Merge branch 'master' into lua-scanner

Conflicts:
	lib/coderay/token_kinds.rb

12 files changed, 320 insertions, 36 deletions

diff --git a/lib/coderay/scanners/c.rb b/lib/coderay/scanners/c.rb
index 8d24b99..84b6e8e 100644
--- a/lib/coderay/scanners/c.rb
+++ b/lib/coderay/scanners/c.rb
@@ -148,7 +148,7 @@ module Scanners
             encoder.text_token match, :char
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group :string
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
             state = :initial
             label_expected = false
           else
diff --git a/lib/coderay/scanners/cpp.rb b/lib/coderay/scanners/cpp.rb
index 9da62f4..e61f56f 100644
--- a/lib/coderay/scanners/cpp.rb
+++ b/lib/coderay/scanners/cpp.rb
@@ -160,7 +160,7 @@ module Scanners
             encoder.text_token match, :char
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group :string
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
             state = :initial
             label_expected = false
           else
diff --git a/lib/coderay/scanners/css.rb b/lib/coderay/scanners/css.rb
index 7b731ef..732f9c5 100644
--- a/lib/coderay/scanners/css.rb
+++ b/lib/coderay/scanners/css.rb
@@ -7,27 +7,25 @@ module Scanners
     
     KINDS_NOT_LOC = [
       :comment,
-      :class, :pseudo_class, :type,
-      :constant, :directive,
+      :class, :pseudo_class, :tag,
+      :id, :directive,
       :key, :value, :operator, :color, :float, :string,
-      :error, :important,
+      :error, :important, :type,
     ]  # :nodoc:
     
     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
@@ -67,13 +64,13 @@ module Scanners
         elsif case states.last
           when :initial, :media
             if match = scan(/(?>#{RE::Ident})(?!\()|\*/ox)
-              encoder.text_token match, :type
+              encoder.text_token match, :tag
               next
             elsif match = scan(RE::Class)
               encoder.text_token match, :class
               next
             elsif match = scan(RE::Id)
-              encoder.text_token match, :constant
+              encoder.text_token match, :id
               next
             elsif match = scan(RE::PseudoClass)
               encoder.text_token match, :pseudo_class
@@ -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/diff.rb b/lib/coderay/scanners/diff.rb
index 38efaf4..af0f755 100644
--- a/lib/coderay/scanners/diff.rb
+++ b/lib/coderay/scanners/diff.rb
@@ -45,7 +45,7 @@ module Scanners
           if match = scan(/--- |\+\+\+ |=+|_+/)
             encoder.begin_line line_kind = :head
             encoder.text_token match, :head
-            if match = scan(/.*?(?=$|[\t\n\x00]|  \(revision)/)
+            if match = scan(/[^\x00\n]+?(?=$|[\t\n]|  \(revision)/)
               encoder.text_token match, :filename
               if options[:highlight_code] && match != '/dev/null'
                 file_type = CodeRay::FileType.fetch(match, :text)
diff --git a/lib/coderay/scanners/java.rb b/lib/coderay/scanners/java.rb
index c1490ac..b282864 100644
--- a/lib/coderay/scanners/java.rb
+++ b/lib/coderay/scanners/java.rb
@@ -147,7 +147,7 @@ module Scanners
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group state
             state = :initial
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
           else
             raise_inspect "else case \" reached; %p not handled." % peek(1), encoder
           end
diff --git a/lib/coderay/scanners/java_script.rb b/lib/coderay/scanners/java_script.rb
index 43ecb18..9eb0a0a 100644
--- a/lib/coderay/scanners/java_script.rb
+++ b/lib/coderay/scanners/java_script.rb
@@ -54,10 +54,17 @@ module Scanners
     
   protected
     
+    def setup
+      @state = :initial
+    end
+    
     def scan_tokens encoder, options
       
-      state = :initial
-      string_delimiter = nil
+      state, string_delimiter = options[:state] || @state
+      if string_delimiter
+        encoder.begin_group state
+      end
+      
       value_expected = true
       key_expected = false
       function_expected = false
@@ -72,9 +79,10 @@ module Scanners
             value_expected = true if !value_expected && match.index(?\n)
             encoder.text_token match, :space
             
