-require 'net/http'
-require 'uri'
-require 'cgi'
+define_structure :Url, :channel, :nick, :time, :url, :info
-Url = Struct.new("Url", :channel, :nick, :time, :url) unless defined?(Url)
-TITLE_RE = /<\s*?title\s*?>(.+?)<\s*?\/title\s*?>/im
-
-UNESCAPE_TABLE = {
- 'raquo' => '>>',
- 'quot' => '"',
- 'micro' => 'u',
- 'copy' => '(c)',
- 'trade' => '(tm)',
- 'reg' => '(R)',
- '#174' => '(R)',
- '#8220' => '"',
- '#8221' => '"',
- '#8212' => '--',
- '#39' => '\'',
-=begin
- # extras codes, for future use...
- 'zwnj' => '‌',
- 'aring' => '\xe5',
- 'gt' => '>',
- 'yen' => '\xa5',
- 'ograve' => '\xf2',
- 'Chi' => 'Χ',
- 'bull' => '•',
- 'Egrave' => '\xc8',
- 'Ntilde' => '\xd1',
- 'upsih' => 'ϒ',
- 'Yacute' => '\xdd',
- 'asymp' => '≈',
- 'radic' => '√',
- 'otimes' => '⊗',
- 'nabla' => '∇',
- 'aelig' => '\xe6',
- 'oelig' => 'œ',
- 'equiv' => '≡',
- 'Psi' => 'Ψ',
- 'auml' => '\xe4',
- 'circ' => 'ˆ',
- 'Acirc' => '\xc2',
- 'Epsilon' => 'Ε',
- 'Yuml' => 'Ÿ',
- 'Eta' => 'Η',
- 'lt' => '<',
- 'Icirc' => '\xce',
- 'Upsilon' => 'Υ',
- 'ndash' => '–',
- 'there4' => '∴',
- 'Prime' => '″',
- 'prime' => '′',
- 'psi' => 'ψ',
- 'Kappa' => 'Κ',
- 'rsaquo' => '›',
- 'Tau' => 'Τ',
- 'darr' => '↓',
- 'ocirc' => '\xf4',
- 'lrm' => '‎',
- 'zwj' => '‍',
- 'cedil' => '\xb8',
- 'Ecirc' => '\xca',
- 'not' => '\xac',
- 'amp' => '&',
- 'AElig' => '\xc6',
- 'oslash' => '\xf8',
- 'acute' => '\xb4',
- 'lceil' => '⌈',
- 'laquo' => '\xab',
- 'shy' => '\xad',
- 'rdquo' => '”',
- 'ge' => '≥',
- 'Igrave' => '\xcc',
- 'Ograve' => '\xd2',
- 'euro' => '€',
- 'dArr' => '⇓',
- 'sdot' => '⋅',
- 'nbsp' => '\xa0',
- 'lfloor' => '⌊',
- 'lArr' => '⇐',
- 'Auml' => '\xc4',
- 'larr' => '←',
- 'Atilde' => '\xc3',
- 'Otilde' => '\xd5',
- 'szlig' => '\xdf',
- 'clubs' => '♣',
- 'diams' => '♦',
- 'agrave' => '\xe0',
- 'Ocirc' => '\xd4',
- 'Iota' => 'Ι',
- 'Theta' => 'Θ',
- 'Pi' => 'Π',
- 'OElig' => 'Œ',
- 'Scaron' => 'Š',
- 'frac14' => '\xbc',
- 'egrave' => '\xe8',
- 'sub' => '⊂',
- 'iexcl' => '\xa1',
- 'frac12' => '\xbd',
- 'sbquo' => '‚',
- 'ordf' => '\xaa',
- 'sum' => '∑',
- 'prop' => '∝',
- 'Uuml' => '\xdc',
- 'ntilde' => '\xf1',
- 'sup' => '⊃',
- 'theta' => 'θ',
- 'prod' => '∏',
- 'nsub' => '⊄',
- 'hArr' => '⇔',
- 'rlm' => '‏',
- 'THORN' => '\xde',
- 'infin' => '∞',
- 'yuml' => '\xff',
- 'Mu' => 'Μ',
- 'le' => '≤',
- 'Eacute' => '\xc9',
- 'thinsp' => ' ',
- 'ecirc' => '\xea',
- 'bdquo' => '„',
- 'Sigma' => 'Σ',
- 'fnof' => 'ƒ',
- 'Aring' => '\xc5',
- 'tilde' => '˜',
- 'frac34' => '\xbe',
- 'emsp' => ' ',
- 'mdash' => '—',
- 'uarr' => '↑',
- 'permil' => '‰',
- 'Ugrave' => '\xd9',
- 'rarr' => '→',
- 'Agrave' => '\xc0',
- 'chi' => 'χ',
- 'forall' => '∀',
- 'eth' => '\xf0',
- 'rceil' => '⌉',
- 'iuml' => '\xef',
- 'gamma' => 'γ',
