- Boost - lists.boost.org

[boost.test] feature request - random order of testing
by Pavel Vozenilek 23 May '04

23 May '04

Would it be possible to add ability to execute tests in random order, instead of as they are added into suite? Example: test_suite* init_unit_test_suite(int argc, char * argv[]) { test_suite* tests = BOOST_TEST_SUITE("..."); tests->add(BOOST_TEST_CASE(&test1)); tests->add(BOOST_TEST_CASE(&test2)); return tests; } Here test1() will be always executed before test2(). This behaviour may hide errors. It would be nice to have: - possibility to say: now execute test in random order, - if error happens framework would print out seed for random, - ability set seed manualy is needed to be able to repeat problem. (If such feature is already present in library my weak reading skills didn't spot it.) /Pavel

3 4

[Serialization] Errors when compiling using GCC 3.4
by Allen Yao 23 May '04

23 May '04

I just updated my gcc from 3.3.3 to 3.4, and found that the serialization17 and serialization18 cannot compile. There are many error messages, some of them seem to caused by lack of "typename", others are very strange, for example, about using boost::noncopyable. Does someone try serialization library on GCC 3.4? Can it be compiled? I suspect I am missing something. Otherwise, I would like to try to "fix" the "bugs" myself. Best regards.

1 0

Re: [boost] Re: [Boost.Test] New testing procedure?
by Reece Dunn 23 May '04

23 May '04

Jeff Garland wrote: >And that's today. Consider during the next couple months 3-4 new libraries >are pending to be added. Serialization tests alone dramatically increase >the >length of the time to run the regression if we always run the full test. >What >will happen in a year when we have say 10 new libraries? > >Robert and I have believe something will need to be done. We've tried to >start a discussion, but no one responded: > >http://lists.boost.org/MailArchives/boost/msg64471.php >http://lists.boost.org/MailArchives/boost/msg64491.php I don't know how feasable this would be, but you might also want to consider library dependancies so that if you modify the iterator adaptors library you can pull a list of tests that need to be re-run because the associated libraries use iterator adaptors. This would reduce the number of tests that need to be performed on each cycle, but you'd need some way of archiving the results and only updating the ones that are re-run. When testing a new platform/compiler/configuration, you can tell the test suite to run all the tests for that setup, but for the others (where it has already been run) it will operate as above, unless an explicit request for the full set of tests is provided (allowing for, say, the full set to be performed weekly while the dependant tests are performed daily/every 12 hours). I like the idea of specifying the test level (basic, torture, concept). Regards, Reece _________________________________________________________________ Want to block unwanted pop-ups? Download the free MSN Toolbar now! http://toolbar.msn.co.uk/

3 6

RE: [boost] Re: [Boost.Test] New testing procedure?
by Reece Dunn 23 May '04

23 May '04

David Abrahams wrote: >"Jeff Garland" <jeff(a)crystalclearsoftware.com> writes: > > > On Sat, 22 May 2004 13:18:41 -0400, David Abrahams wrote > >> > I like the idea of specifying the test level (basic, torture, >concept). > >> > >> I am opposed to the idea of requiring humans to initiate the right > >> tests, at least without proof that mechanically-initiated tests are > >> unworkable. I don't think we've proven that yes. > > > > They shouldn't be human initiated -- just configured to one level or >another. > > Now obviously a human would need to intervene to reset the level, but >that > > would only be in the case of ramping up to release, adding a new >library, > > adding a new compiler, or a specific author request. I believe this >would be > > rather infrequent. I was not suggesting that the basic/full tests be initiated by a human. This is what shell scripts are for: full-tests.sh and basic-tests.sh! Then tell a computer to run full-tests.sh once a week and basic-tests.sh daily. > > Basically, right now we have no way of setting up a Jamfile that can >satisfy > > the need to have full up tests for new compilers, etc and a basic set of >quick > > tests...short of replacing the Jamfile all the time. > >I don't see that as a need, or neccessarily an advantage. Only >outdated tests get run. > >That said, of course we do have a way: > > # foo.jam > > if $(Not_Defined) > { > # torture tests go here > } > >In what way is commenting out the "if" line and checking in the >jamfile significantly different from other mechanisms one might use to >alter the set of tests that get run? Is it possible to add the level as a parameter like how you can configure static/dynamic linking, e.g.: test basic : basic.cpp : <level>basic ; test complete : complete.cpp : <level>torture ; # basic-tests.sh bjam ... basic-test # full-tests.sh bjam ... full-test Regards, Reece _________________________________________________________________ It's fast, it's easy and it's free. Get MSN Messenger today! http://www.msn.co.uk/messenger

