-
- puts "+ connecting to #{url.host}:#{url.port}"
- http = @bot.httputil.get_proxy(url)
- title = http.start { |http|
- url.path = '/' if url.path == ''\r
-\r
- http.request_get(url.path) { |response|\r
-
- case response
- when Net::HTTPRedirection then
- # call self recursively if this is a redirect
- redirect_to = response['location']
- puts "+ redirect location: #{redirect_to}"
- url = URI.join url.to_s, redirect_to
- puts "+ whee, redirecting to #{url.to_s}!"
- title = get_title_for_url(url.to_s, depth-1)
- when Net::HTTPSuccess then
- if response['content-type'] =~ /^text\//
- # since the content is 'text/*' and is small enough to
- # be a webpage, retrieve the title from the page
- puts "+ getting #{url.request_uri}"\r
- data = read_data_from_response(response, 50000)\r
- return get_title_from_html(data)
- else
- # content doesn't have title, just display info.
- size = response['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2')
- return "[Link Info] type: #{response['content-type']}#{size ? ", size: #{size} bytes" : ""}"
- end
- when Net::HTTPClientError then
- return "[Link Info] Error getting link (#{response.code} - #{response.message})"
- when Net::HTTPServerError then
- return "[Link Info] Error getting link (#{response.code} - #{response.message})"
- end # end of "case response"\r
- \r
- } # end of request block
- } # end of http start block\r
-
- rescue SocketError => e
- return "[Link Info] Error connecting to site (#{e.message})"
+
+ # also check the ip, the canonical name and the aliases
+ begin
+ checks = TCPSocket.gethostbyname(url.host)
+ checks.delete_at(-2)
+ rescue => e
+ return "Unable to retrieve info for #{url.host}: #{e.message}"
+ end
+
+ checks << url.host
+ checks.flatten!
+
+ unless checks.grep(@no_info_hosts).empty?
+ return "Sorry, info retrieval for #{url.host} (#{checks.first}) is disabled"
+ end
+
+ logopts = opts.dup
+
+ title = nil
+ extra = []
+
+ begin
+ debug "+ getting info for #{url.request_uri}"
+ info = @bot.filter(:htmlinfo, url)
+ debug info
+ resp = info[:headers]
+
+ logopts[:title] = title = info[:title]
+
+ if info[:content]
+ logopts[:extra] = info[:content]
+ extra << "#{Bold}text#{Bold}: #{info[:content]}" if @bot.config['url.first_par']
+ else
+ logopts[:extra] = String.new
+ logopts[:extra] << "Content Type: #{resp['content-type']}"
+ extra << "#{Bold}type#{Bold}: #{resp['content-type']}" unless title
+ if enc = resp['content-encoding']
+ logopts[:extra] << ", encoding: #{enc}"
+ extra << "#{Bold}encoding#{Bold}: #{enc}" if @bot.config['url.first_par'] or not title
+ end
+
+ size = resp['content-length'].first.gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2') rescue nil
+ if size
+ logopts[:extra] << ", size: #{size} bytes"
+ extra << "#{Bold}size#{Bold}: #{size} bytes" if @bot.config['url.first_par'] or not title
+ end
+ end
+ rescue Exception => e
+ case e
+ when UrlLinkError
+ raise e
+ else
+ error e
+ raise "connecting to site/processing information (#{e.message})"
+ end
+ end
+
+ call_event(:url_added, url.to_s, logopts)
+ if title
+ extra.unshift("#{Bold}title#{Bold}: #{title}")
+ end
+ 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'])
+ return if urls.empty?
+ debug "found urls #{urls.inspect}"
+ list = m.public? ? @registry[m.target] : nil
+ debug "display link info: #{display_info}"
+ urls_displayed = 0
+ urls.each do |urlstr|
+ debug "working on #{urlstr}"
+ next unless urlstr =~ /^https?:/
+ title = nil
+ debug "Getting title for #{urlstr}..."
+ reply = nil
+ begin
+ title = get_title_for_url(urlstr,
+ :nick => m.source.nick,
+ :channel => m.channel,
+ :ircline => m.message)
+ debug "Title #{title ? '' : 'not '} found"
+ reply = "#{LINK_INFO} #{title}" if title
+ rescue => e
+ 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 reply
+ m.plainreply(reply, :overlong => :truncate)
+ urls_displayed += 1
+ end
+ end
+
+ next unless list
+
+ # check to see if this url is already listed
+ next if list.find {|u| u.url == urlstr }
+
+ url = Url.new(m.target, m.sourcenick, Time.new, urlstr, title)
+ debug "#{list.length} urls so far"
+ list.pop if list.length > @bot.config['url.max_urls']
+ debug "storing url #{url.url}"
+ list.unshift url
+ debug "#{list.length} urls now"
+ end
+ @registry[m.target] = list
+ 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) }