-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',
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
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[0..255] if title.length > 255
- "[Link Info] title: #{title}"
+ $1.ircify_html
end
- def read_data_from_response(response, amount)
+ def get_title_for_url(uri_str)
- amount_read = 0
- chunks = []
-
- response.read_body do |chunk| # read body now
+ url = uri_str.kind_of?(URI) ? uri_str : URI.parse(uri_str)
+ return if url.scheme !~ /https?/
- amount_read += chunk.length
+ title = nil
- # if amount_read > amount
- # amount_of_overflow = amount_read - amount
- # chunk = chunk[0...-amount_of_overflow]
- # end
+ begin
+ range = @bot.config['http.info_bytes']
+ response = @bot.httputil.get_response(url, :range => "bytes=0-#{range}")
+ if response.code != "206" && response.code != "200"
+ return "Error getting link (#{response.code} - #{response.message})"
+ end
+ extra = String.new
- chunks << chunk
+ if response['content-type'] =~ /^text\//
- break if amount_read >= amount
+ body = response.body.slice(0, range)
+ title = String.new
- end
+ # since the content is 'text/*' and is small enough to
+ # be a webpage, retrieve the title from the page
+ debug "+ getting #{url.request_uri}"
- chunks.join('')
-
- end
+ title = get_title_from_html(body)
+ if @bot.config['url.first_par']
+ first_par = Utils.ircify_first_html_par(body, :strip => title)
+ extra << ", #{Bold}text#{Bold}: #{first_par}" unless first_par.empty?
+ return "#{Bold}title#{Bold}: #{title}#{extra}" if title
+ else
+ return "#{Bold}title#{Bold}: #{title}" if title
+ end
- def get_title_for_url(uri_str, depth=@bot.config['http.max_redir'])
- # This god-awful mess is what the ruby http library has reduced me to.
- # Python's HTTP lib is so much nicer. :~(
+ # if nothing was found, provide more basic info
+ end
- if depth == 0
- raise "Error: Maximum redirects hit."
+ debug response.to_hash.inspect
+ unless @bot.config['url.titles_only']
+ # content doesn't have title, just display info.
+ size = response['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2') rescue nil
+ if response.code == '206'
+ if response['content-range'] =~ /bytes\s*[^\/]+\/(\d+)/
+ size = $1.to_s.reverse.scan(/\d{1,3}/).join(',').reverse
+ end
+ end
+ size = size ? ", #{Bold}size#{Bold}: #{size} bytes" : ""
+ return "#{Bold}type#{Bold}: #{response['content-type']}#{size}#{extra}"
+ end
+ rescue Exception => e
+ error e.inspect
+ debug e.backtrace.join("\n")
+ return "Error connecting to site (#{e.message})"
end
-
- debug "+ Getting #{uri_str.to_s}"
- url = uri_str.kind_of?(URI) ? uri_str : URI.parse(uri_str)
- return if url.scheme !~ /https?/
-
- title = nil
-
- debug "+ connecting to #{url.host}:#{url.port}"
- http = @bot.httputil.get_proxy(url)
- http.start { |http|
-
- http.request_get(url.request_uri(), @bot.httputil.headers) { |response|
-
- case response
- when Net::HTTPRedirection
- # call self recursively if this is a redirect
- redirect_to = response['location'] || '/'
- debug "+ redirect location: #{redirect_to.inspect}"
- url = URI.join(url.to_s, redirect_to)
- debug "+ whee, redirecting to #{url.to_s}!"
- return get_title_for_url(url, depth-1)
- when Net::HTTPSuccess
- if response['content-type'] =~ /^text\//
- # since the content is 'text/*' and is small enough to
- # be a webpage, retrieve the title from the page
- debug "+ getting #{url.request_uri}"
- # was 5*10^4 ... seems to much to me ... 4k should be enough for everybody ;)
- data = read_data_from_response(response, 4096)
- return get_title_from_html(data)
- else
- unless @bot.config['url.titles_only']
- # content doesn't have title, just display info.
- size = response['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2')
- size = size ? ", size: #{size} bytes" : ""
- return "[Link Info] type: #{response['content-type']}#{size}"
- end
- 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
def listen(m)
list = @registry[m.target]
if @bot.config['url.display_link_info']
- debug "Getting title for #{urlstr}..."
- begin
- title = get_title_for_url urlstr
- if title
- m.reply title
- debug "Title found!"
- else
- debug "Title not found!"
- end
- rescue => e
- debug "Failed: #{e}"
+ Thread.start do
+ debug "Getting title for #{urlstr}..."
+ begin
+ title = get_title_for_url urlstr
+ if title
+ m.reply "#{LINK_INFO} #{title}", :overlong => :truncate
+ debug "Title found!"
+ else
+ debug "Title not found!"
+ end
+ rescue => e
+ debug "Failed: #{e}"
+ end
end
end
projects / user / henk / code / ruby / rbot.git / blobdiff