New: *CSS Scanner* (closes #29).

* Based on Stefan Walk's implementation, with fixes, enhancements and speedups.
* It seems to be fairly fast and highlights nicely.
* I added the styles for Ignis Draconis, S5, and YUI as example code for testing.

More changes:
* Added three new token classes, :important, :pseudo_class, and :value, along with CSS styles.

3 files changed, 188 insertions, 1 deletions

diff --git a/lib/coderay/scanners/css.rb b/lib/coderay/scanners/css.rb
new file mode 100644
index 0000000..897e27a
--- /dev/null
+++ b/lib/coderay/scanners/css.rb
@@ -0,0 +1,181 @@
+module CodeRay
+module Scanners
+
+  class CSS < Scanner
+
+    register_for :css
+
+    module RE
+      NonASCII = /[\x80-\xFF]/
+      Hex = /[0-9a-fA-F]/
+      Unicode = /\\#{Hex}{1,6}(?:\r\n|\s)?/ # differs from standard because it allows uppercase hex too
+      Escape = /#{Unicode}|\\[^\r\n\f0-9a-fA-F]/
+      NMChar = /[-_a-zA-Z0-9]|#{NonASCII}|#{Escape}/
+      NMStart = /[_a-zA-Z]|#{NonASCII}|#{Escape}/
+      NL = /\r\n|\r|\n|\f/
+      String1 = /"(?:[^\n\r\f\\"]|\\#{NL}|#{Escape})*"?/  # FIXME: buggy regexp
+      String2 = /'(?:[^\n\r\f\\']|\\#{NL}|#{Escape})*'?/  # FIXME: buggy regexp
+      String = /#{String1}|#{String2}/
+
+      HexColor = /#(?:#{Hex}{6}|#{Hex}{3})/
+      Color = /#{HexColor}/
+
+      Num = /-?(?:[0-9]+|[0-9]*\.[0-9]+)/
+      Name = /#{NMChar}+/
+      Ident = /-?#{NMStart}#{NMChar}*/
+      AtKeyword = /@#{Ident}/
+      Percentage = /#{Num}%/
+
+      reldimensions = %w[em ex px]
+      absdimensions = %w[in cm mm pt pc]
+      Unit = Regexp.union(*(reldimensions + absdimensions))
+
+      Dimension = /#{Num}#{Unit}/
+
+      Comment = %r! /\* (?: .*? \*/ | .* ) !mx
+      Function = /(?:url|alpha)\((?:[^)\n\r\f]|\\\))*\)?/
+
+      Id = /##{Name}/
+      Class = /\.#{Name}/
+      PseudoClass = /:#{Name}/
+
+    end
+
+    def scan_tokens tokens, options
+      
+      value_expected = nil
+      states = [:initial]
+
+      until eos?
+
+        kind = nil
+        match = nil
+
+        if scan(/\s+/)
+          kind = :space
+
+        elsif case states.last
+          when :initial
+            if scan(/#{RE::Ident}|\*/ox)
+              kind = :keyword
+            elsif scan RE::Class
+              kind = :class
+            elsif scan RE::Id
+              kind = :constant
+            elsif scan RE::PseudoClass
+              kind = :pseudo_class
+            elsif scan RE::Name
+              kind = :identifier
+            end
+
+          when :block
+            if scan(/(?>#{RE::Ident})(?!\()/ox)
+              if value_expected
+                kind = :value
+              else
+                kind = :key
+              end
+            end
+
+          when :comment
+            if scan(/(?:[^*\s]|\*(?!\/))+/)
+              kind = :comment
+            elsif scan(/\*\//)
+              kind = :comment
+              states.pop
+            elsif scan(/\s+/)
+              kind = :space
+            end
+
+          else
+            raise_inspect 'Unknown state', tokens
+
+          end
+
+        elsif scan(/\/\*/)
+          kind = :comment
+          states.push :comment
+
+        elsif scan(/\{/)
+          value_expected = false
+          kind = :operator
+          states.push :block
+
+        elsif scan(/\}/)
+          value_expected = false
+          if states.last == :block
+            kind = :operator
+            states.pop
+          else
+            kind = :error
+          end
+
+        elsif match = scan(/#{RE::String}/o)
+          tokens << [:open, :string]
+          tokens << [match[0, 1], :delimiter]
+          tokens << [match[1..-2], :content] if match.size > 2
+          tokens << [match[-1, 1], :delimiter] if match.size >= 2
+          tokens << [:close, :string]
+          next
+
+        elsif match = scan(/#{RE::Function}/o)
+          tokens << [:open, :string]
+          start = match[/^\w+\(/]
+          tokens << [start, :delimiter]
+          if match[-1] == ?)
+            tokens << [match[start.size..-2], :content]
+            tokens << [')', :delimiter]
+          else
+            tokens << [match[start.size..-1], :content]
+          end
+          tokens << [:close, :string]
+          next
+
+        elsif scan(/(?: #{RE::Dimension} | #{RE::Percentage} | #{RE::Num} )/ox)
+          kind = :float
+
+        elsif scan(/#{RE::Color}/o)
+          kind = :color
+
+        elsif scan(/! *important/)
+          kind = :important
+
+        elsif scan(/rgb\([^()\n]*\)?/)
+          kind = :color
+
+        elsif scan(/#{RE::AtKeyword}/o)
+          kind = :directive
+
+        elsif match = scan(/ [+>:;,.=()\/] /x)
+          if match == ':'
+            value_expected = true
+          elsif match == ';'
+            value_expected = false
+          end
+          kind = :operator
+
+        else
+          getch
+          kind = :error
+
+        end
+
+        match ||= matched
+        if $DEBUG and not kind
+          raise_inspect 'Error token %p in line %d' %
+            [[match, kind], line], tokens
+        end
+        raise_inspect 'Empty token', tokens unless match
+
+        tokens << [match, kind]
+        # tokens << [states.inspect, :error]
+
+      end
+
+      tokens
+    end
+
+  end
+
+end
+end
diff --git a/lib/coderay/styles/cycnus.rb b/lib/coderay/styles/cycnus.rb
index 7b68a66..d552d50 100644
--- a/lib/coderay/styles/cycnus.rb
+++ b/lib/coderay/styles/cycnus.rb
@@ -80,6 +80,7 @@ ol.CodeRay li { white-space: pre }
 .il .il .il { background: #ccc }
 .il .idl { font-weight: bold; color: #888 }
 
