4 # :title: Standard classes extensions
6 # Author:: Giuseppe "Oblomov" Bilotta <giuseppe.bilotta@gmail.com>
7 # Copyright:: (C) 2006,2007 Giuseppe Bilotta
10 # This file collects extensions to standard Ruby classes and to some core rbot
11 # classes to be used by the various plugins
13 # Please note that global symbols have to be prefixed by :: because this plugin
14 # will be read into an anonymous module
17 # Extensions to the Array class
21 # This method returns a random element from the array, or nil if the array is
25 return nil if self.empty?
26 self[rand(self.length)]
34 $we_have_iconv = false
37 # Extensions to the String class
39 # TODO make ircify_html() accept an Hash of options, and make riphtml() just
40 # call ircify_html() with stronger purify options.
44 # This method will try to transcode a String supposed to hold an XML or HTML
45 # document from the original charset to UTF-8.
47 # To find the original encoding, it will first see if the String responds to
48 # #http_headers(), and if it does it will assume that the charset indicated
49 # there is the correct one. Otherwise, it will try to detect the charset from
50 # some typical XML and HTML headers
52 return self unless $we_have_iconv
56 if self.respond_to?(:http_headers) and headers = self.http_headers
57 if headers['content-type'].first.match(/charset=(\S+?)\s*(?:;|\Z)/i)
58 debug "charset #{charset} set from header"
65 when /<\?xml.*encoding="(\S+)".*\?>/i
67 when /<meta\s+http-equiv\s*=\s*["']?Content-Type["']?.*charset\s*=\s*(\S+?)(?:;|["']|\s).*>/i
70 debug "charset #{charset} set from string"
74 return Iconv.iconv('utf-8', charset, self).join rescue self
76 debug "Couldn't find charset for #{self.inspect}"
82 # This method will return a purified version of the receiver, with all HTML
83 # stripped off and some of it converted to IRC formatting
89 txt.gsub!(/<script(?:\s+[^>]*)?>.*?<\/script>/im, "")
92 txt.gsub!(/<style(?:\s+[^>]*)?>.*?<\/style>/im, "")
94 # bold and strong -> bold
95 txt.gsub!(/<\/?(?:b|strong)(?:\s+[^>]*)?>/im, "#{Bold}")
97 # italic, emphasis and underline -> underline
98 txt.gsub!(/<\/?(?:i|em|u)(?:\s+[^>]*)?>/im, "#{Underline}")
100 ## This would be a nice addition, but the results are horrible
101 ## Maybe make it configurable?
102 # txt.gsub!(/<\/?a( [^>]*)?>/, "#{Reverse}")
104 # Paragraph and br tags are converted to whitespace
105 txt.gsub!(/<\/?(p|br)(?:\s+[^>]*)?\s*\/?\s*>/, ' ')
109 # All other tags are just removed
110 txt.gsub!(/<[^>]+>/, '')
112 # Convert HTML entities. We do it now to be able to handle stuff
114 txt = Utils.decode_html_entities(txt)
116 # Remove double formatting options, since they only waste bytes
117 txt.gsub!(/#{Bold}(\s*)#{Bold}/, '\1')
118 txt.gsub!(/#{Underline}(\s*)#{Underline}/, '\1')
120 # Simplify whitespace that appears on both sides of a formatting option
121 txt.gsub!(/\s+(#{Bold}|#{Underline})\s+/, ' \1')
122 txt.sub!(/\s+(#{Bold}|#{Underline})\z/, '\1')
123 txt.sub!(/\A(#{Bold}|#{Underline})\s+/, '\1')
125 # And finally whitespace is squeezed
126 txt.gsub!(/\s+/, ' ')
128 # Decode entities and strip whitespace
132 # This method will strip all HTML crud from the receiver
135 self.gsub(/<[^>]+>/, '').gsub(/&/,'&').gsub(/"/,'"').gsub(/</,'<').gsub(/>/,'>').gsub(/&ellip;/,'...').gsub(/'/, "'").gsub("\n",'')
140 # Extensions to the Regexp class, with some common and/or complex regular
145 # A method to build a regexp that matches a list of something separated by
146 # optional commas and/or the word "and", an optionally repeated prefix,
148 def Regexp.new_list(reg, pfx = "")
149 if pfx.kind_of?(String) and pfx.empty?
150 return %r(#{reg}(?:,?(?:\s+and)?\s+#{reg})*)
152 return %r(#{reg}(?:,?(?:\s+and)?(?:\s+#{pfx})?\s+#{reg})*)
159 # Match a list of channel anmes separated by optional commas, whitespace
160 # and optionally the word "and"
161 CHAN_LIST = Regexp.new_list(GEN_CHAN)
163 # Match "in #channel" or "on #channel" and/or "in private" (optionally
164 # shortened to "in pvt"), returning the channel name or the word 'private'
165 # or 'pvt' as capture
166 IN_CHAN = /#{IN_ON}\s+(#{GEN_CHAN})|(here)|/
167 IN_CHAN_PVT = /#{IN_CHAN}|in\s+(private|pvt)/
169 # As above, but with channel lists
170 IN_CHAN_LIST_SFX = Regexp.new_list(/#{GEN_CHAN}|here/, IN_ON)
171 IN_CHAN_LIST = /#{IN_ON}\s+#{IN_CHAN_LIST_SFX}|anywhere|everywhere/
172 IN_CHAN_LIST_PVT_SFX = Regexp.new_list(/#{GEN_CHAN}|here|private|pvt/, IN_ON)
173 IN_CHAN_LIST_PVT = /#{IN_ON}\s+#{IN_CHAN_LIST_PVT_SFX}|anywhere|everywhere/
175 # Match a list of nicknames separated by optional commas, whitespace and
176 # optionally the word "and"
177 NICK_LIST = Regexp.new_list(GEN_NICK)
187 class BasicUserMessage
189 # We extend the BasicUserMessage class with a method that parses a string
190 # which is a channel list as matched by IN_CHAN(_LIST) and co. The method
191 # returns an array of channel names, where 'private' or 'pvt' is replaced
192 # by the Symbol :"?", 'here' is replaced by the channel of the message or
193 # by :"?" (depending on whether the message target is the bot or a
194 # Channel), and 'anywhere' and 'everywhere' are replaced by Symbol :*
196 def parse_channel_list(string)
197 return [:*] if [:anywhere, :everywhere].include? string.to_sym
199 /(?:^|,?(?:\s+and)?\s+)(?:in|on\s+)?(#{Regexp::Irc::GEN_CHAN}|here|private|pvt)/