c++.announce - Wanted: people to test new MFCSTL array adaptors: CArray_facade and CArray_proxy
- Matthew (37/37) Aug 05 2005 Title pretty much says it all.
- Gabor.Fischer systecs.com (Gabor Fischer) (4/5) Aug 05 2005 Count me in. I can test it under Visual C++ 7.0.
- Jan Knepper (7/12) Aug 05 2005 We should not curse in these newsgroups... ;-)
- Matthew (9/15) Aug 05 2005 Ok, thanks Gabor and Pablo. I suspect you two will easily suffice to
- Pablo Aguilar (1/2) Aug 05 2005 Sure... I'll take a shot at it as well...
- Dejan Lekic (9/9) Aug 13 2005 Mr. Wilson,
- Matthew (28/39) Aug 16 2005 Hi Dejan
- Dejan Lekic (11/11) Aug 16 2005 I do not want to disappoint You, but I am neither using, nor intend to u...
Title pretty much says it all.
I've spent the last few days totally rewriting the existing MFCSTL
CArray-family adaptors, and the new classes I've got support a very
large amount of the functionality of std::vector. The CArray_facade
is
a compile-time adaptor, and the CArray_proxy is a runtime adaptor.
They're both derived - with a little CRTP/SCTP magic - from a common
abstract CArray_adaptor class. (CRTP=Curiously Recurring Template
Pattern; SCTP=Simulated Compile Time Polymorphism)
I'm looking for volunteers to play around with it, and provide
feedback on functionality/correctness/etc.
I've been concurrently writing up the adaptation as a chapter for
"Extended STL" - on which I'm now heavily focused - so it's entirely
possible that salient feedback may find it's way into print. I'm
also
considering whether I do a write-up in my online column, or an
article.
Some small additional motivation, then, if such fame lights your
candle. ;-)
Depending on amount and form (i.e. ng posts vs emails) feedback,
I'll either post an alpha zip here, or will email volunteers
directly
Cheers
--
Matthew Wilson
Author: "Extended STL", Addison-Wesley, 2006
(http://www.extendedstl.com)
Author: "Imperfect C++", Addison-Wesley, 2004
(http://www.imperfectcplusplus.com)
Contributing editor, C/C++ Users Journal
(http://www.synesis.com.au/articles.html#columns)
STLSoft moderator
(http://www.stlsoft.org)
-------------------------------------------------------------------------------
"So far, C++ is the best language I've discovered to say what I want
to say" -- Alex Stepanov
-------------------------------------------------------------------------------
Aug 05 2005
Hello Matthew!Title pretty much says it all.Count me in. I can test it under Visual C++ 7.0. So Long... Gabor
Aug 05 2005
Jan Knepper <jan smartsoft.us> Gabor Fischer wrote:Hello Matthew!We should not curse in these newsgroups... ;-) -- ManiaC++ Jan Knepper But as for me and my household, we shall use Mozilla... www.mozilla.orgTitle pretty much says it all.Count me in. I can test it under Visual C++ 7.0.
Aug 05 2005
Ok, thanks Gabor and Pablo. I suspect you two will easily suffice to give this thing a thorough going over. I've already done a comprehensive unit-test, so I'm hopeful it works right, but I'm also keen to see if I've hit the sweet spot with the feature set. I'll email a zip to you both, in a few hours. Cheers Matthew "Gabor Fischer" <Gabor.Fischer systecs.com> wrote in message news:9bJvydTpQNB systecs.com...Hello Matthew!Title pretty much says it all.Count me in. I can test it under Visual C++ 7.0. So Long... Gabor
Aug 05 2005
Title pretty much says it all.Sure... I'll take a shot at it as well...
Aug 05 2005
Mr. Wilson, may I ask what are You using for unit-testing? Some in-house solution, or some open-source solution? In both cases - would You like to give me some pointer to it in form of an URL. :) -- ........... Dejan Lekic http://dejan.lekic.org
Aug 13 2005
Hi Dejan
The unit-testing framework is one I came up with for STLSoft, rather
than an external solution. It qualifies as open-source, since all
the pieces are available in the STLSoft distributions and free for
people to use, but I've not documented it as yet.
The way it works is that, in the context of the symbol
STLSOFT_UNITTEST being defined, every header will include a block of
code that is automatically registered with a unit-tester instance.
When this instance is invoked to run the tests, all included files
are tested. Hence, for the internal unit-testing purposes, each
sub-project consists of
- a driver program, e.g. unittest/winstl/unittest.cpp,
containing an invocation of
stlsoft::unittest::simple_unittest_host::get_host()->test();
- the implementation for the simple_unittest_host class,
contained in include/unittest/simple_unittester.cpp
- several script-generated files that include all the headers
for a given sub-project, e.g. unittest/winstl/unittest.0.cpp,
unittest/winstl/unittest.1.cpp, unittest/winstl/unittest.2.cpp, ...
unittest/winstl/unittest.11.cpp
When compiled in the context of STLSOFT_UNITTEST, these build a
program that runs _all_ the WinSTL unittests.
If you like, we could have a dialogue in this newsgroup about it,
which I would then be glad to turn into an FAQ?
Cheers
Matthew
"Dejan Lekic" <leka entropy.tmok.com> wrote in message
news:ddlsm8$m5t$1 digitaldaemon.com...
Mr. Wilson,
may I ask what are You using for unit-testing? Some in-house
solution, or
some open-source solution? In both cases - would You like to give
me some
pointer to it in form of an URL. :)
--
...........
Dejan Lekic
http://dejan.lekic.org
Aug 16 2005
I do not want to disappoint You, but I am neither using, nor intend to use STLSoft. I just would like to have small unittest framework for my little c++ framework I am working on, which is based on STDC++ classes and templates. Sure, I would gladly participate in any C++ related discussion here, whether it does, or not, involve STLSoft. :) -- ........... Dejan Lekic http://dejan.lekic.org
Aug 16 2005