: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
@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
checks.flatten!
unless checks.grep(@no_info_hosts).empty?
- return "Sorry, info retrieval for #{url.host} (#{checks.first}) is disabled"
+ return ( opts[:always_reply] ? "Sorry, info retrieval for #{url.host} (#{checks.first}) is disabled" : false )
end
logopts = opts.dup
begin
debug "+ getting info for #{url.request_uri}"
- info = Utils.get_html_info(url)
+ info = @bot.filter(:htmlinfo, url)
debug info
resp = info[:headers]
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
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
- if e.message =~ /\(404 - Not Found\)/i
- # see if we failed to find the thing because of trailing punctuation
- # but check that we still have 'something' in the URL
+ 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}"
if display_info > urls_displayed
if reply
- m.plainreply(reply, :overlong => :truncate)
+ m.reply reply, :overlong => :truncate, :to => :public,
+ :nick => (m.address? ? :auto : false)
urls_displayed += 1
end
end
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={})