2 2

scoped_ptr - custom deleter?
by Robert Ramey 23 May '04

23 May '04

Question: Shouldn't scoped_ptr have a constructor of the form: template<class Y, class D> scoped_ptr(Y * p, D d) where D is a custom deleter. Shared_ptr has it. I have a need for this now. This has been raised during a discussion with Martin Ecker re serialization library implementation. Robert Ramey.

1 0

any interest in a producer consumer template?
by Arash Partow 23 May '04

23 May '04

Hi all, I was wondering if there would be any interest in a template library based around producers and consumers. I've developed something already and have a series of working examples, I propose a library that consist of 3 basic entities, a producer a consumer and a hybrid entity called an interchange. the entities themselves will be threaded and will be linked together via a linking mechanism that is itself a base template but which can be extended to provide data transfer between same system processes and threads and also extended versions which provide data transfer over pipes, sockets, serial lines etc.. the idea of the producer-consumer is very simple the producer produces a type of data ie: string, a class, a struct etc.. and a consumer that is instantiated as a consumer of the type will consume the data provided by the producer and do something with it. The interchange entity is a transformation entity, it acts both as a consumer and as a producer, it consumes from a producer or another interchange, and then invokes the user implemented transmute method to convert/munge/process the data from one type to another then sends it to its consumer. Simple class syntax: template < typename OutputType // class IdleProductionPolicy = Policy class for implementing producer's //behavior during idle production periods. > class Producer : public Thread { }; template <typename InputType> class Consumer : public Thread { }; template < typename InputType, typename OutputType // template < typename > class InternalProducer possible future policy > class Interchange : public Thread { }; The link mechanism is a basically a type that supports the common type between a producer and its consumer, it is a one way flow of data, the consumer is signaled when data is ready for it to collect if its asleep it will awake and fetch the data if its doing something else when it comes back to go and wait for the condition from the producer to be signaled it will instantly be awaken and go off and fetch the data off the link. the data on the link or more precisely in the container which resides in the link is made thread safe via mutexes around the add and get methods. A link has a basic get and add functionality but depending what entity is being implemented they will only see it via an interface and hence only be able to invoke 1 kind of method, a producer can only invoke "add" where-as a consumer can only invoke a get, an interchange can invoke both because it is instantiated with both a consumer and producer link interface, but the links are separate for an interchange, because it requires 2 links, one link for the incoming data and one link for the out going data. both link maybes of different types, ie: Producer <string> ----> Interchange <string, AClass> ----> Consumer <AClass> I need some opinions, about whether or not people will find this library useful and hence have it as part of Boost, if so could people give me their opinions about my design, its just a simple first trial run at designing it, i intended for the producers and interchange entities to be policy driven as far as production of data is concerned. I think this library can be used in distributed computing, in network packet assembly, and also for interfacing and passing data up and down threaded software layers. Any advice, opinions would be very much appreciated. Kind regards Arash Partow __________________________________________________ Be one who knows what they don't know, Instead of being one who knows not what they don't know, Thinking they know everything about all things. http://www.partow.net A more detailed class syntax and usage is here: template < typename OutputType // class