- Boost - lists.boost.org

C++ Builder 2009 and Boost compliance
by joaquin＠tid.es 15 Sep '08

15 Sep '08

C++ Builder 2009 has been recently launched and there are some claims about its having much better Boost support than previous versions of the product (see for instance http://tinyurl.com/5r2yxu or http://tinyurl.com/6zjnvx ). Does anyone have first-hand experience of C++ Builder 2009 and can verify these claims? It'd be nice if we could have a regression test runner for this product. Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

5 6

[review] FSM Library review result
by Martin Vuille 15 Sep '08

15 Sep '08

This is a summary of the comments received during the review of Andrey Semashev's Finite State Machines (FSM) library. Apologies for the delay in getting these out. DOCUMENTATION * Relocate ToC to make it stand out. * Several people used the Wikipedia article http://en.wikipedia.org/wiki/Finite_state_machine as a reference for FSM terminology and operation, and found discrepancies between Boost.FSM and that description. Perhaps Boost.FSM should explicitly describe its FSM metamodel, or refer to an existing metamodel (Wikipedia FSM or Harel or UML Statechart, for example). The latter was mentioned several times as being preferred since it is a standard. * If possible, use standard terminology. For example, "transition map" vs. "state transition table". Or define the non-standard terms. * Include a comprehensive rationale mapping the design choices to the design objectives. * Compare/contrast Boost.FSM with the FSM implementation as described in "C++ Template Metaprogramming". * Add an example of event deferral using boost::function. * Clarify support for multithreading vs. asynchronous FSM, as some were confused by existing text. * Clarify lifetime of objects. * Add implementation of example using transition map/STT. * Add more complex example(s). SCOPE/FEATURES These were features that were suggested or asked about. Some or all may not be suited for addition to the library. In that event, it may be useful to explicitly mention in the documentation that they are not supported, and why. * Support nested states. Return-to-history was cited as an example of a benefit of nested states over nested FSM. * In STT, support matching "any" from state. * Add syntactic sugar to support "synthetic" states (cf. http://lists.boost.org/Archives/boost/2008/08/140714.php) * Support event deferral. * Support data-dependent transitions (perhaps this is the same as the synthetic state suggestion above). * Improve modeling of the "transition" concept. Specifically, there was an issue with the fact that the action was associated with a state rather than a transition. INTERFACE * There was a lot of discussion comparing the amount of code required to implement an "ad hoc" FSM, vs. the amount of code required to implement the same FSM using Boost.FSM. * The "transition map" representation was deemed to contain too much syntactic noise and not sufficiently similar to a traditional STT, and it was suggested that it should include the transition action. * Some felt that the interface was too procedural rather than being declarative. IMPLEMENTATION * There was quite a bit of discussion of the Boost.FSM implementation vs. the C++TM implementation, but I don't believe there were any specific conclusions reached. * Handling of event is different depending on whether or not transition map is used (handled by "to" state vs. current state, respectively). This is also the case when an event is not found in the transition map. This could lead to errors. Thanks again to Andrey for his submission and to all the reviewers and other who made comments for their input. MV

2 1

[bind] mem_fn_eq_test test failure with VC9 in release builds
by Richard Webb 15 Sep '08

15 Sep '08

When running the Trunk regression tests using VC9SP1 and building in release mode, the bind/mem_fn_eq_test test fails. Results available @ http://tinyurl.com/6nfm9m. The same test passes when the tests are run in the default debug mode. Thanks, Richard Webb

