X-Git-Url: https://git.netwichtig.de/gitweb/?a=blobdiff_plain;f=lib%2Frbot%2Fcore%2Futils%2Fextends.rb;h=7b733994b74d323b010db0bc2ee6d66dad1f598d;hb=a209adb0e00e2a335333c1cb696d5807a47072c6;hp=b2b4f5951b69c869e27a4b4280edc1aa8fbcdbd5;hpb=cba46be94c0b711b6aa0709140b08ef541a345bb;p=user%2Fhenk%2Fcode%2Fruby%2Frbot.git diff --git a/lib/rbot/core/utils/extends.rb b/lib/rbot/core/utils/extends.rb index b2b4f595..7b733994 100644 --- a/lib/rbot/core/utils/extends.rb +++ b/lib/rbot/core/utils/extends.rb @@ -1,10 +1,43 @@ #-- vim:sw=2:et #++ # -# Extensions to standard classes, to be used by the various plugins +# :title: Standard classes extensions +# +# Author:: Giuseppe "Oblomov" Bilotta +# Copyright:: (C) 2006,2007 Giuseppe Bilotta +# License:: GPL v2 +# +# This file collects extensions to standard Ruby classes and to some core rbot +# classes to be used by the various plugins +# # Please note that global symbols have to be prefixed by :: because this plugin # will be read into an anonymous module +# Extensions to the Module class +# +class ::Module + + # Many plugins define Struct objects to hold their data. On rescans, lots of + # warnings are echoed because of the redefinitions. Using this method solves + # the problem, by checking if the Struct already exists, and if it has the + # same attributes + # + def define_structure(name, *members) + sym = name.to_sym + if Struct.const_defined?(sym) + kl = Struct.const_get(sym) + if kl.new.members.map { |member| member.intern } == members.map + debug "Struct #{sym} previously defined, skipping" + const_set(sym, kl) + return + end + end + debug "Defining struct #{sym} with members #{members.inspect}" + const_set(sym, Struct.new(name.to_s, *members)) + end +end + + # Extensions to the Array class # class ::Array @@ -18,45 +51,138 @@ class ::Array end end +# Extensions to the Range class +# +class ::Range + + # This method returns a random number between the lower and upper bound + # + def pick_one + len = self.last - self.first + len += 1 unless self.exclude_end? + self.first + Kernel::rand(len) + end + alias :rand :pick_one +end + +# Extensions for the Numeric classes +# +class ::Numeric + + # This method forces a real number to be not more than a given positive + # number or not less than a given positive number, or between two any given + # numbers + # + def clip(left,right=0) + raise ArgumentError unless left.kind_of?(Numeric) and right.kind_of?(Numeric) + l = [left,right].min + u = [left,right].max + return l if self < l + return u if self > u + return self + end +end + # Extensions to the String class # -# TODO make ircify_html() accept an Hash of options, and make riphtml() just -# call ircify_html() with stronger purify options. +# TODO make riphtml() just call ircify_html() with stronger purify options. # class ::String # This method will return a purified version of the receiver, with all HTML # stripped off and some of it converted to IRC formatting # - def ircify_html - txt = self + def ircify_html(opts={}) + txt = self.dup + + # remove scripts + txt.gsub!(/]*)?>.*?<\/script>/im, "") + + # remove styles + txt.gsub!(/]*)?>.*?<\/style>/im, "") # bold and strong -> bold - txt.gsub!(/<\/?(?:b|strong)\s*>/, "#{Bold}") + txt.gsub!(/<\/?(?:b|strong)(?:\s+[^>]*)?>/im, "#{Bold}") # italic, emphasis and underline -> underline - txt.gsub!(/<\/?(?:i|em|u)\s*>/, "#{Underline}") + txt.gsub!(/<\/?(?:i|em|u)(?:\s+[^>]*)?>/im, "#{Underline}") ## This would be a nice addition, but the results are horrible ## Maybe make it configurable? # txt.gsub!(/<\/?a( [^>]*)?>/, "#{Reverse}") + case val = opts[:a_href] + when Reverse, Bold, Underline + txt.gsub!(/<(?:\/a\s*|a (?:[^>]*\s+)?href\s*=\s*(?:[^>]*\s*)?)>/, val) + when :link_out + # Not good for nested links, but the best we can do without something like hpricot + txt.gsub!(/]*\s+)?href\s*=\s*(?:([^"'>][^\s>]*)\s+|"((?:[^"]|\\")*)"|'((?:[^']|\\')*)')(?:[^>]*\s+)?>(.*?)<\/a>/) { |match| + debug match + debug [$1, $2, $3, $4].inspect + link = $1 || $2 || $3 + str = $4 + str + ": " + link + } + else + warning "unknown :a_href option #{val} passed to ircify_html" if val + end - # Paragraph and br tags are converted to whitespace. - txt.gsub!(/<\/?(p|br)\s*\/?\s*>/, ' ') + # Paragraph and br tags are converted to whitespace + txt.gsub!(/<\/?(p|br)(?:\s+[^>]*)?\s*\/?\s*>/i, ' ') txt.gsub!("\n", ' ') + txt.gsub!("\r", ' ') + + # Superscripts and subscripts are turned into ^{...} and _{...} + # where the {} are omitted for single characters + txt.gsub!(/(.*?)<\/sup>/, '^{\1}') + txt.gsub!(/(.*?)<\/sub>/, '_{\1}') + txt.gsub!(/(^|_)\{(.)\}/, '\1\2') + + # List items are converted to *). We don't have special support for + # nested or ordered lists. + txt.gsub!(/