+.im { color:#f00; }
 .in { color:#B2B; font-weight:bold }
 .iv { color:#33B }
 .la { color:#970; font-weight:bold }
@@ -89,11 +90,13 @@ ol.CodeRay li { white-space: pre }
 .op { }
 .pc { color:#038; font-weight:bold }
 .pd { color:#369; font-weight:bold }
-.pp { color:#579 }
+.pp { color:#579; }
+.ps { color:#00C; font-weight: bold; }
 .pt { color:#339; font-weight:bold }
 .r, .kw  { color:#080; font-weight:bold }
 
 .ke { color: #808; }
+.vl { color: #088; }
 
 .rx { background-color:#fff0ff }
 .rx .k { color:#808 }
diff --git a/lib/coderay/token_classes.rb b/lib/coderay/token_classes.rb
index c352e48..9b5e83c 100755
--- a/lib/coderay/token_classes.rb
+++ b/lib/coderay/token_classes.rb
@@ -29,6 +29,7 @@ module CodeRay
       :function => 'fu',
       :global_variable => 'gv',
       :hex => 'hx',
+      :important => 'im',
       :include => 'ic',
       :inline => 'il',
       :inline_delimiter => 'idl',
@@ -46,6 +47,7 @@ module CodeRay
       :pre_type => 'pt',
       :predefined => 'pd',
       :preprocessor => 'pp',
+      :pseudo_class => 'ps',
       :regexp => 'rx',
       :reserved => 'r',
       :shell => 'sh',
@@ -56,6 +58,7 @@ module CodeRay
       :tag_special => 'ts',
       :type => 'ty',
       :variable => 'v',
+      :value => 'vl',
       :xml_text => 'xt',
 
       :ident => :NO_HIGHLIGHT, # 'id'