use pretty printing with Logger (debug, info, warning and error messages)

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

diff --git a/data/rbot/plugins/url.rb b/data/rbot/plugins/url.rb
index 0d85d473a1e6be42bc4a7be89b05d5d058761758..a1a325eb3d5f9bdfbc60f0e5ef368bd2687194b4 100644 (file)
--- a/data/rbot/plugins/url.rb
+++ b/data/rbot/plugins/url.rb
@@ -1,7 +1,6 @@
-require 'uri'
-
 Url = Struct.new("Url", :channel, :nick, :time, :url)
 TITLE_RE = /<\s*?title\s*?>(.+?)<\s*?\/title\s*?>/im
+LINK_INFO = "[Link Info]"
 
 class UrlPlugin < Plugin
   BotConfig.register BotConfigIntegerValue.new('url.max_urls',
@@ -13,6 +12,9 @@ class UrlPlugin < Plugin
   BotConfig.register BotConfigBooleanValue.new('url.titles_only',
     :default => false,
     :desc => "Only show info for links that have <title> tags (in other words, don't display info for jpegs, mpegs, etc.)")
+  BotConfig.register BotConfigBooleanValue.new('url.first_par',
+    :default => false,
+    :desc => "Also try to get the first paragraph of a web page")
 
   def initialize
     super
@@ -25,9 +27,7 @@ class UrlPlugin < Plugin
 
   def get_title_from_html(pagedata)
     return unless TITLE_RE.match(pagedata)
-    title = $1.strip.gsub(/\s*\n+\s*/, " ")
-    title = Utils.decode_html_entities title
-    "title: #{title}"
+    $1.ircify_html
   end
 
   def get_title_for_url(uri_str)
@@ -38,43 +38,54 @@ class UrlPlugin < Plugin
     title = nil
 
     begin
-      @bot.httputil.get_response(url) { |response|
-        case response
-        when Net::HTTPSuccess
-          if response['content-type'] =~ /^text\//
-            # since the content is 'text/*' and is small enough to
-            # be a webpage, retrieve the title from the page
-            debug "+ getting #{url.request_uri}"
-
-            # we look for the title in the first 4k bytes
-            # TODO make the amount of data configurable
-            response.partial_body(4096) { |part|
-              title = get_title_from_html(part)
-              return title if title
-            }
-            # if nothing was found, return nothing
-            return
-          else
-            unless @bot.config['url.titles_only']
-              # content doesn't have title, just display info.
-              size = response['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2')
-              size = size ? ", size: #{size} bytes" : ""
-              return "type: #{response['content-type']}#{size}"
-            end
-          end
-        when Net::HTTPResponse
-          return "Error getting link (#{response.code} - #{response.message})"
+      range = @bot.config['http.info_bytes']
+      response = @bot.httputil.get_response(url, :range => "bytes=0-#{range}")
+      if response.code != "206" && response.code != "200"
+        return "Error getting link (#{response.code} - #{response.message})"
+      end
+      extra = String.new
+
+      if response['content-type'] =~ /^text\//
+
+        body = response.body.slice(0, range)
+        title = String.new
+
+        # since the content is 'text/*' and is small enough to
+        # be a webpage, retrieve the title from the page
+        debug "+ getting #{url.request_uri}"
+
+       title = get_title_from_html(body)
+        if @bot.config['url.first_par']
+          first_par = Utils.ircify_first_html_par(body, :strip => title)
+          extra << ", #{Bold}text#{Bold}: #{first_par}" unless first_par.empty?
+          return "#{Bold}title#{Bold}: #{title}#{extra}" if title
         else
-          raise response
+          return "#{Bold}title#{Bold}: #{title}" if title
         end
-      }
-    rescue Object => e
-      if e.class <= StandardError
-        error e.inspect
-        debug e.backtrace.join("\n")
+
+        # if nothing was found, provide more basic info
       end
 
-      msg = e.respond_to?(:message) ? e.message : e.to_s
+      debug response.to_hash.inspect
+
+      enc = response['content-encoding']
+
+      extra << ", #{Bold}encoding#{Bold}: #{enc}" if enc
+
+      unless @bot.config['url.titles_only']
+        # content doesn't have title, just display info.
+        size = response['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2') rescue nil
+        if response.code == '206'
+          if response['content-range'] =~ /bytes\s*[^\/]+\/(\d+)/
+            size = $1.to_s.reverse.scan(/\d{1,3}/).join(',').reverse
+          end
+        end
+        size = size ? ", #{Bold}size#{Bold}: #{size} bytes" : ""
+        return "#{Bold}type#{Bold}: #{response['content-type']}#{size}#{extra}"
+      end
+    rescue Exception => e
+      error e.inspect
+      debug e.backtrace.join("\n")
       return "Error connecting to site (#{e.message})"
     end
   end
@@ -94,7 +105,7 @@ class UrlPlugin < Plugin
             begin
               title = get_title_for_url urlstr
               if title
-                m.reply "[Link Info] #{title}"
+                m.reply "#{LINK_INFO} #{title}", :overlong => :truncate
                 debug "Title found!"
               else
                 debug "Title not found!"