- 'lambda' => 'λ',
- 'harr' => '↔',
- 'rang' => '〉',
- 'xi' => 'ξ',
- 'dagger' => '†',
- 'divide' => '\xf7',
- 'Ouml' => '\xd6',
- 'image' => 'ℑ',
- 'alefsym' => 'ℵ',
- 'igrave' => '\xec',
- 'otilde' => '\xf5',
- 'Oacute' => '\xd3',
- 'sube' => '⊆',
- 'alpha' => 'α',
- 'frasl' => '⁄',
- 'ETH' => '\xd0',
- 'lowast' => '∗',
- 'Nu' => 'Ν',
- 'plusmn' => '\xb1',
- 'Euml' => '\xcb',
- 'real' => 'ℜ',
- 'sup1' => '\xb9',
- 'sup2' => '\xb2',
- 'sup3' => '\xb3',
- 'Oslash' => '\xd8',
- 'Aacute' => '\xc1',
- 'cent' => '\xa2',
- 'oline' => '‾',
- 'Beta' => 'Β',
- 'perp' => '⊥',
- 'Delta' => 'Δ',
- 'loz' => '◊',
- 'pi' => 'π',
- 'iota' => 'ι',
- 'empty' => '∅',
- 'euml' => '\xeb',
- 'brvbar' => '\xa6',
- 'iacute' => '\xed',
- 'para' => '\xb6',
- 'micro' => '\xb5',
- 'cup' => '∪',
- 'weierp' => '℘',
- 'uuml' => '\xfc',
- 'part' => '∂',
- 'icirc' => '\xee',
- 'delta' => 'δ',
- 'omicron' => 'ο',
- 'upsilon' => 'υ',
- 'Iuml' => '\xcf',
- 'Lambda' => 'Λ',
- 'Xi' => 'Ξ',
- 'kappa' => 'κ',
- 'ccedil' => '\xe7',
- 'Ucirc' => '\xdb',
- 'cap' => '∩',
- 'mu' => 'μ',
- 'scaron' => 'š',
- 'lsquo' => '‘',
- 'isin' => '∈',
- 'Zeta' => 'Ζ',
- 'supe' => '⊇',
- 'deg' => '\xb0',
- 'and' => '∧',
- 'tau' => 'τ',
- 'pound' => '\xa3',
- 'hellip' => '…',
- 'curren' => '\xa4',
- 'int' => '∫',
- 'ucirc' => '\xfb',
- 'rfloor' => '⌋',
- 'ensp' => ' ',
- 'crarr' => '↵',
- 'ugrave' => '\xf9',
- 'notin' => '∉',
- 'exist' => '∃',
- 'uArr' => '⇑',
- 'cong' => '≅',
- 'Dagger' => '‡',
- 'oplus' => '⊕',
- 'times' => '\xd7',
- 'atilde' => '\xe3',
- 'piv' => 'ϖ',
- 'ni' => '∋',
- 'Phi' => 'Φ',
- 'lsaquo' => '‹',
- 'Uacute' => '\xda',
- 'Omicron' => 'Ο',
- 'ang' => '∠',
- 'ne' => '≠',
- 'iquest' => '\xbf',
- 'eta' => 'η',
- 'yacute' => '\xfd',
- 'Rho' => 'Ρ',
- 'uacute' => '\xfa',
- 'Alpha' => 'Α',
- 'zeta' => 'ζ',
- 'Omega' => 'Ω',
- 'nu' => 'ν',
- 'sim' => '∼',
- 'sect' => '\xa7',
- 'phi' => 'φ',
- 'sigmaf' => 'ς',
- 'macr' => '\xaf',
- 'minus' => '−',
- 'Ccedil' => '\xc7',
- 'ordm' => '\xba',
- 'epsilon' => 'ε',
- 'beta' => 'β',
- 'rArr' => '⇒',
- 'rho' => 'ρ',
- 'aacute' => '\xe1',
- 'eacute' => '\xe9',
- 'omega' => 'ω',
- 'middot' => '\xb7',
- 'Gamma' => 'Γ',
- 'Iacute' => '\xcd',
- 'lang' => '〈',
- 'spades' => '♠',
- 'rsquo' => '’',
- 'uml' => '\xa8',
- 'thorn' => '\xfe',
- 'ouml' => '\xf6',
- 'thetasym' => 'ϑ',
- 'or' => '∨',
- 'raquo' => '\xbb',
- 'acirc' => '\xe2',
- 'ldquo' => '“',
- 'hearts' => '♥',
- 'sigma' => 'σ',
- 'oacute' => '\xf3',
-=end
-}
+class ::UrlLinkError < RuntimeError
+end
class UrlPlugin < Plugin
+ TITLE_RE = /<\s*?title\s*?>(.+?)<\s*?\/title\s*?>/im
+ LINK_INFO = "[Link Info]"
+
BotConfig.register BotConfigIntegerValue.new('url.max_urls',
: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.)")
