- Boost-users - lists.boost.org

Re: [Boost-users] [mpl] cross product and gcc specific compiler error
by Josh Quigley 14 May '12

14 May '12

Hi, I grabbed an mpl cross product implementation (product_view) of this mailing list. It works fine in MSVC but fails in gcc: // typedef boost::mpl::vector<decimal32, decimal64, decimal128> DecimalTypes; // Original typedef boost::mpl::vector<float, double, long double> DecimalTypes; // for demonstration typedef boost::mpl::vector<int, unsigned int, long, unsigned long, long long, unsigned long long> IntegralTypes; typedef product_view<boost::mpl::vector<DecimalTypes, IntegralTypes> > EachDecimalWithEachIntegralT; BOOST_AUTO_TEST_CASE_TEMPLATE(MyTest, DecimalIntegralT, EachDecimalWithEachIntegralT) { { typename DecimalIntegralT::item1 r(1234567); // Fails on gcc: no type named ‘item0’ in ‘struct boost::mpl::v_item<long long unsigned int, boost::mpl::v_item<ext::decimal::decimal128, boost::mpl::vector0<mpl_::na>, 0>, 0> typename DecimalIntegralT::item0 d(r); BOOST_CHECK_EQUAL(static_cast<long long>(d), r); } } A complete listing is below. I'm wondering if there is just something simple that needs to be tweaked? The thing that has me really stumped is the type exsits in MSVC, so I'm at a bit of loss of how to proceed debugging. Any help appreciated! // // LISTING // #include <boost/mpl/vector.hpp> #include <boost/mpl/if.hpp> #include <boost/mpl/at.hpp> #include <boost/mpl/copy.hpp> #include <boost/mpl/transform.hpp> #include <boost/type_traits/is_same.hpp> #include <boost/function.hpp> #include <boost/mpl/vector.hpp> #include <boost/mpl/list.hpp> #include <boost/mpl/joint_view.hpp> #include <boost/test/unit_test.hpp> namespace detail { using boost::mpl::at_c; using boost::mpl::at; using boost::mpl::vector3; using boost::mpl::true_; using boost::mpl::false_; using boost::mpl::if_; using boost::mpl::next; using boost::mpl::bool_; using boost::mpl::front_inserter; using boost::mpl::clear; using boost::mpl::forward_iterator_tag; using boost::mpl::reverse_copy; using boost::mpl::transform1; using boost::mpl::deref; using boost::mpl::_1; using boost::mpl::_2; using boost::mpl::int_; using boost::mpl::fold; using boost::mpl::or_; using boost::mpl::nested_begin_end_tag; template<typename S, typename O> struct state_with_overflow { typedef S state; typedef O overflow; }; template<typename I3S> struct i3s_increment_inserter { template<typename I3, typename O> struct i3_ { typedef I3 i3; typedef O overflow; }; template<typename I3> struct incr_ { typedef typename next< typename at_c< I3, 1 >::type >::type next_; typedef bool_< boost::is_same< next_, typename at_c<I3,2>::type >::value > overflow_; typedef i3_< vector3< typename at_c<I3,0>::type, typename if_< overflow_, typename at_c<I3,0>::type, next_ >::type, typename at_c<I3,2>::type >, overflow_ > type; }; typedef front_inserter< typename clear<I3S>::type > base_; typedef state_with_overflow< typename base_::state, true_ > state; struct operation { template<typename S, typename I3> struct apply { typedef typename if_< typename S::overflow, typename incr_<I3>::type, i3_<I3,false_> >::type next_; typedef state_with_overflow< typename base_::operation::template apply< typename S::state, typename next_::i3 >::type, typename next_::overflow > type; }; }; }; template< typename SWO > struct product_iterator { private: typedef typename SWO::state I3S; public: typedef forward_iterator_tag category; typedef typename transform1< I3S, deref< at< _1, int_<1> > > >::type type; typedef product_iterator< typename reverse_copy< I3S, i3s_increment_inserter<I3S> >::type > next; }; template<typename S> struct product_view { private: typedef typename transform1< S, vector3< boost::mpl::begin<_1>, boost::mpl::begin<_1>, boost::mpl::end<_1> > >::type begin_impl; typedef typename fold< S, false_, or_< _1, boost::is_same< boost::mpl::begin<_2>, boost::mpl::end<_2> > // empty<_2> refuses to compile on MSVC 7.1 > >::type empty_result; public: typedef nested_begin_end_tag tag; typedef product_iterator< state_with_overflow< begin_impl, empty_result > > begin; typedef product_iterator< state_with_overflow< begin_impl, true_ > > end; }; } using detail::product_view; // typedef boost::mpl::vector<decimal32, decimal64, decimal128> DecimalTypes; // Original typedef boost::mpl::vector<float, double, long double> DecimalTypes; // for demonstration typedef boost::mpl::vector<int, unsigned int, long, unsigned long, long long, unsigned long long> IntegralTypes; typedef product_view<boost::mpl::vector<DecimalTypes, IntegralTypes> > EachDecimalWithEachIntegralT; BOOST_AUTO_TEST_CASE_TEMPLATE(MyTest, DecimalIntegralT, EachDecimalWithEachIntegralT) { { typename DecimalIntegralT::item1 r(1234567); // Fails on gcc: no type named ‘item0’ in ‘struct boost::mpl::v_item<long long unsigned int, boost::mpl::v_item<ext::decimal::decimal128, boost::mpl::vector0<mpl_::na>, 0>, 0> typename DecimalIntegralT::item0 d(r); BOOST_CHECK_EQUAL(static_cast<long long>(d), r); } }

