- Boost-users - lists.boost.org

[lexical_cast] float-to-string rounding error (not found with cout)
by Grant Birchmeier 30 Apr '09

30 Apr '09

I'm seeing unexpected behavior when using lexical_cast<string>. This code: const float f = 00123.12300; cout << f << endl; cout << boost::lexical_cast<std::string>(f) << endl; gives the result: 123.123 123.123001 Two questions: (1) Is there a way to eliminate this trailing '1' with lexical_cast (presumably some wierd rounding error)? (2) When casting float-to-string, given this behavior, why should I use lexical_cast instead of, say, stringstream? Thanks -Grant

4 5

Default argument values with boost::function
by Artyom 30 Apr '09

30 Apr '09

Hi, Is there any way to set default argument values with Boost::Function? On VS2008 SP1 I tried: boost::function< void( unsigned i = 1 ) > increment; But this didn't work. Code such as increment( 1 ); compiles OK but if you try doing increment(); the following compiler error is triggered: error C2064: term does not evaluate to a function taking 0 arguments class does not define an 'operator()' or a user defined conversion operator to a pointer-to-function or reference-to-function that takes appropriate number of arguments Thanks, Artyom

5 4

[boost-users][python] Pointer argument failed conversion?
by Germán Diago 30 Apr '09

30 Apr '09

Hello. After much googling I didn't find anything useful. My problem is: I have a virtual function like this class MyClass { void monitor(Object * o); } I have wrapped another class, which inherits from Object: class Character : public Object {...}; which is also exposed to python. I register my virtual function as the documentation says I have to do: class_<MyClass, boost::noncopyable>("MyClass"). def("monitor", &MyClass::monitor); When I try to call it from python, I get the following error: a = MyClass() b = Character(....) a.monitor(b) MyClass.monitor(MyClass, Character) did not match C++ signature: monitorizar(MyClass {lvalue}, Object *) Shouldn't my Character be automatically converted to Object * and then to make a successful call to the function? Should I do anything special to use pointers as parameters of my functions? I don't know why it tries to pass a "Character". What I want is to get it converted to Object *. Thanks in advance.

2 1

Program options required arguments?
by David Doria 30 Apr '09

30 Apr '09

If I have a list of 10 parameters, and all are required, I end up with a giant block of these: if(!vm.count("input")) { std::cout << "Input was not set." << std::endl; exit(-1); } Is there a way to do something like: desc.add_options() ("input", po::value<std::string>(&InputFile)->required(), "Set input file.") That will perform that check for me? Thanks, David

2 1

[interprocess]message_queue multi-client
by Laurence Chambenois 30 Apr '09

30 Apr '09

Hi, I need to communicate from x clients processes to one server process. I want to use the message_queue to send messages from the clients to the server, and when the server is done, it sends back its answer to the client. Do you have any examples ? with one bidirectionnal message_queue, or do I have to create a message_queue for each "direction" ? How can I deal with one message_queue and multiple process so that the request sent to one process is read by the good process ? Thanks for your help Laurence

1 0

Boost (1.38) on Intel Compiler (11.0)
by Jared White 30 Apr '09

30 Apr '09

Hey there, Was curious if someone can answer for me, does Boost support the intel compiler? I seem to be having some trouble linking to the libraries that are created when I "bjam" on Mac OSX using the Intel compiler. It builds the libraries as you would expect, and a whole bunch of them get dropped into the "lib" folder. In VC++, I give the linker this path, and it finds the libs, just fine. However, When I compile my application (with XCode), it can't find the libs. It would appear that the <boost/auto_link.hpp> is configuring the lib name to have a toolset of "iw" and not "il", as expected for Mac OS X. Any help is appreciated, -Jared

2 1

German Boost book updated
by Boris Schaeling 30 Apr '09

30 Apr '09

For those who know German: I've added five more chapters to my Boost book at http://www.highscore.de/cpp/boost/. There are now altogether 16 chapters covering 29 Boost libraries. The book does not try to replace the documentation but is written in a tutorial style to give developers a kick start. It explains what the numerous libraries are good for and how they are used without going too much into detail. Any feedback is, as always, appreciated! Boris

3 2

Multi Index: Nested std::pair
by Etienne Philip Pretorius 30 Apr '09

30 Apr '09

Hello List, I would greatly appreciate if someone could assist me in getting this structure correct. boost::multi_index_container< std::list< std::pair< /*value*/ unsigned char, std::pair< /*x co-ordinate*/ unsigned char, /*y co-ordinate*/ unsigned char > > >, boost::multi_index::indexed_by< /*value*/ boost::multi_index::ordered_non_unique< boost::multi_index::identity<unsigned char> >, /*x co-ordinate*/ boost::multi_index::ordered_non_unique< boost::multi_index::identity<unsigned char> >, /*y co-ordinate*/ boost::multi_index::ordered_non_unique< boost::multi_index::identity<unsigned char> > > > m; I do not see how "boost::multi_index::ordered_non_unique" will index the appropriate field in the nested std::pair structures. This is why I suspect there to be a problem with this structure. So before I continue, could someone please tell me if I am on the correct path. Grateful, Etienne

7 24

[Serialization] Getting crash in serialization library specific to Linux and 1.37
by Reusser, Edward 29 Apr '09

29 Apr '09

I created several DLL's that save and restore data using the serialization library. These libraries can be used on both the Windows and Linux and have been quite reliable since they were first created using the 1.34 boost release. The purpose of these libraries is to read and write serialized versions of our projects. We are using both XML and Text archives. The problem I am now looking at is in reading a Text archive. We recently upgraded to boost release 1.37. On Windows we are still using the 7.1 compiler, while on Linux we are using the gcc 3.2 compiler. We had previoously upgraded compilers without any problems, so its not a compiler upgrade issue. Upon upgrading to boost release 1.37, everything has continued to work just fine on Windows. However in the Linux builds, any designs written with the older 1.34 boost release now cause segmentation faults when being read with the more recent build. I have attached a stack trace. To complicate this picture, I can take a project written using the older 1.34 release on Linux, read it in Windows, then write it using Windows, then read it just fine in the new Linux build. Any ideas? Edward Reusser Principal Engineer Actel Corporation (650) 318-4972 The information contained in or attached to this email may be subject to the Export Administration Regulations (EAR), administered by the U.S. Department of Commerce, or the International Traffic in Arms Regulations (ITAR), administered by the U.S. Department of State, and may require an export license from the Commerce or State Department prior to its export. An export can include a release or disclosure to a foreign national inside or outside the United States. Include this notice with any reproduced portion of this information.

1 0

Cannot find -lboost_thhread with upgrade from 1.35 to 1.37
by Johan Mazel 29 Apr '09

29 Apr '09

Hi I just upgraded my OS from Ubuntu 8.10 to 9.04 and Boost upgraded at the same time from 1.35 to 1.37. I get some errors linked to the compilation command with g++ : /usr/bin/ld: cannot find -lboost_thread /usr/bin/ld: cannot find -lboost_filesystem These compilation options were working with the 1.35 version of Boost. I'd like to know wether there has been a change in Boost concerning the compilation options or if anything else have changed that would cause that problem. Thanks in advance for the answers. Johan Mazel

4 6