- Boost - lists.boost.org

program_options: informative exceptions
by Llew Sion Goodstadt 17 Sep '04

17 Sep '04

Dear Volodya, The version of Boost.program_options I am using throws lots of informative exceptions to indicate precisely what has gone wrong. The exception is where an incorrect argument to an option has been supplied e.g. --some_arg "abc" where --some_arg 0.6 is expected. The error message is annoyingly non-descript and it can be difficult to figure out which particular optional argument has been mis-specified. This imprecision is because when program_options::validation_error is being thrown, it is not clear which option is being parsed. Could I suggest that this particular exception is caught in: void store(const parsed_options& options, variables_map& xm, bool utf8) and the option context added back in (and the same exception type re-thrown): void store(const parsed_options& options, variables_map& xm, bool utf8) { ... // First, convert/store all given options for ( size_t i = 0; i < options.options.size(); ++i ) { const string& name = options.options[i].string_key; try { ... } catch ( validation_error& e ) { throw validation_error("[--" + name + "]: " + e.what()); } } ... } What do you think? Leo Goodstadt MRC Functional Genetics Unit University of Oxford

2 4

problem with intersection test_tools and mpl/list.hpp and borland
by Robert Ramey 17 Sep '04

17 Sep '04

OK - I've followed this farther down a very long chain Here is as deep as I'm in a position to go: Anytime this pair of headers is included either directly or indirectly, and the compiler is borland, you'll get a syntax error: #include <boost/mpl/list.hpp> #include <boost/next_prior.hpp> Robert Ramey

1 0

[serialization 1.32]: vc7.1 : link error with free_saver
by Francis ANDRE 16 Sep '04

16 Sep '04

Hi There is a link error with boost::serialization::free_saver in the serialization 1.32 lib under vc7.1. The workaround put in place for vc7.0 (BOOST_MSVC 1300) should also apply to vc7.1 (BOOST_MSVC 1310) Index: split_free.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/serialization/split_free.hpp,v retrieving revision 1.4 diff -u -r1.4 split_free.hpp --- split_free.hpp 4 Aug 2004 05:23:40 -0000 1.4 +++ split_free.hpp 16 Sep 2004 22:17:23 -0000 @@ -34,7 +34,7 @@ namespace serialization { -#if ! BOOST_WORKAROUND(BOOST_MSVC, <= 1300) +#if ! BOOST_WORKAROUND(BOOST_MSVC, <= 1310) template<class Archive, class T> void save(Archive &ar, const T & t, const unsigned int); template<class Archive, class T> Regards. Francis ANDRE

1 0

RE: [boost] Re: problem with intersection test_tools and mpl/list .hpp andborland
by Rozental, Gennadiy 16 Sep '04

16 Sep '04

How do I manage to step in all these compiler bugs??;)) I am quite sure I do not have nor function nor class template named next. How does it happened? Gennadiy > -----Original Message----- > From: boost-bounces(a)lists.boost.org > [mailto:boost-bounces@lists.boost.org] On Behalf Of Pavel Vozenilek > Sent: Thursday, September 16, 2004 2:42 PM > To: boost(a)lists.boost.org > Subject: [boost] Re: problem with intersection test_tools and > mpl/list.hpp andborland > > > > "Robert Ramey" wrote: > > > The following program: > > > > > > #include <boost/mpl/list.hpp> > > #include <boost/test/test_tools.hpp> > > > > fails to compile with borland compilers with the messages: > > > > c:\BoostMainCVS\libs\serialization\test\test_zmisc.cpp: > > Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 30: Cannot > have both > > a template class and function named 'next' > > > > > Old BCB bug, e.g. > http://groups.google.com/groups?hl=en&lr=&ie=UTF-8&c2coff=1&th readm=3BDD798D.130E38F8%40divalsim.it&rnum=1&prev=/groups%3Fnum%3D100%26hl%3 Den%26lr%3D%26ie%3DUTF-8%26c2coff%3D1%26q%3DE2479%2BBorland%2BOR%2BBCB%2BOR% 2Bcompiler%26btnG%3DSearch (http://tinyurl.com/5ltk9) It won't be fixed, ever. One ugly way is to use macro: #include <boost/mpl/list.hpp> #define next xyz #include <boost/test/test_tools.hpp> #undef next other is to separate code into two TUs. /Pavel _______________________________________________ Unsubscribe & other changes: http://lists.boost.org/mailman/listinfo.cgi/boost

2 1

Re: problem with intersection test_tools and mpl/list.hpp andborland
by Robert Ramey 16 Sep '04

16 Sep '04

Pavel Vozenilekwrote: >> The following program: >> >> >> #include <boost/mpl/list.hpp> >> #include <boost/test/test_tools.hpp> >> >> fails to compile with borland compilers with the messages: >> >> c:\BoostMainCVS\libs\serialization\test\test_zmisc.cpp: >> Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 30: Cannot have both >> a template class and function named 'next' >> >> >Old BCB bug, e.g. > (http://tinyurl.com/5ltk9) >It won't be fixed, ever. >One ugly way is to use macro: >#include <boost/mpl/list.hpp> >#define next xyz >#include <boost/test/test_tools.hpp> >#undef next >other is to separate code into two TUs. I don't think this will work for me as it shows up due to file header inclusion about 7 levels deep. Note that this is a new phenomenon that I believed was introduced with the latest MPL update. I realize that's it's a lot, but would it possible for the authors of the two libraries to agree that the name be changed in one of them just so I that us users don't have to resort to a hack such as the above? Robert Ramey