-          elsif match = scan(%r! // [^\n\\]* (?: \\. [^\n\\]* )* | /\* (?: .*? \*/ | .* ) !mx)
+          elsif match = scan(%r! // [^\n\\]* (?: \\. [^\n\\]* )* | /\* (?: .*? \*/ | .*() ) !mx)
             value_expected = true
             encoder.text_token match, :comment
+            state = :open_multi_line_comment if self[1]
             
           elsif check(/\.?\d/)
             key_expected = value_expected = false
@@ -175,20 +183,36 @@ module Scanners
             encoder.text_token match, :content
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group state
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
+            string_delimiter = nil
             key_expected = value_expected = false
             state = :initial
           else
-            raise_inspect "else case \" reached; %p not handled." % peek(1), encoder
+            raise_inspect "else case #{string_delimiter} reached; %p not handled." % peek(1), encoder
           end
           
+        when :open_multi_line_comment
+          if match = scan(%r! .*? \*/ !mx)
+            state = :initial
+          else
+            match = scan(%r! .+ !mx)
+          end
+          value_expected = true
+          encoder.text_token match, :comment if match
+          
         else
-          raise_inspect 'Unknown state', encoder
+          #:nocov:
+          raise_inspect 'Unknown state: %p' % [state], encoder
+          #:nocov:
           
         end
         
       end
       
+      if options[:keep_state]
+        @state = state, string_delimiter
+      end
+      
       if [:string, :regexp].include? state
         encoder.end_group state
       end
diff --git a/lib/coderay/scanners/json.rb b/lib/coderay/scanners/json.rb
index 0c90c34..4e0f462 100644
--- a/lib/coderay/scanners/json.rb
+++ b/lib/coderay/scanners/json.rb
@@ -70,7 +70,7 @@ module Scanners
             encoder.text_token match, :content
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group state
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
             state = :initial
           else
             raise_inspect "else case \" reached; %p not handled." % peek(1), encoder
diff --git a/lib/coderay/scanners/python.rb b/lib/coderay/scanners/python.rb
index cbdbbdb..a9492ab 100644
--- a/lib/coderay/scanners/python.rb
+++ b/lib/coderay/scanners/python.rb
@@ -133,7 +133,7 @@ module Scanners
           elsif match = scan(/ \\ | $ /x)
             encoder.end_group string_type
             string_type = nil
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
             state = :initial
           else
             raise_inspect "else case \" reached; %p not handled." % peek(1), encoder, state
diff --git a/lib/coderay/scanners/ruby.rb b/lib/coderay/scanners/ruby.rb
index c5cf1e2..c282f31 100644
--- a/lib/coderay/scanners/ruby.rb
+++ b/lib/coderay/scanners/ruby.rb
@@ -96,7 +96,7 @@ module Scanners
                                       /#{patterns::METHOD_NAME}/o)
               