2 1

Boost.Filesystem (?): How to get the platform-specific path separator (not directory separator) ?
by Johannes Stallkamp 14 May '12

14 May '12

Hi all, Boost.Filesystem eases use of path names across platforms w.r.t to the directory separator (slash vs. backslash, as I am only concerned with Linux <-> Windows). Now, I need to handle _lists_ of path names (in an environment variable). Is there a similar transparent way a) to automatically handle or b) access os-specific _path_ separators (colon vs. semicolon) in Boost.Filesystem or Boost in general? In other words, I am looking for the Boost (well, C++) equivalent of Pyhton's os.pathsep, Java's path.separator or PHP's PATH_SEPARATOR. And I am looking for a library function/constant rather than a self-implemented platform-dependent preprocessor directive. The only C++-solution I am currently aware of is "apr_filepath_list_merge"/"apr_filepath_list_split" in Apache Portable Runtime (as mentioned in this answer to my question on stackoverflow: (http://stackoverflow.com/a/10514401/1228449) Is there a similar way in Boost? Kind regards Johannes

1 0

Building boost with a locally built g++
by Mark Summerfield 14 May '12

14 May '12

Hi, I have two locally built g++'s and want to build boosts to go with them. Unfortunately I've never used jam and despite reading the docs haven't succeeded. I created a user-config.jam file with the single line: using gcc : 4.7 : /home/mark/opt/gcc47/bin/g++ : <compileflags>-std=c++11 -pedantic <linkflags>-lstdc++ -lpthread -L/home/mark/opt/gcc47/lib -L/home/mark/opt/gcc47/lib64 <address-model>64; but this seems to be ignored. All I want to do is build boost for my local g++ 4.6 in ~/opt/gcc46 and my local g++ 4.7 in ~/opt/gcc47. Can anyone give me or point me to some simple instructions for doing this? I did not find http://www.boost.org/doc/libs/1_49_0/more/getting_started/unix-variants.html at all helpful for my situation. In particular, section 5.2 didn't help and b2 just keeps using the system's /usr/bin/g++. (I want boost primarily for its regex library.) Thanks! PS I'm not a subscriber so would appreciate being cc'd in any reply. -- Mark Summerfield, Qtrac Ltd, www.qtrac.eu C++, Python, Qt, PyQt - training and consultancy "Programming in Go" - ISBN 0321774639 http://www.qtrac.eu/gobook.html

