October 2013 - Boost-users - lists.preview.boost.org

Fwd: support for bjam projects in IDE
by Witold E Wolski 06 Dec '13

06 Dec '13

Hi, I will have to work on bjam based project (rather big). I normally use QT Creator for programming and work with cmake based project. I could't find any hint how to set up an bjam project to work with QT neither information about any IDE which supports bjam. Do boost developer use vim and debug using gdb only? Actually the other guys on the project wrote some python scripts which parse bjam output to generated vcproj files :-@ I would be happy to hear from someone who are happy using bjam with qtcreator or any other IDE for larger projects and how to set this up. All my searches so far regarding bjam return almost exclusively FUD (http://en.wikipedia.org/wiki/Fear,_uncertainty_and_doubt) but looking at the Jamroot.jam file of the project (more than 1000 lines) I am inclinded to believe it. ;-( -- Witold Eryk Wolski -- Witold Eryk Wolski

2 2

Boost.Log: Set formatter for named_scope attr.
by David Puffer 14 Nov '13

14 Nov '13

Hi, Being new to Boost.Log (and Boost in general) I have spent the last few days trying to implement Boost.Log - which Is working fairly well by now. However - I encountered one problem that I've not been able to solve by now, although it seems to me quite trivial: I have successfully registered the "named_scope" attribute and I am populating its value by a call to BOOST_LOG_FUNCTION. This works well and I can see the function-name in my log-output. However - I also want the line-number and the file-name, and discovered that there's a possibility to set the output-format of the "named_scope" attribute (which already incorporates file-name and line-number info) like so: #include <boost/log/expressions/formatters/named_scope.hpp> sink->set_formatter(expr::stream << expr::format_named_scope("Scopes", "%n")); As documented here: http://www.boost.org/doc/libs/1_54_0/libs/log/doc/html/log/detailed/express… I just don't seem to find any way to get to my sink - which is configured in a settings-file and evaluated by a call to the init_from_stream() function. Is there any way to query for all the sinks - or get a sink by name? How is one supposed to set a sink formatter for a sink which was generated by the library and not the user? Or is there a way to specify the output-format for named attributes via a settings-file as well? As far as I can see it's only possible to format the output of a log record as a whole - but not single attributes. Also: Is it correct, that the above function will only set the format for the "Scopes" attribute - which means I could call this function repeatedly for lets say the "TimeStamp" attribute as well? Best regards, David

2 2

[LOG]: set_filter error in tutorial_filtering.cpp and per-sink instance filtering
by Johan Lundberg 05 Nov '13

05 Nov '13

Dear all, (this is my first post to this list) I'm trying to make a boost log sink that would filter on a severity level which should be thread-safely modifiable while logging is ongoing. On my way there I get errors while trying to compile this tutorial example: http://www.boost.org/doc/libs/1_54_0/libs/log/example/doc/tutorial_filterin… The file as-is compiles and runs fine, but if I switch on the user-provided filtering function example by activating the block in #if 0 (and commenting out the other definition of init()) I get errors related to phoenix bind: The offending line is 155: sink->set_filter(phoenix::bind(&my_filter, severity, tag_attr)); I would be grateful for advice on how to get that working. More details on that at the end of the mail. Now, assuming I get past the errors, what is a good way to add a per-sink log level filter? I may have several instances of the sink. To clarify: I know this is not the right syntax, but If I could use a custom functor or a C++11 lambda for a filter I would do something like this: using ATseverity=std::atomic<severity_level> ; // (severity is an enum) shared_ptr<ATseverity> shared_log_level = std::make_shared<ATseverity>(warning); sink->set_filter( [=shared_log_level]->bool(severity_level const& level, std::string const& tag){ return level >= shared_log_level->get(); }); I could then store and modify the shared_log_level from elsewhere. Is there a right way to do this? cheers /Johan Details on the compilation error: I think this is the relevant part of the error message: /myboost154/include/boost/phoenix/core/preprocessed/expression_10.hpp:140:17: error: could not convert ‘boost::proto::make_expr<boost::phoenix::detail::tag::function_eval, boost::proto::domainns_::basic_default_domain, boost::phoenix::detail::function_ptr<2, bool, bool (*)(const boost::log::v2_mt_posix::value_ref<severity_level>&, const boost::log::v2_mt_posix::value_ref<std::basic_string<char> >&)>, boost::log::v2_mt_posix::expressions::attribute_keyword<tag::severity>, boost::log::v2_mt_posix::expressions::attribute_keyword<tag::tag_attr> >((* & a0), (* & a1), (* & a2))’ from ‘const type {aka const boost::proto::exprns_::basic_expr<boost::phoenix::detail::tag::function_eval, boost::proto::argsns_::list3<boost::proto::exprns_::basic_expr<boost::proto::tagns_::tag::terminal, boost::proto::argsns_::term<boost::phoenix::detail::function_ptr<2,bool, bool (*)(const boost::log::v2_mt_posix::value_ref<severity_level>&, const boost::log::v2_mt_posix::value_ref<std::basic_string<char>>&)> >, 0l>, boost::log::v2_mt_posix::expressions::attribute_keyword<tag::severity>, boost::log::v2_mt_posix::expressions::attribute_keyword<tag::tag_attr> >, 3l>}’ to ‘boost::proto::exprns_::basic_expr<boost::proto::tagns_::tag::terminal, boost::proto::argsns_::term<boost::phoenix::detail::function_ptr<2, bool, bool (*)(const boost::log::v2_mt_posix::value_ref<severity_level>&, const boost::log::v2_mt_posix::value_ref<std::basic_string<char> >&)> >, 0l>::proto_child0 { aka boost::phoenix::detail::function_ptr<2, bool, bool (*)(const boost::log::v2_mt_posix::value_ref<severity_level>&, const boost::log::v2_mt_posix::value_ref<std::basic_string<char> >&)>}’ }; I tried boost 1.54 and compiled with gcc 4.7.1 and 4.8.1 under Linux, with and without C++11 mode: g++ -std=gnu++11 -DBOOST_ALL_DYN_LINK boostlogexample.cpp -L $BOOST/lib/ -I $BOOST/include/ -pthread -lboost_system -lboost_log_setup -lboost_log \ -lboost_date_time -lboost_thread -lrt -lboost_filesystem

