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)]
30 # Extensions to the Range class
34 # This method returns a random number between the lower and upper bound
37 len = self.last - self.first
38 len += 1 unless self.exclude_end?
39 self.first + Kernel::rand(len)
44 # Extensions for the Numeric classes
48 # This method forces a real number to be not more than a given positive
49 # number or not less than a given positive number, or between two any given
52 def clip(left,right=0)
53 raise ArgumentError unless left.kind_of?(Numeric) and right.kind_of?(Numeric)
62 # Extensions to the String class
64 # TODO make ircify_html() accept an Hash of options, and make riphtml() just
65 # call ircify_html() with stronger purify options.
69 # This method will return a purified version of the receiver, with all HTML
70 # stripped off and some of it converted to IRC formatting
72 def ircify_html(opts={})
76 txt.gsub!(/<script(?:\s+[^>]*)?>.*?<\/script>/im, "")
79 txt.gsub!(/<style(?:\s+[^>]*)?>.*?<\/style>/im, "")
81 # bold and strong -> bold
82 txt.gsub!(/<\/?(?:b|strong)(?:\s+[^>]*)?>/im, "#{Bold}")
84 # italic, emphasis and underline -> underline
85 txt.gsub!(/<\/?(?:i|em|u)(?:\s+[^>]*)?>/im, "#{Underline}")
87 ## This would be a nice addition, but the results are horrible
88 ## Maybe make it configurable?
89 # txt.gsub!(/<\/?a( [^>]*)?>/, "#{Reverse}")
90 case val = opts[:a_href]
91 when Reverse, Bold, Underline
92 txt.gsub!(/<(?:\/a\s*|a (?:[^>]*\s+)?href\s*=\s*(?:[^>]*\s*)?)>/, val)
94 # Not good for nested links, but the best we can do without something like hpricot
95 txt.gsub!(/<a (?:[^>]*\s+)?href\s*=\s*(?:([^"'>][^\s>]*)\s+|"((?:[^"]|\\")*)"|'((?:[^']|\\')*)')(?:[^>]*\s+)?>(.*?)<\/a>/) { |match|
97 debug [$1, $2, $3, $4].inspect
103 warn "unknown :a_href option #{val} passed to ircify_html" if val
106 # Paragraph and br tags are converted to whitespace
107 txt.gsub!(/<\/?(p|br)(?:\s+[^>]*)?\s*\/?\s*>/i, ' ')
111 # Superscripts and subscripts are turned into ^{...} and _{...}
112 # where the {} are omitted for single characters
113 txt.gsub!(/<sup>(.*?)<\/sup>/, '^{\1}')
114 txt.gsub!(/<sub>(.*?)<\/sub>/, '_{\1}')
115 txt.gsub!(/(^|_)\{(.)\}/, '\1\2')
117 # All other tags are just removed
118 txt.gsub!(/<[^>]+>/, '')
120 # Convert HTML entities. We do it now to be able to handle stuff
122 txt = Utils.decode_html_entities(txt)
124 # Remove double formatting options, since they only waste bytes
125 txt.gsub!(/#{Bold}(\s*)#{Bold}/, '\1')
126 txt.gsub!(/#{Underline}(\s*)#{Underline}/, '\1')
128 # Simplify whitespace that appears on both sides of a formatting option
129 txt.gsub!(/\s+(#{Bold}|#{Underline})\s+/, ' \1')
130 txt.sub!(/\s+(#{Bold}|#{Underline})\z/, '\1')
131 txt.sub!(/\A(#{Bold}|#{Underline})\s+/, '\1')
133 # And finally whitespace is squeezed
134 txt.gsub!(/\s+/, ' ')
136 # Decode entities and strip whitespace
140 # As above, but modify the receiver
142 def ircify_html!(opts={})
144 replace self.ircify_html(opts)
145 return self unless self.hash == old_hash
148 # This method will strip all HTML crud from the receiver
151 self.gsub(/<[^>]+>/, '').gsub(/&/,'&').gsub(/"/,'"').gsub(/</,'<').gsub(/>/,'>').gsub(/&ellip;/,'...').gsub(/'/, "'").gsub("\n",'')
156 # Extensions to the Regexp class, with some common and/or complex regular
161 # A method to build a regexp that matches a list of something separated by
162 # optional commas and/or the word "and", an optionally repeated prefix,
164 def Regexp.new_list(reg, pfx = "")
165 if pfx.kind_of?(String) and pfx.empty?
166 return %r(#{reg}(?:,?(?:\s+and)?\s+#{reg})*)
168 return %r(#{reg}(?:,?(?:\s+and)?(?:\s+#{pfx})?\s+#{reg})*)
175 # Match a list of channel anmes separated by optional commas, whitespace
176 # and optionally the word "and"
177 CHAN_LIST = Regexp.new_list(GEN_CHAN)
179 # Match "in #channel" or "on #channel" and/or "in private" (optionally
180 # shortened to "in pvt"), returning the channel name or the word 'private'
181 # or 'pvt' as capture
182 IN_CHAN = /#{IN_ON}\s+(#{GEN_CHAN})|(here)|/
183 IN_CHAN_PVT = /#{IN_CHAN}|in\s+(private|pvt)/
185 # As above, but with channel lists
186 IN_CHAN_LIST_SFX = Regexp.new_list(/#{GEN_CHAN}|here/, IN_ON)
187 IN_CHAN_LIST = /#{IN_ON}\s+#{IN_CHAN_LIST_SFX}|anywhere|everywhere/
188 IN_CHAN_LIST_PVT_SFX = Regexp.new_list(/#{GEN_CHAN}|here|private|pvt/, IN_ON)
189 IN_CHAN_LIST_PVT = /#{IN_ON}\s+#{IN_CHAN_LIST_PVT_SFX}|anywhere|everywhere/
191 # Match a list of nicknames separated by optional commas, whitespace and
192 # optionally the word "and"
193 NICK_LIST = Regexp.new_list(GEN_NICK)
203 class BasicUserMessage
205 # We extend the BasicUserMessage class with a method that parses a string
206 # which is a channel list as matched by IN_CHAN(_LIST) and co. The method
207 # returns an array of channel names, where 'private' or 'pvt' is replaced
208 # by the Symbol :"?", 'here' is replaced by the channel of the message or
209 # by :"?" (depending on whether the message target is the bot or a
210 # Channel), and 'anywhere' and 'everywhere' are replaced by Symbol :*
212 def parse_channel_list(string)
213 return [:*] if [:anywhere, :everywhere].include? string.to_sym
215 /(?:^|,?(?:\s+and)?\s+)(?:in|on\s+)?(#{Regexp::Irc::GEN_CHAN}|here|private|pvt)/