3 3

[build] How do I get my trial project to use b2 correctly?
by Daryle Walker 13 May '12

13 May '12

I'm trying out GitHub for the first time. I have a new project thatis supposed to reference my Boost-SVN tree for both the Boostheaders and its build system. I copied the project structure,including Jamfiles, from John Maddock's Multiprecision trial library.I don't think I did it right because calling "b2" doesn't do anything.At first, I got directory errors because it couldn't find my Boost tree.(The "../Trunk" alias doesn't work. Is it because I'm an extradirectory away?) Then I defined "BOOST" in my shell, so "b2" canfind the tree. I just get "1 target found" notices without anythinghappening. I know nothing's being compiled because I called it fortool-sets that aren't C++11-enabled and the build just ends withouterror on my C++11 code. My directories are like: + MyProjects++ boost-trunk++ Trunk (aliased to "boost-trunk")++ Unionize (a Git clone of my project on GitHub) Does anyone want to take a look and see how I have to fix my buildfiles? The project is at "https://github.com/CTMacUser/Unionize". Daryle W.

2 1

Interprocess file mapping - fixed addresses
by Steve 13 May '12

13 May '12

I hope someone on this list is sufficiently familiar with file_mapping, mapped_region and "Fixed Address Maping" to be able to offer advice. I have a very large file, and I want to map pairs of system-page-sized blocks from it into memory. The catch: I want to map non-consecutive blocks from the file into consecutive memory addresses. For example, with block 42 mapped to BASE_ADDRESS, I might want block 7 mapped to BASE_ADDRESS+PAGE_SIZE. Under Linux, using mmap() directly, I can use one of two approaches: 1. I can map two pages from the first file-block-start-location, then map one block from the second, using MAP_FIXED, at the address returned by the first mapping plus one page. 2. I could map two pages of /dev/zero - then map the first and second pages explicitly using MAP_FIXED into that address space - this copes with the case where the first block in memory is the last block of the file, but has an overhead. Both approaches seem to work, though I've not done extensive testing. I'd prefer to avoid using mmap directly - both to gain the advantages of a well-tested library using high-level C++ idioms - and to compile cross-platform... definitely targeting Linux and Windows - but maybe other platforms too. When I attempt to use the above strategy, I get a boost::interprocess_exception::library_error - though I'm not sure why. The following code fails on both Linux and Windows: -- size_t page_size=mapped_region::get_page_size(); file_mapping m("file.dat", read_write); mapped_region ra(m,read_write,0,2*page_size); uint8_t *base=reinterpret_cast<uint8_t*>(ra.get_address()); mapped_region rb(m,read_write,2*page_size,page_size,base+page_size); //FAILS -- whereas, the following appeared to work under Windows 7 (but not Linux) - perhaps by fluke of the way address space was allocated: -- size_t page_size=mapped_region::get_page_size(); file_mapping m("file.dat", read_write); mapped_region ra(m,read_write,0,page_size); // N.B. Map 1 page only uint8_t *base=reinterpret_cast<uint8_t*>(ra.get_address()); mapped_region rb(m,read_write,2*page_size,page_size,base+page_size); -- The documentation seems focused on MAP_FIXED being used to map specific data in a file into the same memory address on every invocation - with the obvious reference to validity of pointers referencing the mapped address space. Pointers are not a problem for me - my file contains no pointers - just a long stream of application-specific values. I only need to explicitly control two pages to be mapped into contiguous memory - I can be flexible about the mapped base address for the pair of pages - and it doesn't matter if different addresses are chosen on successive invocations.. as long as, in memory, the byte following the last byte of the first page is the first byte of the second page for an arbitrary pair of pages... irrespective of the locations of those pages in the file. Is this a use case that boost::interprocess intends to support? What is the most appropriate, robust, platform independent, way to reserve address space in order map into fixed addresses to ensure contiguous mappings? Do I need to mmap /dev/zero for POSIX systems - and use VirtualAlloc with MEM_RESERVE on Windows myself in order to use fixed-address memory mapping of files? Does boost offer anything to make this platform independent - or am I 'on my own'?

