- Boost-users - lists.boost.org

using bind with operator new
by Archie14 19 May '09

19 May '09

I am trying to use bind to generate ptr_vector list of objects from another vector #include <boost/shared_ptr.hpp> #include <boost/bind.hpp> class a { }; class b { public: b(a* val) : _val(val) { } private: boost::shared_ptr<a> _val; }; void test() { boost::ptr_vector<a> lsta; boost::ptr_vector<b> lstb; typedef boost::ptr_vector<b> listofb; // create lstb by applying new to each element of lsta and pushing it to lstb. Fails. I will appreciate any help with this. std::for_each (lsta.begin(), lsta.end, boost::bind((&listofb::push_back), lstb, boost::bind((&operator new), b, _1) ); }

2 1

boost::interprocess::managed_shared_memory uses C: - Need to change this
by Moorthi 19 May '09

19 May '09

I am using boost::interprocess::managed_shared_memory to share the data between processes. But the shared memory internally creates memory mapped file in C: dirve. if anyone knows how to change the configuration to use some other drive instead of C: as my production environment doesnt have much space in C: drive. -- View this message in context: http://www.nabble.com/boost%3A%3Ainterprocess%3A%3Amanaged_shared_memory-us… Sent from the Boost - Users mailing list archive at Nabble.com.

2 2

directory_iterator ++operator never return
by gtsml owevwr 19 May '09

19 May '09

Hi, I'm getting a list of the file in a directory like that: fs::directory_iterator end ; for( fs::directory_iterator iter(myfilepath) ; iter != end ; ++iter ) {...} And at some point in this for loop, the ++operator never return, see callstack (single-thread build). I found a related thread here: http://marc.info/?l=boost-users&m=118835374908173&w=2 I compiled boost in threaded and non-threaded mode and still have the same pb in both mode... This bug reproduce with iphone os21 and os3.0, boost 1.38.0 (multi/single thread). Anybody? Thanks, Frank #0 0x31d41270 in swtch_pri #1 0x31dd19cc in sched_yield #2 0x000ac1d4 in boost::detail::yield at yield_k.hpp:110 #3 0x000adb54 in scoped_lock [inlined] at detail:45 #4 0x000adb54 in boost::detail::spinlock::lock at detail:32 #5 0x000adb54 in atomic_exchange_and_add [inlined] at detail:61 #6 0x000adb54 in boost::detail::sp_counted_base::release at detail:100 #7 0x000adb54 in boost::shared_ptr<boost::filesystem::detail::dir_itr_imp<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> > >::reset at detail:216 #8 0x000adb54 in boost::filesystem::basic_directory_iterator<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> >::increment #9 0x0009cc44 in boost::iterator_core_access::increment<boost::filesystem::basic_directory_iterator<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> > > at iterator_facade.hpp:522 #10 0x0009cc64 in boost::iterator_facade<boost::filesystem::basic_directory_iterator<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> >, boost::filesystem::basic_directory_entry<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> >, boost::single_pass_traversal_tag, boost::filesystem::basic_directory_entry<boost::filesystem::basic_path<std::string, boost::filesystem::path_traits> >&, int>::operator++ at iterator_facade.hpp:664