IdleProductionPolicy = Policy class for implementing producer's //behavior during idle production periods. > class Producer : public Thread { private: typedef vector < OutputType > TypeVector; typedef ProducerLink < OutputType > TypeLink; public: Producer(TypeLink* _link, unsigned int _minpendbufsize = MIN_PEND_BUF_SIZE, unsigned int _maxpendbufsize = MAX_PEND_BUF_SIZE, unsigned int _id = 0) { }; ~Producer(){}; unsigned int getMinPenBufferSize() { return minpendbufsize; }; unsigned int getMaxPenBufferSize() { return maxpendbufsize; }; void setMinPenBufferSize(int _minpenbufsize = MIN_PEND_BUF_SIZE) { minpendbufsize = _minpenbufsize; }; void setMaxPenBufferSize(int _maxpenbufsize = MAX_PEND_BUF_SIZE) { maxpendbufsize = _maxpenbufsize; }; virtual void setActiveMode(bool state = true) { }; protected: void add(OutputType t) {}; void add(const TypeVector& tlist) {}; virtual bool manufacture()=0; private: void execute() {}; }; template <typename InputType> class Consumer : public Thread { private: typedef vector < InputType > TypeVector; typedef ConsumerLink < InputType > TypeLink; public: Consumer(TypeLink* _link){}; ~Consumer(){}; protected: void get(TypeVector& tlist) { }; void setLink(TypeLink& _link) { }; virtual bool consume()=0; private: void execute(){}; }; template < typename InputType, typename OutputType // template < typename > class InternalProducer possible future policy > class Interchange : public Thread { public: typedef vector < InputType > ConsDataList; typedef vector < OutputType > ProdDataList; typedef ConsumerLink < InputType > ConsLink; typedef ProducerLink < OutputType > ProdLink; public: Interchange( ConsLink* _consLink, // Link from the producer entity ProdLink* _prodLink, // Link to the consumer entity unsigned int _minpendbufsize = MIN_PEND_BUF_SIZE, unsigned int _maxpendbufsize = MAX_PEND_BUF_SIZE, unsigned int _id = 0 // for debugging purposes ) { }; ~Interchange() {}; unsigned int getMinPenBufferSize() {}; unsigned int getMaxPenBufferSize() {}; void setMinPenBufferSize(int _minpenbufsize = MIN_PEND_BUF_SIZE) { }; void setMaxPenBufferSize(int _maxpenbufsize = MAX_PEND_BUF_SIZE) { }; void setActiveMode(bool state = true){}; protected: void add(OutputType t) { prodLink->add(t); }; void add(const ProdDataList& tlist) { prodLink->add(tlist); }; void get(ConsDataList& tlist) { consLink->get(tlist); }; void setConsumerLink(ConsLink& _link) { consLink = _link; }; void setProducerLink(ProdLink& _link) { prodLink = _link; }; virtual bool transmute()=0; private: void execute(){}; }; Usage would be as follows: class AClass { public: AClass(string str):val(str){} ~AClass(){}; void displa(){cout << val << endl;}; private: string val; }; class MyProducer : public Producer<string> { public: typedef ProducerLink < string > LinkType; MyProducer(LinkType* _link, int minpenbs, int maxpenpbs):Producer<string>(_link, minpenbs, maxpenpbs){}; bool manufacture() { add("abc"); return true; }; }; class MyInterchange : public Interchange<string,AClass> { public: typedef ProducerLink < string > ProdLinkType; typedef ConsumerLink < string > ConsLinkType; MyInterchange( ConsLinkType* _conslink, ProdLinkType* _prodlink, unsigned int minpenbs, unsigned int maxpenbs ):Interchange<string,string>( _conslink, _prodlink, minpenbs, maxpenbs, ){}; bool transmute() { vector <string> slist; get(slist); for(unsigned int i = 0; i < slist.size(); i++) { add(AClass(slist[i])); } add(slist); return true; }; private: }; class MyConsumer: public Consumer<AClass> { public: typedef ConsumerLink < string > LinkType; MyConsumer(LinkType* _link, int mpID):Consumer<string>(_link){}; bool consume() { vector <AClass> alist; get(alist); for(unsigned int i = 0; i < slist.size(); i++) { alist[i].display(); } return true; }; }; int main () { Link < string > link1; Link < AClass > link2; MyProducer myproducer = MyProducer(new ProducerLink <string>(&link1),10,50); MyInterchange myinterchange = MyInterchange(new ConsumerLink <string>(&link1),new ProducerLink <AClass>(&link2),1,10); MyConsumer myconsumer = MyConsumer(new ConsumerLink <AClass>(&link2)); myproducer.start(); myinterchange.start(); myconsumer.start(); while(true) { . . . } return true; }