1 0

NaN and lexical_cast
by Claude 13 May '12

13 May '12

Hi! It is possible to convert the "nan" string in a Non a Number floating point value using lexical_cast? -- View this message in context: http://boost.2283326.n4.nabble.com/NaN-and-lexical-cast-tp4630035.html Sent from the Boost - Users mailing list archive at Nabble.com.

1 0

boost::asio::ip::tcp::resolver::query does not work in Ubuntu!!!
by mozdemir 13 May '12

13 May '12

Hi, I am a newbie with Asio library and i got into crashing problem with the following code: boost::system::error_code ec; boost::asio::ip::tcp::resolver resolver( m_hive->GetService() ); boost::asio::ip::tcp::resolver::query query( m_host, boost::lexical_cast< std::string >( m_port )); boost::asio::ip::tcp::resolver::iterator endpointIterator = resolver.resolve( query ); When i run this in windows, i do not have any isssue but in Ubuntu, it works and after some more resolution of query, it crashes. I did install Miredo to activate Ipv6 but it did not help. When the crash happens, i see this: 23:37.47 [dbg] [0x8ec0650] Exception: Host not found (authoritative) 23:37.47 [dbg] [0x8ec0650] Exception: Host not found (authoritative) terminate called after throwing an instance of 'boost::exception_detail::clone_impl<boost::exception_detail::error_info_injector<boost::system::system_error> >' what(): Host not found (authoritative) Aborted (core dumped) Do you know what else is missing or what is wrong? Thanks a lot, mustafa -- View this message in context: http://boost.2283326.n4.nabble.com/boost-asio-ip-tcp-resolver-query-does-no… Sent from the Boost - Users mailing list archive at Nabble.com.

2 4

Serializing a vector containing an unsupported data type
by Tim Burgess 13 May '12

13 May '12

Hi, I have a vector: Std::vector <std::pair< std::string, wxAny> > myParameters; I need to serialize this as part of the class it belongs to and I'm confused as to how to do it, given that the wxWidgets wxAny class isn't supported by Boost. A colleague suggested using lexical cast and I think I can see how this would help if I was just serializing an individual wxAny variable, but how would I cope with it within the vector? I'm serializing out to XML, if this is an issue. Can anybody help, please? Best wishes. Tim Burgess Raised Bar Ltd Phone: +44 (0)1827 719822 Don't forget to vote for improved access to music and music technology at <http://www.raisedbar.net/petition.htm> http://www.raisedbar.net/petition.htm

2 1

[interprocess] interprocess_barrier missing
by Sebastian Messerschmidt 12 May '12

12 May '12

Hello, I just upgraded my boost version from 1.42 to 1.49. Unfortunately the interprocess_barrier is not longer part of the interprocess library. I've checked the SVN and the file /boost/interprocess/sync/interprocess_barrier has been removed in revision 76106. Is there a reason for the removal? Or can someone point me to replacement? Right now I can live without it, but in order to get some of my components working again, I need a shared memory barrier. Any hints? cheers Sebastian

2 2

Boost libraries on MacOS X: libboost_foo.dylib vs libboost_foo-mt.dylib (+)
by Nick Zavaritsky 12 May '12

12 May '12

Hi all! Please enlighten me what is the difference between libboost_foo.dylib and libboost_foo-mt.dylib? I guess -mt naming convention originates from Windows which once had both signle- and multithreaded CRT (-mt). However there is neither a single-threaded CRT nor a single-threaded program on Mac, thanks to GCD. Is there any use for non-MT libboost_foo.dylib or not? And let me squeeze another vaguely related question in the same letter. Are Boost debug and normal library variants binary-compatible? The reason I am curious is because Mac dynamic library linker (dyld) has a promising feature to load a different library variant without program recompilation. This feature is activated through the environment variable DYLD_IMAGE_SUFFIX. Ex: setting DYLD_IMAGE_SUFFIX=_debug will result in dyld attempting to load libfoo_debug.dylib instead of libfoo.dylib. Thanks in advance.

1 0