2 1

[exception] Virtual destructor warnings (again)
by Chris Newbold 15 Sep '08

15 Sep '08

I'm running into another instance of virtual destructor warnings with Boost.Exception -- this one when building on Solaris using the Sun Studio 12 compiler. The problem is with the boost::exception_detail::exception_clone class template. It uses CRTP and is often instantiated with a type which has a virtual destructor which has been declared 'throw()'. This causes the compiler to complain that the exception specification for exception_clone's implicitly-generated destructor is too wide: "../../../3p_mirror/sol64/boost/include/boost-1_36/boost/exception/enable_current_exception.hpp", line 121: Warning, wlessrestrictedthrow: Function boost::exception_detail::exception_clone<boost::exception_detail::error_info_injector<boost::bad_function_call>>::~exception_clone() can throw only the exceptions thrown by the function boost::exception_detail::error_info_injector<boost::bad_function_call>::~error_info_injector() it overrides. "../../../3p_mirror/sol64/boost/include/boost-1_36/boost/exception/enable_current_exception.hpp", line 130: Where, temwhilespec: While specializing "boost::exception_detail::exception_clone<boost::exception_detail::error_info_injector<boost::bad_function_call>>". "../../../3p_mirror/sol64/boost/include/boost-1_36/boost/exception/enable_current_exception.hpp", line 130: Where, teminstfrom: Instantiated from boost::function3<void, int&, const char**&, const char**>::operator()(int&, const char**&, const char**) const. "../../../3p_mirror/sol64/boost/include/boost-1_36/boost/signals/signal_template.hpp", line 119: Where, teminstfrom: Instantiated from boost::signal3<void, int&, const char**&, const char**, boost::last_value<void>, int, std::less<int>, boost::function<void(int&,const char**&,const char**)>>::operator()(int&, const char**&, const char**). The fix seems simple enough, and is compatible with other compilers: add an empty destructor declared 'throw()' to exception_clone. Here's the diff: ==== //3rdparty/tmw/boost/boost/exception/enable_current_exception.hpp#1 - /mathworks/devel/sandbox/cnewbold/3rdp/3p/sources/boost/boost/exception/enable_current_exception.hpp ==== @@ -96,6 +96,10 @@ *be1 = *be2; } + ~exception_clone() throw() + { + } + private: detail::atomic_count mutable count_; -Chris

2 2

[BENUM] Any interest in auto-generated enum-to/from-string functions, without longer/more-complex usage?
by Tony D 15 Sep '08

15 Sep '08

After two expressions of interest (Keith Burton / Lance Diduck Vol 2186 Issue 4), reviews by Andrew Marlow and a few months, I've finally got around to packaging BENUM and putting it in the vault. Further consideration/feedback/questions appreciated. Specifically, I think the optional features and mark-up require fuller consideration and perhaps maturation.... Tony

3 2

Re: [boost] [switch] Is there any interest in library switch extension?
by Kasra Nassiri(Math & ComSci) 15 Sep '08

15 Sep '08

I am not sure how it could be performed on O(1), however, the current algorithm in O(n). Current algorithm is something like: BOOST_SWITCH( "hello world" ) { #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("invalid case") std::cout << "Reached invalid case!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("another string") std::cout << "Reached invalid case!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE( "hello world" ) std::cout << "Reached the correct code!" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("hello world 1") std::cout << "Reached the correct code! (no breaks)" << std::endl; BOOST_BREAK; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("hello world") std::cout << "Should throw exception! (duplicated value)" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_CASE("another string 1") std::cout << "Reached invalid case! (broken)" << std::endl; #include BOOST_PP_UPDATE_COUNTER() BOOST_DEFAULT std::cout << "Reached default Hello!" << std::endl; } BOOST_SWITCH_END It works fine but it is O(n) and uses hash tables. I want to get rid of "#include BOOST_PP_UPDATE_COUNTER()" somehow, but at this time I can't. Even when you use BOOST_BREAK at early statements it is still O(n), since it checks for duplicated values. But using break would change to O(m) : m < n. The 'default' statement at the moment must be at the end of all other cases, however, it will be updated so it could be anywhere with in the switch. By the way where is this switch_utility? Any other suggestions?

1 0

[BENUM] Any interest in auto-generated enum-to/from-string functions, without longer/more-complex usage?
by Tony D 15 Sep '08

15 Sep '08

Anyone think this functionality belongs in boost? (I have working code) ENUM(X, B = N, C, D = B + C }; // or whatever X x1 = C; cout << x1 << '\n'; // outputs "C\n" X x2("D"); // construct from string... if (x2 == D) { will do } Tony __________________________________________________________ Sent from Yahoo! Mail. A Smarter Email http://uk.docs.yahoo.com/nowyoucan.html

4 3

[boost.test] Customized test suite gives SIGSEGV when setup fails
by Markus Schöpflin 15 Sep '08

