added Sass scanner (work in progress)

4 files changed, 220 insertions, 17 deletions

diff --git a/lib/coderay/helpers/file_type.rb b/lib/coderay/helpers/file_type.rb
index 637001b..b06b9bc 100644
--- a/lib/coderay/helpers/file_type.rb
+++ b/lib/coderay/helpers/file_type.rb
@@ -116,6 +116,7 @@ module CodeRay
       'rpdf'     => :ruby,
       'ru'       => :ruby,
       'rxml'     => :ruby,
+      'sass'     => :sass,
       # 'sch'      => :scheme,
       'sql'      => :sql,
       # 'ss'       => :scheme,
diff --git a/lib/coderay/scanners/css.rb b/lib/coderay/scanners/css.rb
index 7b731ef..003eed6 100644
--- a/lib/coderay/scanners/css.rb
+++ b/lib/coderay/scanners/css.rb
@@ -15,19 +15,17 @@ module Scanners
     
     module RE  # :nodoc:
       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]|#{Escape}/
-      NMStart = /[_a-zA-Z]|#{Escape}/
-      NL = /\r\n|\r|\n|\f/
-      String1 = /"(?:[^\n\r\f\\"]|\\#{NL}|#{Escape})*"?/  # TODO: buggy regexp
-      String2 = /'(?:[^\n\r\f\\']|\\#{NL}|#{Escape})*'?/  # TODO: buggy regexp
+      Unicode = /\\#{Hex}{1,6}\b/ # differs from standard because it allows uppercase hex too
+      Escape = /#{Unicode}|\\[^\n0-9a-fA-F]/
+      NMChar = /[-_a-zA-Z0-9]/
+      NMStart = /[_a-zA-Z]/
+      String1 = /"(?:[^\n\\"]+|\\\n|#{Escape})*"?/  # TODO: buggy regexp
+      String2 = /'(?:[^\n\\']+|\\\n|#{Escape})*'?/  # TODO: buggy regexp
       String = /#{String1}|#{String2}/
       
       HexColor = /#(?:#{Hex}{6}|#{Hex}{3})/
-      Color = /#{HexColor}/
       
-      Num = /-?(?:[0-9]+|[0-9]*\.[0-9]+)/
+      Num = /-?(?:[0-9]*\.[0-9]+|[0-9]+)/
       Name = /#{NMChar}+/
       Ident = /-?#{NMStart}#{NMChar}*/
       AtKeyword = /@#{Ident}/
@@ -35,16 +33,15 @@ module Scanners
       
       reldimensions = %w[em ex px]
       absdimensions = %w[in cm mm pt pc]
-      Unit = Regexp.union(*(reldimensions + absdimensions + %w[s]))
+      Unit = Regexp.union(*(reldimensions + absdimensions + %w[s dpi dppx deg]))
       
       Dimension = /#{Num}#{Unit}/
       
-      Comment = %r! /\* (?: .*? \*/ | .* ) !mx
-      Function = /(?:url|alpha|attr|counters?)\((?:[^)\n\r\f]|\\\))*\)?/
+      Function = /(?:url|alpha|attr|counters?)\((?:[^)\n]|\\\))*\)?/
       
-      Id = /##{Name}/
+      Id = /(?!#{HexColor}\b(?!-))##{Name}/
       Class = /\.#{Name}/
-      PseudoClass = /:#{Name}/
+      PseudoClass = /::?#{Ident}/
       AttributeSelector = /\[[^\]]*\]?/
     end
     
@@ -52,7 +49,7 @@ module Scanners
     
     def setup
       @state = :initial
-      @value_expected = nil
+      @value_expected = false
     end
     
     def scan_tokens encoder, options
@@ -158,7 +155,7 @@ module Scanners
         elsif match = scan(/(?: #{RE::Dimension} | #{RE::Percentage} | #{RE::Num} )/ox)
           encoder.text_token match, :float
           
-        elsif match = scan(/#{RE::Color}/o)
+        elsif match = scan(/#{RE::HexColor}/o)
           encoder.text_token match, :color
           
         elsif match = scan(/! *important/)
@@ -170,7 +167,7 @@ module Scanners
         elsif match = scan(RE::AtKeyword)
           encoder.text_token match, :directive
           