1 1

container with large number of elements and a small number of invalide elements
by MM 05 Nov '13

05 Nov '13

Hello, I have a need for a container where I can store 10^6 elements of type T (PODs) , a small number of which (say 100) can be invalid elements (ie special value of T) The 100 elements can be randomly distributed throughout the 10^6 elements. The container should be as close to a sequence container as possible. The required usages: 0. invalid elements can be set = 10^6 - 100 1. get number of valid elements in O(1) 2. get the max/min element of all valid elements 3. have a bidrectional iterator that iterators only over valid elements On a separate note, is there a version of multi_array where the number of dimensions is known at runtime? regards,

6 17

[Spirit] Parser omits certain characters
by beet 05 Nov '13

05 Nov '13

Dear all, For all of the following, please see the attached test-case. I would like to specify certain characterstics of variables in the following way: d(MY_DPAR_01,-10.3,12.8,100) . So in this case there is a type-identifyer (d stands for "double") and then, in parantheses, a variable name (or alternatively an index), a lower and upper boundary and some integer parameter. What is contained in the list depends on the parameter type. It should then be possible to specify any number of variables, including their properties, in a comma-separated list. So I need to dissect a string like the following: "d(MY_DPAR_01,-10.3,12.8,100), d(0,-10.3,12.8,100), i(SOME_IPAR_17, 0,5), b(SOME_BPAR)" The first step, as I see it, is to seperate the "outer" list and to store each variable description in a boost::tuple<char, std::string>, where the char holds the type identifier and the std::string the part inside the parantheses. Then, in the second step, I want to parse that string according to what is expected for this particular type. I have now run into the problem that the following rule in Boost.Spirit "swallows" all non-alphanumeric characters. qi::rule<std::string::const_iterator, std::string(), ascii::space_type> varSpec = +(alnum | '_' | ',' | '.' | '+' | '-'); varSpec stands for the text inside of the parantheses. I would have expected this to match any string of at least one of the listed character types (i.e. alpha-numeric, underscore, etc.). The rule qi::rule<std::string::const_iterator, boost::tuple<char, std::string>(), ascii::space_type> varString = char_("dfib") >> '(' >> varSpec >> ')'; is then meant to separate the type-descriptor from the description. Parsing with these yields the output d: MYDPAR01103128100 d: 0103128100 i: SOMEIPAR1705 b: SOMEBPAR on a Mac (Mavericks, Boost 1.54) i.e., only the alpha-numeric characters remain. So clearly my assumption about the "varSpec" rule is wrong. On a Ubuntu 13.10 System (Boost 1.54, g++ 4.8.1) I get the output d: d: i: b: which is equally wrong. Funnily, if I define varSpec simply as +char_, varSpec gives the correct result on the Mac (haven't tried Ubuntu), but varString does not match anything. So I am at a loss here and would appreciate some help. Best Regards and thanks for any help you can provide, Beet

3 4

Status of smart_assert
by Daan Baas 03 Nov '13

03 Nov '13

Howdy, I was searching for an intelligent assert facility with serveral debug levels, that asserts when it is run debug mode and throws in release mode. That is how i ran into Boost's SMART_ASSERT ( http://www.drdobbs.com/cpp/enhancing-assertions/184403745). However, I cannot find anything about it that was posted after 2003 nor can I find it on the Boost site. Does anyone know what happened to the project? Was is cancelled or did it make it into Boost under another name? Kind regards, Dano

2 2

[MSM] Parse error in Visual Studio 2012, not in Visual Studio 2013.
by Vincent N. Virgilio 02 Nov '13