               kind = patterns::IDENT_KIND[match]
-              if kind == :ident && value_expected != :colon_expected && scan(/:(?!:)/)
+              if value_expected != :colon_expected && scan(/:(?!:)/)
                 value_expected = true
                 encoder.text_token match, :key
                 encoder.text_token ':',   :operator
diff --git a/lib/coderay/scanners/sass.rb b/lib/coderay/scanners/sass.rb
new file mode 100644
index 0000000..167051d
--- /dev/null
+++ b/lib/coderay/scanners/sass.rb
@@ -0,0 +1,227 @@
+module CodeRay
+module Scanners
+  
+  # A scanner for Sass.
+  class Sass < CSS
+    
+    register_for :sass
+    file_extension 'sass'
+    
+    STRING_CONTENT_PATTERN = {
+      "'" => /(?:[^\n\'\#]+|\\\n|#{RE::Escape}|#(?!\{))+/,
+      '"' => /(?:[^\n\"\#]+|\\\n|#{RE::Escape}|#(?!\{))+/,
+    }
+    
+  protected
+    
+    def setup
+      @state = :initial
+    end
+    
+    def scan_tokens encoder, options
+      states = Array(options[:state] || @state)
+      string_delimiter = nil
+      
+      until eos?
+        
+        if bol? && (match = scan(/(?>( +)?(\/[\*\/])(.+)?)(?=\n)/))
+          encoder.text_token self[1], :space if self[1]
+          encoder.begin_group :comment
+          encoder.text_token self[2], :delimiter
+          encoder.text_token self[3], :content if self[3]
+          if match = scan(/(?:\n+#{self[1]} .*)+/)
+            encoder.text_token match, :content
+          end
+          encoder.end_group :comment
+        elsif match = scan(/\n|[^\n\S]+\n?/)
+          encoder.text_token match, :space
+          if match.index(/\n/)
+            value_expected = false
+            states.pop if states.last == :include
+          end
+        
+        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 : :tag)
+              next
+            elsif !value_expected && (match = scan(/\*/))
+              encoder.text_token match, :tag
+              next
+            elsif match = scan(RE::Class)
+              encoder.text_token match, :class
+              next
+            elsif match = scan(RE::Id)
+              encoder.text_token match, :id
+              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(/@import\b/)
+              encoder.text_token match, :directive
+              states << :include
+              next
+            elsif match = scan(/@media\b/)
+              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 :string
+              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
+          
+          when :include
+            if match = scan(/[^\s'",]+/)
+              encoder.text_token match, :include
+              next
+            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
+          if states.include? :sass_inline
+            content = scan_until(/(?=#{string_delimiter}|\}|\z)/)
+            encoder.text_token content, :content unless content.empty?
+            encoder.text_token string_delimiter, :delimiter if scan(/#{string_delimiter}/)
+            encoder.end_group :string
+          else
+            states.push :string
+          end
+          
+        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(/[a-z][-a-z_]*(?=\()/o)
+          encoder.text_token match, :predefined
+          
+        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
+      end
+      
+      encoder
+    end
+    
+  end
+  
+end
+end
diff --git a/lib/coderay/scanners/sql.rb b/lib/coderay/scanners/sql.rb
index bcbffd5..b757278 100644
--- a/lib/coderay/scanners/sql.rb
+++ b/lib/coderay/scanners/sql.rb
@@ -148,7 +148,7 @@ module CodeRay module Scanners
               encoder.text_token string_content, :content
               string_content = ''
             end
-            encoder.text_token match, :error
+            encoder.text_token match, :error unless match.empty?
             state = :initial
           else
             raise "else case \" reached; %p not handled." % peek(1), encoder
diff --git a/lib/coderay/scanners/taskpaper.rb b/lib/coderay/scanners/taskpaper.rb
new file mode 100644
index 0000000..42670bc
--- /dev/null
+++ b/lib/coderay/scanners/taskpaper.rb
@@ -0,0 +1,36 @@
+module CodeRay
+module Scanners
+  
+  class Taskpaper < Scanner
+    
+    register_for :taskpaper
+    file_extension 'taskpaper'
+    
+  protected
+    
+    def scan_tokens encoder, options
+      until eos?
+        if match = scan(/\S.*:.*$/)                  # project
+          encoder.text_token(match, :namespace)
+        elsif match = scan(/-.+@done.*/)             # completed task
+          encoder.text_token(match, :done)
+        elsif match = scan(/-(?:[^@\n]+|@(?!due))*/) # task
+          encoder.text_token(match, :plain)
+        elsif match = scan(/@due.*/)                 # comment
+          encoder.text_token(match, :important)
+        elsif match = scan(/.+/)                     # comment
+          encoder.text_token(match, :comment)
+        elsif match = scan(/\s+/)                    # space
+          encoder.text_token(match, :space)
+        else                                         # other
+          encoder.text_token getch, :error
+        end
+      end
+      
+      encoder
+    end
+    
+  end
+  
+end
+end