" + sr, title, score, votes,
-        genre]
+      info = "#{title} : http://us.imdb.com#{sr}\n"
+      info << "Ratings: #{score}/10 (#{votes} voters). Genre: #{genre.join('/')}\n"
+      info << plot if plot
+      return info
+    end
+    return nil
+  end
+
+  def info_name(sr)
+    resp = nil
+    begin
+      resp = @bot.httputil.get_response('http://us.imdb.com' + sr,
+                                        :max_redir => -1)
+    rescue Exception => e
+      error e.message
+      warning e.backtrace.join("\n")
+      return nil
+    end
+
+    if resp.code == "200"
+      m = /<title>([^<]*)<\/title>/.match(resp.body)
+      return nil if !m
+      name = Utils.decode_html_entities(m[1])
+
+      birth = nil
+      data = grab_info("Date of Birth", resp.body)
+      if data
+        birth = "Birth: #{data.ircify_html.gsub(/\s+more$/,'')}"
+      end
+
+      death = nil
+      data = grab_info("Date of Death", resp.body)
+      if data
+        death = "Death: #{data.ircify_html.gsub(/\s+more$/,'')}"
+      end
+
+      movies = {}
+
+      filmorate = nil
+      begin
+        filmorate = @bot.httputil.get("http://us.imdb.com" + sr + "filmorate")
+      rescue Exception
+      end
+
+      if filmorate
+        filmorate.scan(/<div class="filmo">.*?<a href="\/title.*?<\/div>/m) { |str|
+          what = str.match(/<a name="[^"]+">([^<]+)<\/a>/)[1] rescue nil
+          # next unless what
+          next unless ['Actor', 'Director'].include?(what)
+          movies[what] = str.scan(/<a href="\/title\/[^"]+">([^<]+)<\/a>/)[0..2].map { |tit|
+            Utils.decode_html_entities(tit)
+          }
+        }
+      end
+      debug movies.inspect
+
+      info = "#{name} : http://us.imdb.com#{sr}\n"
+      info << [birth, death].compact.join('. ') << "\n"
+      unless movies.empty?
+        info << "Top Movies:: "
+        ar = []
+        movies.keys.sort.each { |key|
+          ar << key.dup
+          ar.last << ": " + movies[key].join(', ')
+        }
+        info << ar.join('. ')
+      end
+      return info
+
     end
     return nil
   end
@@ -94,8 +189,7 @@ class ImdbPlugin < Plugin
       m.reply "Nothing found for #{what}"
       return nil
     end
-    m.reply "#{info[1]} : #{info[0]}"
-    m.reply "Ratings: #{info[2]}/10 (#{info[3]} voters). Genre: #{info[4].join('/')}"
+    m.reply info
   end
 end