Bugzilla – Full Text Bug Listing
| Summary: | Errors generating when using @EGLProperty | ||
|---|---|---|---|
| Product: | z_Archived | Reporter: | Justin Spadea <jspadea> |
| Component: | EDT | Assignee: | Project Inbox <edt.javascriptgen-inbox> |
| Status: | CLOSED FIXED | QA Contact: | |
| Severity: | normal | ||
| Priority: | P3 | CC: | greer |
| Version: | unspecified | ||
| Target Milestone: | --- | ||
| Hardware: | PC | ||
| OS: | Linux | ||
| See Also: | https://bugs.eclipse.org/bugs/show_bug.cgi?id=355906 | ||
| Whiteboard: | |||
I took a look at the code, and I think you can ignore my last remark - it does make sure the getter/setter is not the current method. Another problem: looks like it defaults to "get<varName>" if you don't specify getMethod, but it doesn't check if such a method exists:
i int{@EGLProperty{}};
function test() returns(int)
return(i);
end
Generated return stmt in test(): return this.getI();
But there is no getI() method.
Justin, I believe this has been fixed under https://bugs.eclipse.org/bugs/show_bug.cgi?id=355906 With those changes, your testcase appears to gen correctly. The other problem you added, that of verifying that a getter/setter actually exists, would be a validation error. This is something Paul would have to look into, so you should probably split that off into a separate bug.... Verified |
In order to reproduce any of this, you need to change the value of Constants.Annotation_EGLProperty to "eglx.lang.EGLProperty". handler rui1 type RUIhandler {initialUI = [],onConstructionFunction = start, title="rui1"} i int{@EGLProperty{getMethod = geti}}; j int{@EGLProperty{setMethod = setj}}; function start() end function geti() returns(int) return(i); end function setj(x int) j = x; end end handler rui2 type RUIhandler {initialUI = [],onConstructionFunction = start, cssFile="css/ruiProj.css", title="rui2"} function start() r rui1{}; r.j = 33; syslib.writestdout(r.i); end end rui1 fails on "return(i);" with a CCE rui2 fails on "r.j = 33;' with the same CCE. If you comment that line out, it then fails on "syslib.writestdout(r.i);" with the same CCE. The CCE is trying to cast a MemberNameImpl to a String. I looked at the annotation source and it's indeed a FunctionMemberRef for EGLProperty (and a string for Property). Finally, the real error in rui1 is that the getter/setter should not be used if accessed within the part that declared the field - otherwise it would get into an infinite loop.