2 7

building boost 1.39 on iphone os3.0
by gtsml owevwr 19 May '09

19 May '09

Hi, I understand a lot of effort is being spent on making boost easy to compile and so on but as far as my experience goes, that's always a bit painful... So here's what I'm going to do as a background task: I'm going to play it dumb (eg. not spend 30 minutes on the web everytime I hit something) and tell you all about it :-) My goal here is to compile boost 1.39 on the iphone os3.0. Note that I already compiled 1.38 on os21 and os30 so I have gathered some experience already, but as I said, I'm playing it dumb. So: - getting boost from the web, uncompress and so on - boostrap, so far so good - add my user-config with: ################################################################# # Compiler configuration using darwin : 4.0.1~iphone : /Developer/Platforms/iPhoneOS.platform/Developer/usr/bin/gcc-4.2 -arch arm : <striper> : <architecture>arm <target-os>iphone <macosx-version>iphone-3.0 ; using darwin : 4.0.1~iphonesim : /Developer/Platforms/iPhoneSimulator.platform/Developer/usr/bin/gcc-4.2 -arch i386 : <striper> : <architecture>x86 <target-os>iphone <macosx-version>iphonesim-3.0 ; ################################################################# - and type ./bjam BUT, here we go, first problem, my user-config.jam seems to be ignored. the compiler used is "g++": "/Developer/Platforms/iPhoneOS.platform/Developer/usr/bin/gcc-4.2" still the doc says: On startup, Boost.Build searches and reads two configuration files: site-config.jam and user-config.jam. On top of that, the project-config.jam file says: # Compiler configuration. This definition will be used unless # you already have defined some toolsets in your user-config.jam # file. if ! darwin in [ feature.values <toolset> ] { using darwin ; } Anybody get any idea? Frank

2 4

[BGL] Trying to use subgraph
by Lindley M French 19 May '09

19 May '09

I've been trying to work out how to use the subgraph adapter. I've come to realize that in order to use it, my primary graph must *itself* be an object of the subgraph type (unintuitive). However, I'm getting a static assertion error that edge_index is an unknown property. I had *thought* that an adjacency_list<vecS,vecS> would not have that issue, but apparently I was wrong. The page here: http://www.boost.org/doc/libs/1_39_0/libs/graph/doc/subgraph.html says "we add an edge index property to the adjacency list", but the code which follows demonstrates no such thing. The distinction between bundled properties and internal properties is still a bit mysterious to me. Can anyone shed some light on this?

2 1

Serialization Archive Incompatibility between Boost Versions
by David Graf 19 May '09

19 May '09

Hello How is it possible to load serialized objects with a newer boost version than they have been saved with. I have installed boost version 1.36 and I cannot read archives that have been save with boost 1.35. This is not possible because boost 1.35 puts 'serialization::archive 4' into the files but boost 1.36 needs the archive version 5. I found some information in the documentation that is must be possible to read old archives. But I am not able to do it. I always get exceptions. Thank you in advance for help! David

2 8

Boost up for SourceForge Community Choice Award
by David Abrahams 19 May '09

19 May '09

Cast your vote here: http://tinyurl.com/boost-cca09 Cheers, -- Dave Abrahams BoostPro Computing http://www.boostpro.com

2 1

[signals2] connection blocking
by Bryan Green 19 May '09

19 May '09

Hi, I've starting porting some code from signals to signals2, and I've run into a problem because of the removal of the 'block' and 'unblock' methods from a connection object. I understand the rational for using the 'shared_connection_block' object, but it doesn't really apply in my case. I have a GUI widget which receives events via a signal, unless it is in an "inactive" state, during which time events are blocked: class widget { private: boost::signals::connection conn; ... public: void toggle_active_state() { ... if (inactive) conn.block(); else conn.unblock(); } }; It is obviously awkward to be forced to convert this code to use an RAII object, and I'd rather not do a full connect/disconnect for this state change. What do you think? Should I be thinking about this differently? Or can a good case be made for bringing back the block()/unblock() methods? Thanks, -bryan

