Harbs
Harbs
Oops. Hit the close button by mistake...
Well, it seems like that was only partially the issue. I created a branch (`feature/ROYALE_INTERFACE_INFO`) which moves the interface data out into a separate variable and Language uses that. That...
One more thing that we probably need to do to prevent rename issues is to make sure the compiler doesn't add quotes to the property names: `ROYALE_CLASS_INFO`, `ROYALE_INTERFACE_INFO` and `ROYALE_INTERFACE_INFO`.
I added exceptions here: https://github.com/apache/royale-compiler/commit/6c1bd52f8942bff982b09ffcbc7e33dffeb51dfd I'd appreciate if someone who's more familiar with the compiler code looks at what I did. I don't know if there's a better way to...
@pashminakazi Do you have a test case for this, or should we close the issue?
I think the bug was in Language.string. I just committed a fix for that.
It looks like the compiler has a bug. It adds the Language.string call even when using an as cast in a ternary expression. I made a temporary change to Language.string...
Here's a nice summary of the different methods of converting to string: Performance-wise toString() seems to be the poorest performer and string concatenation seems to be the best: https://jsperf.com/cast-to-string/2 Possibly,...
I think we should return `value` rather than `null` to preserve `undefined` if it's `undefined` instead of `null`.
Makes sense. FYI, I just ran that jsperf across Chrome, Firefox and Safari. String concatenation is consistently fastest across Mac browsers. (I have not checked Windows.) String(val) vs. val.toString() varies...