- Boost-users - lists.boost.org

gcc3.2 preprocessor concatenation or not on negative numbers
by Andy Little 21 Jan '06

21 Jan '06

Hi all, Using gcc3.2 the concatenation of 1##Num where Num is negative leaves a space eg 1e#-1 --> 1e -1 as can be seen by the preprocessed output below. The only way round it that I can see is to simply not use a macro for this compiler Is there a preprocessor workaround for this problem? Is this even the correct behaviour? Any help appreciated. #define POW_EVAL(Num, Float_type)\ template <>\ struct pow_eval< \ Float_type BOOST_PP_COMMA() Num BOOST_PP_COMMA()\ 1 BOOST_PP_COMMA() false\ >{\ enum{ required = true};\ typedef Float_type result_type;\ result_type operator()()const\ { result_type result\ = static_cast< Float_type > (1e ## Num);\ return result;\ }\ }; POW_EVAL(-60, double) POW_EVAL(-59, double) POW_EVAL(58, double) POW_EVAL(57, double) // preprocessor output template <> struct pow_eval< double , -60 , 1 , false >{ enum{ required = true}; typedef double result_type; result_type operator()()const { result_type result = static_cast< double > (1e- 60); return result; } }; template <> struct pow_eval< double , -59 , 1 , false >{ enum{ required = true}; typedef double result_type; result_type operator()()const { result_type result = static_cast< double > (1e- 59); return result; } }; template <> struct pow_eval< double , 58 , 1 , false >{ enum{ required = true}; typedef double result_type; result_type operator()()const { result_type result = static_cast< double > (1e58); return result; } }; template <> struct pow_eval< double , 57 , 1 , false >{ enum{ required = true}; typedef double result_type; result_type operator()()const { result_type result = static_cast< double > (1e57); return result; } }; regards Andy Little

2 5

regex format string extension for more than nine captures
by Evan Burkitt 21 Jan '06

21 Jan '06

I need to be able to reformat match_results that contain more than nine captures using format strings. I'm looking for advice as to the easiest way to accomplish this. Is there some existing work I can build on or borrow from? I need syntax similar to the boost extended syntax, possibly using two digits for the capture number, e.g., '$01' It's important to me to be able to use the extended syntax's conditional expression '?Ntrue-expression:false-expression', again probably with two digits 'NN' in place of the single 'N'. Any help would be greatly appreciated. -eb-

2 1

[tuple] no easy access to get function
by Frank Birbacher 21 Jan '06

21 Jan '06

Hi! Today I wanted to std::transform a range of tuples into a sequence of one element. I thought I could just take the address of some tuple::get<N> function. Maybe also specify the Tuple type as template parameter. But from the source I figured the get function is not designed in a way that would allow me easy access. I had to write myself an additional function for this task (see below). Am I trying the wrong way? Or would you fix the implementation? template<size_t N, typename Tuple> typename boost::tuples::element<1,Tuple>::type const& getTupleElement(const Tuple& tuple) { return boost::get<N>(tuple); } to do: transform(..., ..., ..., &::getTupleElement<1,Tuple>); or //using boost::lambda for_each(..., ..., var(i) += bind(&::getTupleElement<1,Tuple>, _1)); I'd rather want to take the address of boost::tuple::get<N,Tuple>, but the template parameters won't match. Frank

1 0

Parse errors when building against Insure++ by ParaSoft.
by Sven Delmas 20 Jan '06

20 Jan '06

I am trying to do some memory allocation debugging using Insure++ on my Windows machine. They use their own parser, and inject all kinds of fun stuff. However, their parser appears to be more restrictive (or just plain different) than the usual parsers. I end up with two parser errors: One in boost/details/indirect_traits.hpp (line 254 in boost 1.33.1): BOOST_STATIC_CONSTANT( bool, value = sizeof(detail::is_function_ref_tester(t,0)) == sizeof(::boost::type_traits::yes_type)); The error is: namespace "boost::detail" has no member is_function_ref_tester I can resolve that by taking the detail:: away, which seems to make sense if it indeed not part of the namespace. Can anyone confirm whether that is a valid solution/issue? The second issue is in boost/pool/detail/ct_gcd_lcm.hpp (line 68 in boost 1.33.1): BOOST_STATIC_CONSTANT(unsigned, value = ct_gcd2<B>::helper<A % B>::value); The error is: the global scope has no value I can resolve this by switching to the other branch with respect to the macro BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION. That compiles fine for VC 7.0, VC 7.1 and VC 8.0, so my question is whether it will produce the correct result, and if so, whether the visualc.hpp should be adjusted for those compiler. Any help would be appreciated. Thanks, Sven

1 0

Serialization "stream error" when using binary_iarchive/binary_oarchive & make_binary_object, but xml_iarchive & text_iarchive suceeds
by Terence Wilson 20 Jan '06

