Hi, after your announce i've the follow. feature request:
Linking OOP-ObjectPascal-Progs to (shared) C++-Libs ! Of course you can't use templates and exceptions and so on, but it would be really useful.
Regards, Stephan
Hallo, Stephan!
Linking OOP-ObjectPascal-Progs to (shared) C++-Libs ! Of course you can't use templates and exceptions and so on, but it would be really useful.
You already can do that if you write a wrapper that translates the linker names of Pascal methods to C++ member functions or vice versa.
If someone knows the exact, non-changing naming conventions for C++ member functions we can discuss about making them the default for Pascal as well. But right now it is really simple to write or to access Pascal methods in plain C which is an adventage, too.
Schöne GrüÃe,
Peter
On Thu, 4 Feb 1999, Peter Gerwinski wrote:
If someone knows the exact, non-changing naming conventions for C++ member functions we can discuss about making them the default for Pascal as well.
... but unfortunately there's no such non-changing mangling scheme. Different versions of GCC and/or EGCS all require their very own copy of libstdc++. Still, maybe you can borrow code from c++filt, from the GCC you're building GPC with, to do something useful.
JanJaap
--- GNU mingw32: http://agnes.dida.physik.uni-essen.de/~janjaap/mingw32
On Thu, 4 Feb 1999, Peter Gerwinski wrote:
If someone knows the exact, non-changing naming conventions for C++ member functions we can discuss about making them the default for Pascal as well. But right now it is really simple to write or to access Pascal methods in plain C which is an adventage, too.
asmname :) It would work if asmname was accepted in an object declaration (I'm sorry if I make it sound so easy, but GPC already has the mechanisms to asmname methods. Think about it, currently you can declare an external assembly language file with the method and object name without having to do anything other than prototype it in GPC. Therefore the next step would be to give GPC a name to use instead of declaring its own especially since it does not call a partial method name anyway). I honestly don't see why asmname could/should not be extended to methods.
See ya! Orlando Llanes
"Meine Damen und Herren, Elvis hat soeben das Gebaeude verlassen!"
"Look out fo' flyeeng feet" O__/ a010111t@bc.seflin.org /|____. O <__. /> / \ ____________|_________ http://ourworld.compuserve.com/homepages/Monkey414
Am Thu, 04 Feb 1999 schrieben Sie:
Hallo, Stephan!
Linking OOP-ObjectPascal-Progs to (shared) C++-Libs ! Of course you can't use templates and exceptions and so on, but it would be really useful.
You already can do that if you write a wrapper that translates the linker names of Pascal methods to C++ member functions or vice versa.
If someone knows the exact, non-changing naming conventions for C++ member functions we can discuss about making them the default for Pascal as well. But right now it is really simple to write or to access Pascal methods in plain C which is an adventage, too.
Sorry I use an new Mailprog and forgot the Subject :( But what I really meant was a _easy_ way to program ObjectPascal in an C++-Library-Enviroment, also with a tool converting C++-Headers with all Objects to units. I dont know, perhaps this is already possible _without_ hacking, but this would also make OO-Pascal Programming much more useful and acceptible in an Unix-Enviroment. Especially for programming GUI, distributed objects and so on... You know, C++ _seems_ to replace C in such parts of software development, so the Pascal<->C-Interface is perhaps not enough. C++ is also more difficult than C, and can be very complex if you use all the features of this language. So people use Java. But this are only thoughts, maybe no one would really use this. And of course if you can only access such things, which are possible in Pascal, so perhaps you cant't use all of the classes of such a lib, and this makes it worthless.
Gruß, Stephan
-
Jan-Jaap van der Heijden -
Orlando Llanes -
Peter Gerwinski -
Stephan Goetter