X-Git-Url: https://git.netwichtig.de/gitweb/?a=blobdiff_plain;f=data%2Frbot%2Fplugins%2Furl.rb;h=ad8951212362b6c26bcad57403c2b73412cfaf38;hb=7f244f32d974cafa74f449853f940f194d08d6d2;hp=524b0858aaab183e9d0b440e87fea03c03f1ac6f;hpb=250407932afa3932c1357de1acf67d54cd12481b;p=user%2Fhenk%2Fcode%2Fruby%2Frbot.git

diff --git a/data/rbot/plugins/url.rb b/data/rbot/plugins/url.rb
index 524b0858..ad895121 100644
--- a/data/rbot/plugins/url.rb
+++ b/data/rbot/plugins/url.rb
@@ -28,7 +28,12 @@ class UrlPlugin < Plugin
     :default => ['localhost', '^192\.168\.', '^10\.', '^127\.', '^172\.(1[6-9]|2\d|31)\.'],
     :on_change => Proc.new { |bot, v| bot.plugins['url'].reset_no_info_hosts },
     :desc => "A list of regular expressions matching hosts for which no info should be provided")
-
+  Config.register Config::ArrayValue.new('url.only_on_channels',
+    :desc => "Show link info only on these channels",
+    :default => [])
+  Config.register Config::ArrayValue.new('url.ignore',
+    :desc => "Don't show link info for urls from users represented as hostmasks on this list. Useful for ignoring other bots, for example.",
+    :default => [])
 
   def initialize
     super
@@ -37,6 +42,8 @@ class UrlPlugin < Plugin
       @bot.config.items[:'url.display_link_info'].set_string(@bot.config['url.display_link_info'].to_s)
     end
     reset_no_info_hosts
+    self.filter_group = :htmlinfo
+    load_filters
   end
 
   def reset_no_info_hosts
@@ -57,8 +64,19 @@ class UrlPlugin < Plugin
     url = uri_str.kind_of?(URI) ? uri_str : URI.parse(uri_str)
     return if url.scheme !~ /https?/
 
-    if url.host =~ @no_info_hosts
-      return "Sorry, info retrieval for #{url.host} is disabled"
+    # also check the ip, the canonical name and the aliases
+    begin
+      checks = TCPSocket.gethostbyname(url.host)
+      checks.delete_at(-2)
+    rescue => e
+      return "Unable to retrieve info for #{url.host}: #{e.message}"
+    end
+
+    checks << url.host
+    checks.flatten!
+
+    unless checks.grep(@no_info_hosts).empty?
+      return ( opts[:always_reply] ? "Sorry, info retrieval for #{url.host} (#{checks.first}) is disabled" : false )
     end
 
     logopts = opts.dup
@@ -68,7 +86,7 @@ class UrlPlugin < Plugin
 
     begin
       debug "+ getting info for #{url.request_uri}"
-      info = Utils.get_html_info(url)
+      info = @bot.filter(:htmlinfo, url)
       debug info
       resp = info[:headers]
 
@@ -109,7 +127,23 @@ class UrlPlugin < Plugin
     return extra.join(", ") if title or not @bot.config['url.titles_only']
   end
 
-  def handle_urls(m, urls, display_info=@bot.config['url.display_link_info'])
+  def handle_urls(m, params={})
+    opts = {
+      :display_info => @bot.config['url.display_link_info'],
+      :channels => @bot.config['url.only_on_channels'],
+      :ignore => @bot.config['url.ignore']
+    }.merge params
+    urls = opts[:urls]
+    display_info= opts[:display_info]
+    channels = opts[:channels]
+    ignore = opts[:ignore]
+
+    unless channels.empty?
+      return unless channels.map { |c| c.downcase }.include?(m.channel.downcase)
+    end
+
+    ignore.each { |u| return if m.source.matches?(u) }
+
     return if urls.empty?
     debug "found urls #{urls.inspect}"
     list = m.public? ? @registry[m.target] : nil
@@ -117,22 +151,35 @@ class UrlPlugin < Plugin
     urls_displayed = 0
     urls.each do |urlstr|
       debug "working on #{urlstr}"
-      next unless urlstr =~ /^https?:/
+      next unless urlstr =~ /^https?:\/\/./
       title = nil
       debug "Getting title for #{urlstr}..."
+      reply = nil
       begin
         title = get_title_for_url(urlstr,
+                                  :always_reply => m.address?,
                                   :nick => m.source.nick,
                                   :channel => m.channel,
                                   :ircline => m.message)
         debug "Title #{title ? '' : 'not '} found"
+        reply = "#{LINK_INFO} #{title}" if title
       rescue => e
-        m.reply "Error #{e.message}"
+        debug e
+        # we might get a 404 because of trailing punctuation, so we try again
+        # with the last character stripped. this might generate invalid URIs
+        # (e.g. because "some.url" gets chopped to some.url%2, so catch that too
+        if e.message =~ /\(404 - Not Found\)/i or e.kind_of?(URI::InvalidURIError)
+          # chop off last character, and retry if we still have enough string to
+          # look like a minimal URL
+          retry if urlstr.chop! and urlstr =~ /^https?:\/\/./
+        end
+        reply = "Error #{e.message}"
       end
 
       if display_info > urls_displayed
-        if title
-          m.reply("#{LINK_INFO} #{title}", :overlong => :truncate)
+        if reply
+          m.reply reply, :overlong => :truncate, :to => :public,
+            :nick => (m.address? ? :auto : false)
           urls_displayed += 1
         end
       end
@@ -155,17 +202,21 @@ class UrlPlugin < Plugin
   def info(m, params)
     escaped = URI.escape(params[:urls].to_s, OUR_UNSAFE)
     urls = URI.extract(escaped)
-    Thread.new { handle_urls(m, urls, params[:urls].length) }
+    Thread.new do
+      handle_urls(m,
+                  :urls => urls,
+                  :display_info => params[:urls].length,
+                  :channels => [])
+    end
   end
 
-  def listen(m)
-    return unless m.kind_of?(PrivMessage)
+  def message(m)
     return if m.address?
 
     escaped = URI.escape(m.message, OUR_UNSAFE)
     urls = URI.extract(escaped, ['http', 'https'])
     return if urls.empty?
-    Thread.new { handle_urls(m, urls) }
+    Thread.new { handle_urls(m, :urls => urls) }
   end
 
   def reply_urls(opts={})