X-Git-Url: https://git.netwichtig.de/gitweb/?a=blobdiff_plain;f=data%2Frbot%2Fplugins%2Fslashdot.rb;h=2c4a23618b8c91d8f551f53e2dd13b855c9befb5;hb=bc7efe2d4b360da0276287e6cc7f6a401609c162;hp=c9e35b9e9592badde10fe4689aa9765e17c55add;hpb=109fa2a5b63af113df2c6b21d44135efa0d94d70;p=user%2Fhenk%2Fcode%2Fruby%2Frbot.git

diff --git a/data/rbot/plugins/slashdot.rb b/data/rbot/plugins/slashdot.rb
index c9e35b9e..2c4a2361 100644
--- a/data/rbot/plugins/slashdot.rb
+++ b/data/rbot/plugins/slashdot.rb
@@ -1,3 +1,8 @@
+#-- vim:sw=2:et
+#++
+#
+# :title: Slashdot plugin for rbot
+
 require 'rexml/document'
 
 class SlashdotPlugin < Plugin
@@ -5,7 +10,59 @@ class SlashdotPlugin < Plugin
   def help(plugin, topic="")
     "slashdot search <string> [<max>=4] => search slashdot for <string>, slashdot [<max>=4] => return up to <max> slashdot headlines (use negative max to return that many headlines, but all on one line.)"
   end
-  
+
+  # This method defines a filter for /. pages. It's needed because the generic
+  # summarization grabs a comment, not the actual article.
+  #
+  # This filter relies on Hpricot being available, since REXML isn't too
+  # happy with the /. pages
+  def slashdot_filter(s)
+    return nil unless defined? Hpricot
+    loc = Utils.check_location(s, /slashdot\.org/)
+    return nil unless loc
+    h = Hpricot(s[:text])
+    # If we have no title tag in a head tag, return as this is not
+    # a /. page (it's probably a Slashdot RSS
+    ht = h/"head/title"
+    return nil if ht.empty?
+    title = ht.first.to_html.ircify_html
+    arts = (h/"div.article")
+    return nil if arts.empty?
+    if arts.length > 1
+      tits = []
+      arts.each { |el|
+        # see if the div tag with generaltitle class is present
+        artitle = (el/"div.generaltitle").first
+        if artitle
+          tits << artitle.to_html.ircify_html
+          next
+        end
+        # otherwise, check for skin+datitle a tags
+        datitle = (el/"a.datitle").first
+        next unless datitle
+        skin = (el/"a.skin").first
+        artitle = [
+          skin ? skin.innerHTML.ircify_html : nil,
+          datitle.innerHTML.ircify_html
+        ].compact.join(" ")
+        tits << artitle
+      }
+      content = tits.join(" | ")
+    else
+      det = (arts.first/"div.details").first.to_html.ircify_html
+      body = (arts.first/"div.body").first.to_html.ircify_html
+      content = [det, body].join(' ')
+    end
+    return {:title => title, :content => content}
+  end
+
+  def initialize
+    super
+    if defined? Hpricot
+      @bot.register_filter(:slashdot, :htmlinfo) { |s| slashdot_filter(s) }
+    end
+  end
+
   def search_slashdot(m, params)
     max = params[:limit].to_i
     search = params[:search].to_s
@@ -18,9 +75,9 @@ class SlashdotPlugin < Plugin
     debug xml.inspect
     begin
       doc = Document.new xml
-    rescue REXML::ParseException => e
-      warning e.inspect
-      m.reply "couldn't parse output XML: #{e.class}"
+    rescue REXML::ParseException => err
+      warning err.inspect
+      m.reply "couldn't parse output XML: #{err.class}"
       return
     end
     unless doc
@@ -42,7 +99,7 @@ class SlashdotPlugin < Plugin
       m.reply "search for #{search} failed"
     end
   end
-  
+
   def slashdot(m, params)
     debug params.inspect
     max = params[:limit].to_i
@@ -66,12 +123,12 @@ class SlashdotPlugin < Plugin
     max = 8 if max > 8
     matches = Array.new
     doc.elements.each("*/story") {|e|
-      matches << [ e.elements["title"].text, 
-                   e.elements["author"].text, 
+      matches << [ e.elements["title"].text,
+                   e.elements["author"].text,
                    e.elements["time"].text.gsub(/\d{4}-(\d{2})-(\d{2})/, "\\2/\\1").gsub(/:\d\d$/, "") ]
       done += 1
       break if done >= max
-    } 
+    }
     if oneline
       m.reply matches.collect{|mat| mat[0]}.join(" | ")
     else