+ BotConfig.register BotConfigBooleanValue.new('url.first_par',
+ :default => false,
+ :desc => "Also try to get the first paragraph of a web page")
+ BotConfig.register BotConfigBooleanValue.new('url.info_on_list',
+ :default => false,
+ :desc => "Show link info when listing/searching for urls")
+
+
def initialize
super
@registry.set_default(Array.new)
"urls [<max>=4] => list <max> last urls mentioned in current channel, urls search [<max>=4] <regexp> => search for matching urls. In a private message, you must specify the channel to query, eg. urls <channel> [max], urls search <channel> [max] <regexp>"
end
- def unescape_title(htmldata)
- # first pass -- let CGI try to attack it...
- htmldata = CGI::unescapeHTML htmldata
-
- # second pass -- destroy the remaining bits...
- htmldata.gsub(/(&(.+?);)/) {
- symbol = $2
-
- # remove the 0-paddng from unicode integers
- if symbol =~ /#(.+)/
- symbol = "##{$1.to_i.to_s}"
- end
-
- # output the symbol's irc-translated character, or a * if it's unknown
- UNESCAPE_TABLE[symbol] || '*'
- }
- end
-
def get_title_from_html(pagedata)
return unless TITLE_RE.match(pagedata)
- title = $1.strip.gsub(/\s*\n+\s*/, " ")
- title = unescape_title title
- title = title[0..255] if title.length > 255
- "[Link Info] title: #{title}"
+ $1.ircify_html
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, nick = nil, channel = nil)
-
- def get_title_for_url(uri_str, depth=10)
- # 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}"
- url = URI.parse(uri_str)
+ 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|
- url.path = '/' if url.path == ''
+ logopts = Hash.new
+ logopts[:nick] = nick if nick
+ logopts[:channel] = channel if channel
- http.request_get(url.path, "User-Agent" => "Mozilla/4.0 (compatible; MSIE 6.0; Windows NT 5.1; SV1; .NET CLR 1.1.4322)") { |response|
-
- case response
- when Net::HTTPRedirection, Net::HTTPMovedPermanently then
- # 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.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
- debug "+ getting #{url.request_uri}"
- data = read_data_from_response(response, 50000)
- return get_title_from_html(data)
+ title = nil
+ extra = String.new
+
+ begin
+ debug "+ getting #{url.request_uri}"
+ @bot.httputil.get_response(url) { |resp|
+ case resp
+ when Net::HTTPSuccess
+
+ debug resp.to_hash
+
+ if resp['content-type'] =~ /^text\/|(?:x|ht)ml/
+ # The page is text or HTML, so we can try finding a title and, if
+ # requested, the first par.
+ #
+ # We act differently depending on whether we want the first par or
+ # not: in the first case we download the initial part and the parse
+ # it; in the second case we only download as much as we need to find
+ # the title
+ #
+ if @bot.config['url.first_par']
+ partial = resp.partial_body(@bot.config['http.info_bytes'])
+ logopts[:title] = title = get_title_from_html(partial)
+ first_par = Utils.ircify_first_html_par(partial, :strip => title)
+ unless first_par.empty?
