Large update: Scanners for HTML, RHTML and Nitro-XHTML added.

CSS style changes/enhancments (mainly the new background color for inline code, affects all Ruby code.)
Demos and tests adjusted.

Plugin: new PluginHost::default method.

Scanner:
- New setup method
- ability to re-use a scanner
- ability to keep the tokens
- minor changes to token caching and string flattening

Encoder: Error if token content is neither String nor Symbol.

HTML encoder:
- more warnings for unclosed tokens
- output now UTF-8

Ruby Scanner:
- bug: symbols before => now do not include =; {:foo=>bar} is valid Ruby code
- try to close all open tokens
- constants now all with specific namespace (for speed, I hope)

Styles: new :entity/en class.

Test suite now gives hinted HTML output.

1 files changed, 4 insertions, 2 deletions

diff --git a/lib/coderay/encoder.rb b/lib/coderay/encoder.rb
index 8b6c22a..8f8375a 100644
--- a/lib/coderay/encoder.rb
+++ b/lib/coderay/encoder.rb
@@ -130,10 +130,12 @@ module CodeRay
 			# By default, it calls text_token or block_token, depending on

 			# whether +text+ is a String.

 			def token text, kind

-				if text.is_a? String

+				if text.is_a? ::String

 					text_token text, kind

-				else

+				elsif text.is_a? ::Symbol

 					block_token text, kind

+				else

+					raise 'Unknown token text type: %p' % text

 				end

 			end