:default => 100, :validate => Proc.new{|v| v > 0},
:desc => "Maximum number of urls to store. New urls replace oldest ones.")
BotConfig.register BotConfigBooleanValue.new('url.display_link_info',
- :default => false,
+ :default => false,
:desc => "Get the title of any links pasted to the channel and display it (also tells if the link is broken or the site is down)")
-
+ 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.)")
+
def initialize
super
@registry.set_default(Array.new)
return unless TITLE_RE.match(pagedata)
title = $1.strip.gsub(/\s*\n+\s*/, " ")
title = Utils.decode_html_entities title
- title = title[0..255] if title.length > 255
- "[Link Info] title: #{title}"
+ "title: #{title}"
end
- def read_data_from_response(response, amount)
-
- amount_read = 0
- chunks = []
-
- response.read_body do |chunk| # read body now
-
- amount_read += chunk.length
-
- if amount_read > amount
- amount_of_overflow = amount_read - amount
- chunk = chunk[0...-amount_of_overflow]
- end
-
- chunks << chunk
-
- break if amount_read >= amount
-
- end
-
- chunks.join('')
-
- end
+ def get_title_for_url(uri_str)
- def get_title_for_url(uri_str, depth=@bot.config['http.max_redir'])
- # This god-awful mess is what the ruby http library has reduced me to.
- # Python's HTTP lib is so much nicer. :~(
-
- if depth == 0
- raise "Error: Maximum redirects hit."
- end
-
- debug "+ Getting #{uri_str.to_s}"
url = uri_str.kind_of?(URI) ? uri_str : URI.parse(uri_str)
return if url.scheme !~ /https?/
title = nil
-
- debug "+ connecting to #{url.host}:#{url.port}"
- http = @bot.httputil.get_proxy(url)
- http.start { |http|
- http.request_get(url.request_uri(), @bot.httputil.headers) { |response|
-
+ begin
+ @bot.httputil.get_response(url) { |response|
case response
- when Net::HTTPRedirection
- # call self recursively if this is a redirect
- redirect_to = response['location'] || '/'
- debug "+ redirect location: #{redirect_to.inspect}"
- url = URI.join(url.to_s, redirect_to)
- debug "+ whee, redirecting to #{url.to_s}!"
- return get_title_for_url(url, depth-1)
- 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}"
- # was 5*10^4 ... seems to much to me ... 4k should be enough for everybody ;)
- data = read_data_from_response(response, 4096)
- return get_title_from_html(data)
- else
+ 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 "[Link Info] type: #{response['content-type']}#{size}"
+ return "type: #{response['content-type']}#{size}"
end
- else
- return "[Link Info] Error getting link (#{response.code} - #{response.message})"
- end # end of "case response"
-
- } # end of request block
- } # end of http start block
-
- return title
-
- rescue SocketError => e
- return "[Link Info] Error connecting to site (#{e.message})"
+ end
+ when Net::HTTPResponse
+ return "Error getting link (#{response.code} - #{response.message})"
+ else
+ raise response
+ end
+ }
+ rescue Object => e
+ if e.class <= StandardError
+ error e.inspect
+ debug e.backtrace.join("\n")
+ end
+
+ msg = e.respond_to?(:message) ? e.message : e.to_s
+ return "Error connecting to site (#{e.message})"
+ end
end
def listen(m)
list = @registry[m.target]
if @bot.config['url.display_link_info']
- debug "Getting title for #{urlstr}..."
- title = get_title_for_url urlstr
- if title
- m.reply title
- debug "Title found!"
- else
- debug "Title not found!"
- end
+ Thread.start do
+ debug "Getting title for #{urlstr}..."
+ begin
+ title = get_title_for_url urlstr
+ if title
+ m.reply "[Link Info] #{title}"
+ debug "Title found!"
+ else
+ debug "Title not found!"
+ end
+ rescue => e
+ debug "Failed: #{e}"
+ end
+ end
end
-
+
# check to see if this url is already listed
return if list.find {|u| u.url == urlstr }
-
+
url = Url.new(m.target, m.sourcenick, Time.new, urlstr)
debug "#{list.length} urls so far"
if list.length > @bot.config['url.max_urls']
projects / user / henk / code / ruby / rbot.git / blobdiff