-        elsif match = scan(/ [+>:;,.=()\/] /x)
+        elsif match = scan(/ [+>~:;,.=()\/] /x)
           if match == ':'
             value_expected = true
           elsif match == ';'
diff --git a/lib/coderay/scanners/sass.rb b/lib/coderay/scanners/sass.rb
new file mode 100644
index 0000000..89d7737
--- /dev/null
+++ b/lib/coderay/scanners/sass.rb
@@ -0,0 +1,203 @@
+module CodeRay
+module Scanners
+  
+  # A scanner for Sass.
+  class Sass < CSS
+    
+    register_for :sass
+    file_extension 'sass'
+    
+    SASS_FUNCTION = /(?:inline-image|linear-gradient|color-stops|mix|lighten|darken|rotate|image-url|image-width|image-height|sprite|sprite-url|sprite-path|sprite-file|sprite-map|sprite-position|unquote|join|round|ceil|floor|nth)/
+    
+    STRING_CONTENT_PATTERN = {
+      "'" => /(?:[^\n\'\#]+|\\\n|#{RE::Escape}|#(?!\{))+/,
+      '"' => /(?:[^\n\"\#]+|\\\n|#{RE::Escape}|#(?!\{))+/,
+    }
+    
+  protected
+    
+    def setup
+      @state = :initial
+      @value_expected = false
+    end
+    
+    def scan_tokens encoder, options
+      states = Array(options[:state] || @state)
+      string_delimiter = nil
+      value_expected = @value_expected
+      
+      until eos?
+        
+        if match = scan(/\s+/)
+          encoder.text_token match, :space
+          value_expected = false if match.index(/\n/)
+          
+        elsif states.last == :sass_inline && (match = scan(/\}/))
+          encoder.text_token match, :inline_delimiter
+          encoder.end_group :inline
+          states.pop
+        
+        elsif case states.last
+          when :initial, :media, :sass_inline
+            if match = scan(/(?>#{RE::Ident})(?!\()/ox)
+              encoder.text_token match, value_expected ? :value : (check(/.*:/) ? :key : :type)
+              next
+            elsif !value_expected && (match = scan(/\*/))
+              encoder.text_token match, :type
+              next
+            elsif match = scan(RE::Class)
+              encoder.text_token match, :class
+              next
+            elsif match = scan(RE::Id)
+              encoder.text_token match, :constant
+              next
+            elsif match = scan(RE::PseudoClass)
+              encoder.text_token match, :pseudo_class
+              next
+            elsif match = scan(RE::AttributeSelector)
+              # TODO: Improve highlighting inside of attribute selectors.
+              encoder.text_token match[0,1], :operator
+              encoder.text_token match[1..-2], :attribute_name if match.size > 2
+              encoder.text_token match[-1,1], :operator if match[-1] == ?]
+              next
+            elsif match = scan(/(\=|@mixin +)#{RE::Ident}/o)
+              encoder.text_token match, :function
+              next
+            elsif match = scan(/@media/)
+              encoder.text_token match, :directive
+              # states.push :media_before_name
+              next
+            end
+          
+          when :block
+            if match = scan(/(?>#{RE::Ident})(?!\()/ox)
+              if value_expected
+                encoder.text_token match, :value
+              else
+                encoder.text_token match, :key
+              end
+              next
+            end
+            
+          when :string
+            if match = scan(STRING_CONTENT_PATTERN[string_delimiter])
+              encoder.text_token match, :content
+            elsif match = scan(/['"]/)
+              encoder.text_token match, :delimiter
+              encoder.end_group :string
+              string_delimiter = nil
+              states.pop
+            elsif match = scan(/#\{/)
+              encoder.begin_group :inline
+              encoder.text_token match, :inline_delimiter
+              states.push :sass_inline
+            elsif match = scan(/ \\ | $ /x)
+              encoder.end_group state
+              encoder.text_token match, :error unless match.empty?
+              states.pop
+            else