+ logopts[:extra] = first_par
+ extra << ", #{Bold}text#{Bold}: #{first_par}"
+ end
+ call_event(:url_added, url.to_s, logopts)
+ return "#{Bold}title#{Bold}: #{title}#{extra}" if title
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" : ""}"
+ resp.partial_body(@bot.config['http.info_bytes']) { |part|
+ logopts[:title] = title = get_title_from_html(part)
+ call_event(:url_added, url.to_s, logopts)
+ return "#{Bold}title#{Bold}: #{title}" if title
+ }
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})"
+ # if nothing was found, provide more basic info, as for non-html pages
else
- return nil
- 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})"
+ resp.no_cache = true
+ end
+
+ enc = resp['content-encoding']
+ logopts[:extra] = String.new
+ logopts[:extra] << "Content Type: #{resp['content-type']}"
+ if enc
+ logopts[:extra] << ", encoding: #{enc}"
+ extra << ", #{Bold}encoding#{Bold}: #{enc}"
+ end
+
+ unless @bot.config['url.titles_only']
+ # content doesn't have title, just display info.
+ size = resp['content-length'].gsub(/(\d)(?=\d{3}+(?:\.|$))(\d{3}\..*)?/,'\1,\2') rescue nil
+ if size
+ logopts[:extra] << ", size: #{size} bytes"
+ size = ", #{Bold}size#{Bold}: #{size} bytes"
+ end
+ call_event(:url_added, url.to_s, logopts)
+ return "#{Bold}type#{Bold}: #{resp['content-type']}#{size}#{extra}"
+ end
+ call_event(:url_added, url.to_s, logopts)
+ else
+ raise UrlLinkError, "getting link (#{resp.code} - #{resp.message})"
+ end
+ }
+ return nil
+ rescue Exception => e
+ case e
+ when UrlLinkError
+ raise e
+ else
+ error e
+ raise "connecting to site/processing information (#{e.message})"
+ end
+ end
end
def listen(m)
urlstr = $1
list = @registry[m.target]
+ title = nil
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, m.source.nick, m.channel
+ if title
+ m.reply "#{LINK_INFO} #{title}", :overlong => :truncate
+ debug "Title found!"
+ else
+ debug "Title not found!"
+ end
+ rescue => e
+ m.reply "Error #{e.message}"
+ 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)
+
+ url = Url.new(m.target, m.sourcenick, Time.new, urlstr, title)
debug "#{list.length} urls so far"
if list.length > @bot.config['url.max_urls']
list.pop
end
end
+ def reply_urls(opts={})
+ list = opts[:list]
+ max = opts[:max]
+ channel = opts[:channel]
+ m = opts[:msg]
+ return unless list and max and m
+ list[0..(max-1)].each do |url|
+ disp = "[#{url.time.strftime('%Y/%m/%d %H:%M:%S')}] <#{url.nick}> #{url.url}"
+ if @bot.config['url.info_on_list']
+ title = url.info || get_title_for_url(url.url, url.nick, channel) rescue nil
+ # If the url info was missing and we now have some, try to upgrade it
+ if channel and title and not url.info
+ ll = @registry[channel]
+ debug ll
+ if el = ll.find { |u| u.url == url.url }
+ el.info = title
+ @registry[channel] = ll
+ end
+ end
+ disp << " --> #{title}" if title
+ end
+ m.reply disp, :overlong => :truncate
+ end
+ end
+
def urls(m, params)
channel = params[:channel] ? params[:channel] : m.target
max = params[:limit].to_i
if list.empty?
m.reply "no urls seen yet for channel #{channel}"
else
- list[0..(max-1)].each do |url|
- m.reply "[#{url.time.strftime('%Y/%m/%d %H:%M:%S')}] <#{url.nick}> #{url.url}"
- end
+ reply_urls :msg => m, :channel => channel, :list => list, :max => max
end
end
max = 1 if max < 1
regex = Regexp.new(string, Regexp::IGNORECASE)
list = @registry[channel].find_all {|url|
- regex.match(url.url) || regex.match(url.nick)
+ regex.match(url.url) || regex.match(url.nick) ||
+ (@bot.config['url.info_on_list'] && regex.match(url.info))
}
if list.empty?
m.reply "no matches for channel #{channel}"
else
- list[0..(max-1)].each do |url|
- m.reply "[#{url.time.strftime('%Y/%m/%d %H:%M:%S')}] <#{url.nick}> #{url.url}"
- end
+ reply_urls :msg => m, :channel => channel, :list => list, :max => max
end
end
end
+
plugin = UrlPlugin.new
plugin.map 'urls search :channel :limit :string', :action => 'search',
:defaults => {:limit => 4},