02 Nov '13

Hello, In the code below, VS 2012 (target v110) underlines the macro name BOOST_MSM_EUML_EVENT_WITH_ATTRIBUTES in the last line with a wavy red line. Hovering a mouse over it shows a tooltip with error report: "Error: not a class or struct name" And if I recall correctly, the code does not compile. The error disappears in VS 2013 (v120). Though in VS 2013 (target v120), if I remove line "namespace msm = boost::msm;", the last line triggers an error again. This time with a different tooltip error report: "Error: name followed by '::' must be a class or namespace name." Am I using MSM incorrectly? Thanks, Vince // BEGIN euml.h #include <boost/msm/back/state_machine.hpp> #include <boost/msm/front/state_machine_def.hpp> #include <boost/msm/front/euml/euml.hpp> using namespace boost::msm::front::euml; namespace msm = boost::msm; BOOST_MSM_EUML_DECLARE_ATTRIBUTE(char*, charp_attr) BOOST_MSM_EUML_ATTRIBUTES((attributes_ << charp_attr), found_attrS) // next line gives error in VS 2012 (target v110), "Error: not a class or struct name". Not in VS 2013 (targ. v12). BOOST_MSM_EUML_EVENT_WITH_ATTRIBUTES(found, found_attrS) // END euml.h

3 3

[1.55.0] Release candidates available
by Marshall Clow 01 Nov '13

01 Nov '13

Release candidate files for 1.55.0 are available at http://boost.cowic.de/rc/ As always, the release managers would appreciate it if you download the candidate of your choice and give building it a try. Please report both success and failure, and anything else that is noteworthy. This helps ensure the candidates build OK before we push them out to SourceForge. The files (and associated md5s) are: ba1ae723b8b4d1edeca684bb14b3213b *boost_1_55_0_rc1.tar.gz 66e01be2f328c556835c42be7b6d17d5 *boost_1_55_0_rc1.tar.bz2 925bb6c96082033687e5742b4e77af7b *boost_1_55_0_rc1.zip 830d61075ecf7d5ebf7c0843033393bf *boost_1_55_0_rc1.7z Thanks! -- The release managers

9 16

[python] Crash right after initialization after ubgrading from 1.48 to 1.54
by Richi Lists 31 Oct '13

31 Oct '13

Now I * downloaded the python 2.75 source * compiled python 2.75 * compiled boost 1.54 against python 2.75 * built our software against these new libs -> The behavior is still the same. ob_type of the first python object I create is 0xffffffff, and trying to access it results in an immediate crash. This only applies to the debug version. Release is fine. Also, the behavior is the same, if I link the boost debug build against python release or debug. Rgds Richard

1 0

[Iostreams] Problem to chain gzip filter with home-made one
by Christophe hugon 31 Oct '13

31 Oct '13

Hello everybody, I meeting a execution issue while using the gzip filter of the Boost.Iostreams, combined with a buffered home-made one. I would like to know what is wrong in my way to do it. Firstly, my filter is a general DualFilter. It uses an buffer (of 16 bytes) before the transformation and the put/return. The categories of my filter is closable and dual_use. You can find code details in the end of this mail. My problem is ([a/b] means for the separates case with the letter a and b): If I take the 6 possibilities for write and read: 1) io::filtering_[i/o]stream stm; 2) stm.push(io::gzip_[de/]compressor()); 3) stm.push(my_[Unt/T]ransFilter); at the moment to use the stm.[get/put] to read/write 1)-2) works perfectly 1)-3) works perfectly 1)-3)-2) works perfectly 1)-2)-3) Stay stuck (obviously it is the one I really need) In fact in the last case, the gzip filter give this infinite loop to the put of the TransFilter : 31 -117 8 0 0 0 0 0 0 31 -117 8 0 0... A similar case occur with the bzip filter (even if it is a bit longer loop). What do I do wrong ? Somebody can help me ? Thanks in advance Christophe PS: here a simplified implementation of my filter template<typename Source> int get(Source& src) { //at the very beginning or if the buffer are fully returned if (fcurrent_pos == fblkLength) { fcurrent_pos =0; FillThem (src); //fill&transform the buffers with the right number of src.get() fc=CheckNSolveBufferIsEnd (src); //eventually prepare read the suffix of the file } //The effective return if (fcurrent_pos < fUntrans.size ()) return (int)fUntrans[fcurrent_pos++]; else return fc; } template<typename Sink> bool put(Sink& dest, int c) { if (c==EOF || c ==boost::iostreams::WOULD_BLOCK) return false; if (fUntrans.size () < fblkLength - 1) { fUntrans.AddData((char*)(&c),1); return true; } fUntrans.AddData((char*)(&c),1); fenc->TransformThis (fUntrans, fTrans, fKey); for (int i=0; i<fblkLength; i++) { if (!boost::iostreams::put(dest, (int)fTrans[i])) return false; } return true; } template<typename Sink> bool close (Sink &dest, BOOST_IOS::openmode m) { if (fFlushed) return true; PrepareSuffix (); for (int i=0; i< fblkLength*2; //suffix length i++) { if (!boost::iostreams::put(dest, (int)fTrans[i])) return false; } return true; }

1 0