The attached patch changes the support.class.ruby regex to match
variable.other.constant.ruby, since class names are in fact constants.
Specifically, it removes the requirement that the second letter be a lowercase
letter. Prototype[1], for instance, has an IEBrowser class in its jstest.rb
file.
This also allows class names to be one character long while they were formerly
at least two. Is that a problem for anything else in the grammar?
[1] http://prototypejs.org/
Index: Ruby.tmbundle/Syntaxes/Ruby.plist
===================================================================
--- Ruby.tmbundle/Syntaxes/Ruby.plist (revision 7206)
+++ Ruby.tmbundle/Syntaxes/Ruby.plist (working copy)
@@ -336,7 +336,7 @@
</dict>
<dict>
<key>match</key>
- <string>\b[A-Z][a-z]\w*(?=((\.|::)[A-Za-z]|\[))</string>
+ <string>\b[A-Z]\w*(?=((\.|::)[A-Za-z]|\[))</string>
<key>name</key>
<string>support.class.ruby</string>
</dict>
_______________________________________________
textmate-dev mailing list
[email protected]
http://lists.macromates.com/mailman/listinfo/textmate-dev