+              raise_inspect "else case #{string_delimiter} reached; %p not handled." % peek(1), encoder
+            end
+          
+          else
+            #:nocov:
+            raise_inspect 'Unknown state', encoder
+            #:nocov:
+            
+          end
+          
+        elsif match = scan(/\$#{RE::Ident}/o)
+          encoder.text_token match, :variable
+          next
+        
+        elsif match = scan(/&/)
+          encoder.text_token match, :local_variable
+          
+        elsif match = scan(/\+#{RE::Ident}/o)
+          encoder.text_token match, :include
+          value_expected = true
+          
+        elsif match = scan(/\/\*(?:.*?\*\/|.*)|\/\/.*/)
+          encoder.text_token match, :comment
+          
+        elsif match = scan(/#\{/)
+          encoder.begin_group :inline
+          encoder.text_token match, :inline_delimiter
+          states.push :sass_inline
+          
+        elsif match = scan(/\{/)
+          value_expected = false
+          encoder.text_token match, :operator
+          states.push :block
+          
+        elsif match = scan(/\}/)
+          value_expected = false
+          encoder.text_token match, :operator
+          if states.last == :block || states.last == :media
+            states.pop
+          end
+          
+        elsif match = scan(/['"]/)
+          encoder.begin_group :string
+          string_delimiter = match
+          encoder.text_token match, :delimiter
+          states.push :string
+          
+        elsif match = scan(/#{SASS_FUNCTION}/o)
+          encoder.text_token match, :predefined
+          
+        elsif match = scan(/#{RE::Function}/o)
+          encoder.begin_group :function
+          start = match[/^[-\w]+\(/]
+          encoder.text_token start, :delimiter
+          if match[-1] == ?)
+            encoder.text_token match[start.size..-2], :content
+            encoder.text_token ')', :delimiter
+          else
+            encoder.text_token match[start.size..-1], :content
+          end
+          encoder.end_group :function
+          
+        elsif match = scan(/(?: #{RE::Dimension} | #{RE::Percentage} | #{RE::Num} )/ox)
+          encoder.text_token match, :float
+          
+        elsif match = scan(/#{RE::HexColor}/o)
+          encoder.text_token match, :color
+          
+        elsif match = scan(/! *(?:important|optional)/)
+          encoder.text_token match, :important
+          
+        elsif match = scan(/(?:rgb|hsl)a?\([^()\n]*\)?/)
+          encoder.text_token match, :color
+          
+        elsif match = scan(/@else if\b|#{RE::AtKeyword}/)
+          encoder.text_token match, :directive
+          value_expected = true
+          
+        elsif match = scan(/ == | != | [-+*\/>~:;,.=()] /x)
+          if match == ':'
+            value_expected = true
+          elsif match == ';'
+            value_expected = false
+          end
+          encoder.text_token match, :operator
+          
+        else
+          encoder.text_token getch, :error
+          
+        end
+        
+      end
+      
+      if options[:keep_state]
+        @state = states
+        @value_expected = value_expected
+      end
+      
+      encoder
+    end
+    
+  end
+  
+end
+end
diff --git a/lib/coderay/styles/alpha.rb b/lib/coderay/styles/alpha.rb
index 8506d10..1f073b6 100644
--- a/lib/coderay/styles/alpha.rb
+++ b/lib/coderay/styles/alpha.rb
@@ -78,6 +78,7 @@ table.CodeRay td { padding: 2px 4px; vertical-align: top; }
 .exception { color:#C00; font-weight:bold }
 .float { color:#60E }
 .function { color:#06B; font-weight:bold }
+.function .delimiter { color:#024; font-weight:bold }
 .global-variable { color:#d70 }
 .hex { color:#02b }
 .imaginary { color:#f00 }
@@ -86,6 +87,7 @@ table.CodeRay td { padding: 2px 4px; vertical-align: top; }
 .inline-delimiter { font-weight: bold; color: #666 }
 .instance-variable { color:#33B }
 .integer  { color:#00D }
+.important { color:#D00 }
 .key .char { color: #60f }
 .key .delimiter { color: #404 }
 .key { color: #606 }