Re: another look at namespaces
- Posted by Robert Craig <rds at RapidEuphoria.com> Jul 11, 2003
- 479 views
Thanks to everyone who helped develop the recent proposed namespace refinement. (I won't name names since it's a bit fuzzy who came up with what idea when). I think it makes sense that when there are conflicting global symbols, and one of the symbols was intentionally defined or included (directly or indirectly) by the current file, while the other symbol(s) are visible for reasons outside the control of the current file, then it's pretty clear which symbol the current file wants to use. It's already the case that if a symbol is directly defined in the current file, that symbol will be used, and there will be no conflict. We'll extend that idea to symbols included by the current file. I'll implement this refinement to the namespace feature for the next release. Matt, I'll be interested in seeing your code, but keep in mind, for me there are many places affected by any change to symbol look-up. All are coded slightly differently. - main parser/scanner look-up for Interpreter and Translator - Interpreter routine_id() look-up - interactive trace(1) variable look-up - binder look-up - Translator routine-id() look-up The last 3 are not part of the Source Code product. Regards, Rob Craig Rapid Deployment Software http://www.RapidEuphoria.com