1 0

Re:[boost] Serializing bitsets and tuples
by Robert Ramey 16 Sep '04

16 Sep '04

>Again, Robert, I have to thank you immensely for the Serialization >library. It has made my life much easier. >Anyway, I was wondering, since you state in the manual that all STL >containers are supported, why I can't seem to find std::bitset in that. >I realize that it is probably not the most common container to >serialize, but still. Also, considering the is Boost library now, I'm >wondering about the apparent absence of serialization for boost::tuple. >It's not the end of the world (I can write them myself rather easily, I >suppose), but for completeness I'd think at least Boost.Tuple should be >supported. LOL - I never heard of std::bitset until you mentioned it. Also I think I overlooked vector<bool> as well. Feel free to implement serialization for them and put your name on it. Make a test and we'll add it in. Robert Ramey

3 3

bug in recursive_mutex
by James Ahlborn 16 Sep '04

16 Sep '04

I previously posted about issues when using a recursive_mutex and a condition. after some more inspection, it turns out it is a minor bug in the recursive_mutex implementation (the pthread version). I've included the patch below, which was generated against the 1.31.0 version of boost (but the bug is still in the latest cvs as well). Would be great if someone could integrate this into the latest code, or possibly a bug fix release for 1.31 (this is a major bug for anyone using recursive_mutex with condition). thanks, -james *** boost_1_31_0/libs/thread/src/recursive_mutex.cpp Fri Dec 26 18:22:00 2003 --- boost_1_31_0.jta/libs/thread/src/recursive_mutex.cpp Thu Sep 16 10:01:10 2004 *************** *** 412,417 **** --- 412,418 ---- state.pmutex = &m_mutex; state.count = m_count; + m_count = 0; } recursive_try_mutex::recursive_try_mutex() *************** *** 608,613 **** --- 609,615 ---- state.pmutex = &m_mutex; state.count = m_count; + m_count = 0; } recursive_timed_mutex::recursive_timed_mutex() *************** *** 787,792 **** --- 789,795 ---- state.pmutex = &m_mutex; state.count = m_count; + m_count = 0; } #elif defined(BOOST_HAS_MPTASKS)

3 3

problem with intersection test_tools and mpl/list.hpp and borland
by Robert Ramey 16 Sep '04

16 Sep '04

The following program: #include <boost/mpl/list.hpp> #include <boost/test/test_tools.hpp> fails to compile with borland compilers with the messages: c:\BoostMainCVS\libs\serialization\test\test_zmisc.cpp: Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 30: Cannot have both a template class and function named 'next' Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 34: Cannot have both a template class and function named 'next' Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 40: Cannot have both a template class and function named 'prior' Error E2479 C:\BoostMainCVS\boost/next_prior.hpp 44: Cannot have both a template class and function named 'prior' *** 4 errors in Compile *** Could someone shed some light on this? Robert Ramey

2 1

Re:Re: [boost] Serializing bitsets and tuples
by Robert Ramey 16 Sep '04

16 Sep '04

>> I'll see if I can get a test case thrown together for std::bitset. >> I'll take a look at boost::tuple as well (but that should be big time >> trivial, so... :) ) >Erh... trivial, but tedious. Seeing as I have to redo the serialize >funtion for all possible lengths of a tuple (0 to 10 contained >elements). Unless I missed something obvious? (again) >Anyway, is serialization for tuples supposed to be a Boost.Serialization >add-on or a Boost.Tuples add-on? As Jeff Garland mentioned, generally >std::* should be in Serialization, and anything else in the respective >library. However, Boost.Serialization already provides support for e.g. >shared_ptr<> and Boost.Optional. So...? This question has come up before without being explicitly/publicly agreed upon. Jeff Garland and I have come to the following consensus: Serialization for standard containers and objects, ie members of namespace std:: will go into the boost::serialization namespace and their corresponding test will go into the serialization test library. Other serialization implementations will go into the namespace/test set of their respective libraries. All maintainence of serialization modules will be responsibility of their respective authors. There are a few boost classes in the serialization library which ended up there during development of the library. shared_ptr is and example. These are anomalous cases and not a pattern that will be continued. Obviously, its my desire to keep the serialization library as task of O(1000) rather than O(1000 + N). So that's our policy. Robert Ramey

1 0

State of giallo, (sockets and demultiplexing)
by Hugo Duncan 16 Sep '04

16 Sep '04

Spend a few days off line and you suddenly miss a whole conversation :-( It's going to take a while to catch up... Recent IOstreams discussion has indeed been very interesting, espescially as I am finally getting round to implementing buffers and filters for giallo (boost.sockets). At the moment I am leaning towards ACE style message buffer, using the apache filter idea of flush/eos/eoc control (but trying to avoid any knowledge of the underlying resource). So where is giallo today (apart from being temporarily broken on gcc due to mpl changes)? i) Low level, platform header hiding, object based wrappers of sockets, named pipes, select, and completion ports (and a POSIX aio wrapper on linux that I have tried to get to work in many different ways and failed so far). ii) proactor and reactor interfaces, with policy based implementation, provided by bridge classes to the low level completion port and select wrappers above. iii) 'event completion notification' style connector, acceptor and connection classes for sockets and named pipes. These classes combine a demultiplexer with a resource (socket, pipe, etc), and can use a proactor or reactor demultiplexer interchangeably (via template parameter). The connection selects the synchronisation mode (block, non-blocking, asynchronous) of the socket based the demultiplexer. The library is thread aware (can run multithreaded, will create threads if asked to, uses threads to do asynchronous connect if not natively supported, etc) There is a streambuf implementation, but I have not found a way to implement this without it degenerating to a thread per connection model. I feel quite strongly that the decision to implement connection, connector and acceptor concepts with a single (in this case 'event completion' style) notification interface, regardless of the demultiplexer mechanism used, is very important, and can lead to a stream/filter concept that is independent of demultiplexer, of blocking/non-blocking/asynch mode and thus independent of the resource used.

5 12