2 2

MSDN News of the week
by Edward Diener 22 May '04

22 May '04

The news of the week at MSDN, first item, http://msdn.microsoft.com/ mentions Boost. All Boost developers can take a bow and to them I also offer my applause.

1 1

Re:Re: [boost] Re: [Boost.Test] New testing procedure?
by Robert Ramey 22 May '04

22 May '04

Reece Dunn wrote: > I don't know how feasable this would be, but you might also want to > consider library dependancies so that if you modify the iterator adaptors > library you can pull a list of tests that need to be re-run because the > associated libraries use iterator adaptors. This would reduce the number > of tests that need to be performed on each cycle, but you'd need some way > of archiving the results and only updating the ones that are re-run. I thought that's what it already did! Isn't that the whole point of the BJAM system? Is there not a way to "compose" Jamfile(s) into a one so that only dependent tests are run? Just a thought. More food for thought. The serialization library relies on spirit. The current version of spirit (1.8?) doesn't build on older platforms such as VC 6. For me this isn't a problem. For testing on other platforms spirit 1.61 will have to be installed in a "known" place. These kind of testing tweaks can occur inside the runtest.sh but that would inhibit the composition of Jamfiles described above. - maybe this is not a consideration. I recognize that the current testing procedure is exhaustive and that's a good thing. If I were to add exhaustive testing of the serialization library (35 tests * 5 archives (maybe 10) * 2 (debug/release) * 2 (dll/lib) * 8 compilers ) that would be 5600 build and run tests. This would take a while. (I havn't included polymorphic archive versions which would double this). So I'm resigned to the fact that I'm not going to be able to test exhaustively on a continuing basis. I would like to be able to solicite a test for a given combination on demand. Robert Ramey

1 0

boost.css validation
by Gennaro Prota 22 May '04

22 May '04

Hi, the W3C online validator accessible at http://jigsaw.w3.org/css-validator/ gives a warning on our main .css file: Warnings URI : http://www.boost.org/boost.css * Line : 0 font-family: You are encouraged to offer a generic family as a last alternative This looks like a good idea. What about adding 'monospace' as the last choice for both the PRE and the CODE rules? Genny.

2 1

[MPL] Integral constant expressions and MSVC.
by Cromwell Enage 22 May '04

22 May '04

Morning, all. I'm trying to implement a compile-time greatest common divisor metafunction, a key component of which is boost::mpl::modulus_c. Am I correct if I say that the following is typical usage of this class template? #include <iostream> #include <boost/mpl/modulus.hpp> int main() { typedef typename boost::mpl::modulus<int,14,6>::type remainder_type; std::cout << "The remainder of 14 divided by 6 is "; std::cout << remainder_type::value << std::endl; return 0; } MinGW 3.2 accepts this usage, but MSVC 7.1 files the following complaint: ..\..\Libraries\boost\boost\mpl\modulus.hpp(39) : error C2975: 'N' : invalid template argument for 'boost::mpl::integral_c', compile-time evaluatable constant expression expected It seems MSVC doesn't recognize A % B as an integral constant expression. What's the proper way of handling this? Cromwell Enage __________________________________ Do you Yahoo!? Yahoo! Domains � Claim yours for only $14.70/year http://smallbusiness.promotions.yahoo.com/offer

2 2