4 7

[Regex] Source-only link error in MSVC
by dtexg＠hush.com 19 May '09

19 May '09

Hi, I'm trying to include the Boost Regex source in a Visual C++ project, but I get the following linker error: (indented for legibility) error LNK2019: unresolved external symbol "private: class boost::basic_regex< char, struct boost::regex_traits< char, class boost::w32_regex_traits<char> > > & __thiscall boost::basic_regex< char, struct boost::regex_traits< char, class boost::w32_regex_traits<char> > >:: do_assign( char const *, char const *, unsigned int)" (?do_assign@?$basic_regex@DU?$regex_traits@DV?$w32_regex_traits@D@bo ost@@@boost@@@boost@@AAEAAV12@PBD0I@Z) referenced in function "public: class boost::basic_regex< char, struct boost::regex_traits< char, class boost::w32_regex_traits<char> > > & __thiscall boost::basic_regex< char, struct boost::regex_traits< char, class boost::w32_regex_traits<char> > >:: assign( char const *, char const *, unsigned int)" (?assign@?$basic_regex@DU?$regex_traits@DV?$w32_regex_traits@D@boost @@@boost@@@boost@@QAEAAV12@PBD0I@Z) Here are the steps I'm taking: I downloaded the latest Boost (boost_1_39_0.zip), and extracted it all into its own directory; I've tried extracting with both the built-in Windows Zip Folder support and the latest WinRAR. I did not do any making (nmake) or building on the Boost source -- after extracting, I don't touch this directory directly at all. In both Visual Studio 2005 and Visual C++ 2008 Express Edition (http://www.microsoft.com/express/vc/) I created a new empty console project, and I added a new .cpp file to the project. I went to Project -> Properties, Configuration Properties -> C/C++ - > General: Additional Include Directories, and (for the "Active(Debug)" configuration) added the full path of the directory containing the source, so that "boost" is a subdirectory in the selected directory. I wrote the following code: #define BOOST_REGEX_SOURCE #include <boost/regex.hpp> int main() { boost::regex r("a"); return 0; } Everything compiles, but linking fails with the above error. The same error results if I try this: boost::regex * p = new boost::regex("a"); Or if I use boost::wregex, or boost::basic_regex<char,boost::regex_traits<char,boost::w32_regex_tr aits<char>>>, etc. Also, the outcome is the same whether I #define BOOST_REGEX_SOURCE, BOOST_REGEX_NO_LIB, or BOOST_ALL_NO_LIB, and whether I define them in the .cpp or in Project -> Properties, Configuration Properties -> C/C++ -> Preprocessor: Preprocessor Definitions, and whether I #include <boost/regex.hpp> or <boost/regex/v4/regex.hpp>, and whether I use Debug or Release configuration. If I put a #error by the do_assign implementation in boost/regex/v4/basic_regex.hpp, it gets hit, so the implementation isn't being #if'ed away somehow. It will build successfully if, in the .cpp file in which main() resides, I write an implementation and replace all the template arguments, but besides being vulgar, this kludge also fails (with compiler errors) if I try to copy-and-paste the full implementation from basic_regex.hpp, so it's untenable as a workaround. In a post from 2006, John wrote, "if you are doing something, umm, strange, and you can't track down the cause of the error, you can always just add the regex source directly to your application," but I've not seen this successfully demonstrated in Visual C++. I'm just looking for an absolutely parsimonious way to add Boost Regex to a VC++ project. /Eric -- Free information on Digital Photography. Click Now! http://tagline.hushmail.com/fc/BLSrjkqZZHWE3MuglDmDyK5F8gPPnxz3V9yL7kN5dsQj…

2 1

PropertyTree's XML parsers survey
by Tan, Tom (Shanghai) 19 May '09

19 May '09

>From: Nat Goodspeed <nat(a)lindenlab.com> >> - I often embed some comment nodes inside the xml to facilitate the >> manual edit. E.g, enumeration of accepted value some fields. I wish >> boost.serialize ignores them when parsing, > >That seems plausible to me... > >> and embed them when saving (is this possible). >This request feels more difficult. Boost.Serialize is designed to read >from a file into /your/ data structures, and/or write those data >structures to a file. XML is only one of several possible archive >formats for such a file. > >Where in your own data structures would you recommend the Serialize >library store any XML comments it may encounter? > >Supposing there were an appropriate place to collect such comments, how >would you write them again in something like their original file location? Just for the fun of discussion, how about allowing only one comment node only and fixing the position to be the end of the file, like this: First, add an optional parameter to signature. template<class Archive> void serialize(Archive & ar, const unsigned int version, basic_string<Ch> comment = basic_string<Ch>()) Then, toward the end of the file, append this node <!== {comment string here} ==> This does not break the compatibility of the existing version, and that the xml archive version already includes more info(i.e. names in additional to values) than other variants(e.g. bins), it does not actually violate the current philosophy much. >>Instead of incrementally enhancing a minimal XML parser, wouldn't it >>make more sense to jump right to an existing full-blown parser >>implementation? Ah, while this is too limited, a full-blown is too heavy.

1 0