15 Sep '08

Hello, I'm getting a SIGSEGV when trying the following with Boost 1.35.0 and gcc 4.1.2: ---%<--- #define BOOST_TEST_ALTERNATIVE_INIT_API #include <boost/test/included/unit_test.hpp> using namespace boost::unit_test; struct my_test_suite : public test_suite { my_test_suite() : test_suite("my_test_suite") { throw std::runtime_error("runtime_error"); } }; bool init_unit_test() { test_suite *test = BOOST_TEST_SUITE(""); test->add(new my_test_suite()); framework::master_test_suite().add(test); return false; } ---%<--- When run in the gdb, I get the following output: ---%<--- (gdb) run Starting program: a.out Test setup error: std::runtime_error: runtime_error Program received signal SIGSEGV, Segmentation fault. 0x0806e2fa in ~framework_impl (this=0x8089740) at boost/test/impl/framework.ipp:126 126 delete (test_suite const*)tu.second; --->%--- The code in question has been ported from an old Boost.Test version to 1.35. When looking at the documentation for Boost.Test (1.36, as this seems more in sync with Boost.Test 1.35) I'm unsure that custom test suites are still supported in this way. What I need to do is to set up and tear down a database connection at the start/end of a test suite execution. Is there some other way for this? TIA, Markus

2 3

[Boost-Test] Patch xml_log_formatter.hpp for sun-5.9 on Linux
by K. Noel Belcourt 15 Sep '08

15 Sep '08

Hi, Attached is a patch to fix this problem with Boost.test trunk and sun-5.9 on Linux. sun.compile.c++ ../../../bin.v2/libs/test/build/sun-5.9/debug/link- static/unit_test_main.o "../../../boost/test/impl/unit_test_main.ipp", line 115: Error: Could not find a match for std::vector<boost::unit_test::test_case_filter::single_filter>::vector (boost::unit_test::basic_string_token_iterator<char, boost::unit_test::ut_detail::default_char_compare<char>>, boost::unit_test::basic_string_token_iterator<char, boost::unit_test::ut_detail::default_char_compare<char>>) needed in boost::unit_test::test_case_filter::test_case_filter (boost::unit_test::basic_cstring<const char>). "../../../boost/test/impl/unit_test_main.ipp", line 115: Error: Formal argument x of type const std::vector<boost::unit_test::test_case_filter::single_filter>& in call to std::vector<std::vector<boost::unit_test::test_case_filter::single_filte r> >::push_back(const std::vector<boost::unit_test::test_case_filter::single_filter>&) is being passed int. 2 Error(s) detected. "/home/sntools/extras/compilers/studio12/sunstudio12/bin/CC" +d - g -DBOOST_ALL_NO_LIB=1 -I"../../.." -c -o "../../../bin.v2/libs/test/ build/sun-5.9/debug/link-static/unit_test_main.o" "../../../libs/test/ src/unit_test_main.cpp" The patch adds a using declaration to bring the base class function into the overload set. Sun's interpretation of the standard seems to be correct, from what I can tell. Here's their rational: [ from their sun-5.9 FAQ ] Q. Why does the compiler say that a derived-class virtual function hides a base-class virtual function with a different signature? My other compiler doesn't complain about the code. A. The C++ rule is that overloading occurs only within one scope, never across scopes. A base class is considered to be in a scope that surrounds the scope of a derived class. Any name declared in a derived class therefore hides, and cannot overload, any function in a base class. This fundamental C++ rule predates the ARM. If another compiler does not complain, it is doing you a disservice, because the code will not behave as you probably expect. Our compiler issues a warning while accepting the code. (The code is legal, but probably does not do what you want.) If you wish to include base-class functions in an overloaded set, you must do something to bring the base-class functions into the current scope. If you are compiling in default standard mode, you can add a using-declaration: Okay to commit to trunk? -- Noel

2 1

Boost with comeau c++
by Sean Farrow 14 Sep '08

14 Sep '08

Hi: Does anybody have any experience/information as to how well boost compiles with comeau c++? Any information would be gratefully apreciated. Cheers Sean. __________ Information from ESET NOD32 Antivirus, version of virus signature database 3440 (20080913) __________ The message was checked by ESET NOD32 Antivirus. http://www.eset.com

2 1