20 Jan '06

I've been scratching my head trying to figure this one out for a few days. My code serializes xml_iarchive & text_iarchive with no problems, however, when I substitute the binary_iarchive I get a "stream error" serialization exception thrown during the make_binary_object call when reading a binary file: template<class Archive> void serialize(Archive & ar, const unsigned int) { ar & BOOST_SERIALIZATION_BASE_OBJECT_NVP(CImageDataBase2D); if (Archive::is_loading::value) { Alloc(); } //ar & boost::serialization::make_nvp("ImageData", boost::serialization::make_binary_object(m_pData, sizeof(T)*m_x*m_y)); const int blobSize=sizeof(T)*m_x*m_y; _ASSERTE(blobSize); ar & boost::serialization::make_nvp("ImageData", boost::serialization::make_binary_object(m_pData, blobSize)); } I've tried to excise the relevant code to a console project, however, it has tentacles and will take a few more days. In the mean time I would appreciate help from anyone that might have a suggestion. I've looked at the code and can find nothing wrong, as I said, ASCII and XML serialization works flawlessly I'm using Visual Studio 8 with Boost 1.33.1.

2 2

Getting Environment Variables
by Merrill Cornish 20 Jan '06

20 Jan '06

Is there a Boost way to get environment variables? I _thought_ I had seen a reference to that in the program options libarary, but I can't find it now. Merrill

5 7

Re: [Boost-users] gcc3.2 preprocessor concatenation or not on negative numbers
by Merrill Cornish 20 Jan '06

20 Jan '06

Andy, Thinking about it some more, I think I was wrong. Merrill

1 0

Re: [Boost-users] gcc3.2 preprocessor concatenation or not on negative numbers
by Merrill Cornish 20 Jan '06

20 Jan '06

Andy, I _think_ I may know why concatenation is working like it is. I believe the purpose of the ## concatentation operator is to cause one symbol (i.e., letters, numbers, underscore) to be created out of two symbols. Since the result of the concatenation that you specified is _not_ a legal symbol, the preprocessor appears to have made a best guess at what should be done rather than flag an error. Merrill

2 1

Compiling with egcs-2.91.66
by Emre Birol 20 Jan '06

20 Jan '06

Hi all, My application uses Graph library files of Boost 1.33.1. It is a very very simple shortest path implementation. When I try to compile it with egcs-2.91.66, Red Hat 6.2 (don't ask why), I get this weird error: boost/mpl/aux_/integral_wrapper.hpp:72: sorry, not implemented: conversion of template_type_parm as template parameter In file included from boost/mpl/integral_c.hpp:32, .... ... It uses following headers from boost: #include <boost/config.hpp> #include <boost/graph/graph_traits.hpp> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/dijkstra_shortest_paths.hpp> Any idea? How can I fix it? thanks, Emre

2 2

Re: [Boost-users] STL-PORT and VC 7.1
by Nat Goodspeed 20 Jan '06

20 Jan '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Sliwa, Przemyslaw (London) > Sent: Friday, January 20, 2006 10:32 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] STL-PORT and VC 7.1 > > I have a quick question - I would like to build the static library and > not the DLL. > Could you tell me what should be changed in the project apart form the > "General->static lib"? [Nat] Um, no. But since .vcproj files are specially-formatted text files, I assume the worst case would involve building a new static-library project by manually setting things the same way in your new project file as in my old one. The IDE may let you change the root type of a project after the fact; I've never tried. > Do you know how to disable the auto-link of boost? [Nat] These are things we changed in boost/boost/config/user.hpp: // BOOST_WHATEVER_DYN_LINK: Forces library "whatever" to be linked as a dll // rather than a static library on Microsoft Windows: replace the WHATEVER // part of the macro name with the name of the library that you want to // dynamically link to, for example use BOOST_DATE_TIME_DYN_LINK or // BOOST_REGEX_DYN_LINK etc (this macro is used to turn on __declspec(dllimport) // modifiers, so that the compiler knows which symbols to look for in a dll // rather than in a static library). // Note that there may be some libraries that can only be statically linked // (Boost.Test for example) and others which may only be dynamically linked // (Boost.Threads for example), in these cases this macro is unsupported. // #define BOOST_WHATEVER_DYN_LINK #define BOOST_FILESYSTEM_DYN_LINK #define BOOST_SIGNALS_DYN_LINK #define BOOST_PYTHON_DYN_LINK // BOOST_ALL_NO_LIB: Tells the config system not to automatically select // which libraries to link against. // Normally if a compiler supports #pragma lib, then the correct library // build variant will be automatically selected and linked against, // simply by the act of including one of that library's headers. // This macro turns that feature off. #define BOOST_ALL_NO_LIB

1 0