  • /, ' *) ') # All other tags are just removed txt.gsub!(/<[^>]+>/, '') + # Convert HTML entities. We do it now to be able to handle stuff + # such as   + txt = Utils.decode_html_entities(txt) + + # Keep unbreakable spaces or conver them to plain spaces? + case val = opts[:nbsp] + when :space, ' ' + txt.gsub!([160].pack('U'), ' ') + else + warning "unknown :nbsp option #{val} passed to ircify_html" if val + end + # Remove double formatting options, since they only waste bytes txt.gsub!(/#{Bold}(\s*)#{Bold}/, '\1') txt.gsub!(/#{Underline}(\s*)#{Underline}/, '\1') + # Simplify whitespace that appears on both sides of a formatting option + txt.gsub!(/\s+(#{Bold}|#{Underline})\s+/, ' \1') + txt.sub!(/\s+(#{Bold}|#{Underline})\z/, '\1') + txt.sub!(/\A(#{Bold}|#{Underline})\s+/, '\1') + # And finally whitespace is squeezed txt.gsub!(/\s+/, ' ') + txt.strip! + + if opts[:limit] && txt.size > opts[:limit] + txt = txt.slice(0, opts[:limit]) + "#{Reverse}...#{Reverse}" + end # Decode entities and strip whitespace - return Utils.decode_html_entities(txt).strip! + return txt + end + + # As above, but modify the receiver + # + def ircify_html!(opts={}) + old_hash = self.hash + replace self.ircify_html(opts) + return self unless self.hash == old_hash end # This method will strip all HTML crud from the receiver @@ -64,8 +190,26 @@ class ::String def riphtml self.gsub(/<[^>]+>/, '').gsub(/&/,'&').gsub(/"/,'"').gsub(/</,'<').gsub(/>/,'>').gsub(/&ellip;/,'...').gsub(/'/, "'").gsub("\n",'') end + + # This method tries to find an HTML title in the string, + # and returns it if found + def get_html_title + if defined? ::Hpricot + Hpricot(self).at("title").inner_html + else + return unless Irc::Utils::TITLE_REGEX.match(self) + $1 + end + end + + # This method returns the IRC-formatted version of an + # HTML title found in the string + def ircify_html_title + self.get_html_title.ircify_html rescue nil + end end + # Extensions to the Regexp class, with some common and/or complex regular # expressions. # @@ -84,23 +228,9 @@ class ::Regexp IN_ON = /in|on/ - # We start with some IRC related regular expressions, used to match - # Irc::User nicks and Irc::Channel names - # - # For each of them we define three versions of the regular expression: - # * a generic one, which should match for any server but may turn out to - # match more than a specific server would accept - # * an RFC-compliant matcher - # * TODO a server-specific one that uses the Irc::Server#supports method to build - # a matcher valid for a particular server. - # module Irc - CHAN_FIRST = /[#&+]/ - CHAN_SAFE = /![A-Z0-9]{5}/ - CHAN_ANY = /[^\x00\x07\x0A\x0D ,:]/ - GEN_CHAN = /(?:#{CHAN_FIRST}|#{CHAN_SAFE})#{CHAN_ANY}+/ - RFC_CHAN = /#{CHAN_FIRST}#{CHAN_ANY}{1,49}|#{CHAN_SAFE}#{CHAN_ANY}{1,44}/ - + # Match a list of channel anmes separated by optional commas, whitespace + # and optionally the word "and" CHAN_LIST = Regexp.new_list(GEN_CHAN) # Match "in #channel" or "on #channel" and/or "in private" (optionally @@ -115,30 +245,15 @@ class ::Regexp IN_CHAN_LIST_PVT_SFX = Regexp.new_list(/#{GEN_CHAN}|here|private|pvt/, IN_ON) IN_CHAN_LIST_PVT = /#{IN_ON}\s+#{IN_CHAN_LIST_PVT_SFX}|anywhere|everywhere/ - SPECIAL_CHAR = /[\x5b-\x60\x7b-\x7d]/ - NICK_FIRST = /#{SPECIAL_CHAR}|[[:alpha:]]/ - NICK_ANY = /#{SPECIAL_CHAR}|[[:alnum:]]|-/ - GEN_NICK = /#{NICK_FIRST}#{NICK_ANY}+/ - RFC_NICK = /#{NICK_FIRST}#{NICK_ANY}{0,8}/ - # Match a list of nicknames separated by optional commas, whitespace and # optionally the word "and" - NICK_LIST = Regexp.new_list(GEN_CHAN) + NICK_LIST = Regexp.new_list(GEN_NICK) end - # Next, some general purpose ones - DIGITS = /\d+/ - HEX_DIGIT = /[0-9A-Fa-f]/ - HEX_DIGITS = /#{HEX_DIGIT}+/ - HEX_OCTET = /#{HEX_DIGIT}#{HEX_DIGIT}?/ - DEC_OCTET = /[01]?\d?\d|2[0-4]\d|25[0-5]/ - DEC_IP = /#{DEC_OCTET}.#{DEC_OCTET}.#{DEC_OCTET}/ - HEX_IP = /#{HEX_OCTET}.#{HEX_OCTET}.#{HEX_OCTET}/ - IP = /#{DEC_IP}|#{HEX_IP}/ - end + module ::Irc