[Issue 1595] Multiple typedefs of Exception should be treated as different types.

[Issue 1595] (D1 only) Multiple typedefs of Exception should be treated as different types.

Feb 05, 2012

http://d.puremagic.com/issues/show_bug.cgi?id=1595 thomas-dloop@kuehne.cn changed: What |Removed |Added ---------------------------------------------------------------------------- AssignedTo|dvdfrdmn@users.sf.net |bugzilla@digitalmars.com OS/Version|Linux |All Platform|PC |All Version|1.022 |1.018 ------- Comment #2 from thomas-dloop@kuehne.cn 2007-10-20 02:04 ------- David, what part of the typedef / try-catch documentation renders this issue an enhancement instead of a bug? Added to DStress as http://dstress.kuehne.cn/run/c/catch_08_A.d http://dstress.kuehne.cn/run/c/catch_08_B.d http://dstress.kuehne.cn/run/c/catch_08_C.d http://dstress.kuehne.cn/run/c/catch_08_D.d http://dstress.kuehne.cn/run/c/catch_08_E.d http://dstress.kuehne.cn/run/c/catch_08_F.d http://dstress.kuehne.cn/run/c/catch_08_G.d --

http://d.puremagic.com/issues/show_bug.cgi?id=1595 ------- Comment #3 from dvdfrdmn@users.sf.net 2007-10-20 06:54 ------- Exception handling is based on class inheritance. Supporting typedefs (using TypeInfo RTTI) is above and beyond what is described in the spec. --

October 20, 2007

[Issue 1595] Multiple typedefs of Exception should be treated as different types.

Posted by d-bugmail

Permalink

d-bugmail

Permalink

http://d.puremagic.com/issues/show_bug.cgi?id=1595





------- Comment #4 from thomas-dloop@kuehne.cn  2007-10-20 10:31 -------
http://www.digitalmars.com/d/declaration.html
# Strong types can be introduced with the typedef. Strong
# types are semantically a distinct type to the type checking
# system, for function overloading, and for the debugger.

As far as I understand the quote, typedefs create distinct types. For classes I always asumed that

#
# typedef Master Slave;
#

was syntactic sugar for

#
# class Slave : Master{
#       this(...){
#               super.this(...);
#       }
# }
#

however the current implementation only creates a
strong-at-compiletime and weak-at-runtime type.
Test code:

# import std.stdio;
#
# class Master{
# }
#
# typedef Master Slave;
#
# void main(){
#    Object o;
#
#    o = new Master(); // 1
#    writefln("1: %s", o);
#
#    o = new Slave(); // 2
#    writefln("2: %s", o);
# }

As the debugger currently can't differentiate the types of object 1 and 2 I'd rate this a general bug that was triggered in this case by "catch".


--

http://d.puremagic.com/issues/show_bug.cgi?id=1595 Yao Gomez <yao.gomez@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |yao.gomez@gmail.com --- Comment #5 from Yao Gomez <yao.gomez@gmail.com> 2012-02-04 22:32:00 PST --- This is a D1 only issue (obviously, as D2 will deprecate typedef). However, the version currently selected is 1.018. What should be done with this reports (besides fixing them) with specific version? Should be changed to the more generic D1? -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email ------- You are receiving this mail because: -------

http://d.puremagic.com/issues/show_bug.cgi?id=1595 yebblies <yebblies@gmail.com> changed: What |Removed |Added ---------------------------------------------------------------------------- CC| |yebblies@gmail.com Version|1.018 |D1 Summary|Multiple typedefs of |(D1 only) Multiple typedefs |Exception should be treated |of Exception should be |as different types. |treated as different types. --- Comment #7 from yebblies <yebblies@gmail.com> 2012-02-05 18:41:43 EST --- Yes, the individual versions are not used any more. Having (D1 only) as a prefix in the title also helps differentiate the bugs. -- Configure issuemail: http://d.puremagic.com/issues/userprefs.cgi?tab=email ------- You are receiving this mail because: -------

Forums