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

[serialization] Question about pointers and casting
by Ben 24 Oct '09

24 Oct '09

Dear Boost Experts -- I'm using g++ 4.2.1, OS X 10.6.1, boost 1.40, although I have also seen this on linux. In attempting to understand how boost::serialization works, I've stumbled across the following curiosity. The following program compiles and runs fine (I've eliminated diagnostic output which verifies the serialization.) #include <iostream> #include <fstream> #include <boost/archive/binary_oarchive.hpp> #include <boost/archive/binary_iarchive.hpp> struct A { int v; template<class Ar> void serialize(Ar & ar, const unsigned int ver) { ar & v; } }; int main(int ac, char ** argv) { A a; a.v = 10; A * a1p = &a; const char * fname = "test.bin"; { std::ofstream ofs(fname); boost::archive::binary_oarchive ar(ofs); ar << a1p; } A * a2p; { std::ifstream ifs(fname); boost::archive::binary_iarchive ar(ifs); ar >> a2p; } return 0; } However, if I add a static_cast<A*> to either a1p or a2p (which are already declared to be type A*) in the serialization / deserialization lines, it fails to compile: #include <iostream> #include <fstream> #include <boost/archive/binary_oarchive.hpp> #include <boost/archive/binary_iarchive.hpp> struct A { int v; template<class Ar> void serialize(Ar & ar, const unsigned int ver) { ar & v; } }; int main(int ac, char ** argv) { A a; a.v = 10; A * a1p = &a; const char * fname = "test.bin"; { std::ofstream ofs(fname); boost::archive::binary_oarchive ar(ofs); ar << static_cast<A*>(a1p); } A * a2p; { std::ifstream ifs(fname); boost::archive::binary_iarchive ar(ifs); ar >> static_cast<A*>(a2p); } return 0; } Compilation started at Fri Oct 23 14:31:17 make -k debug make all "CXX_FLAGS = -g -fno-inline -m64 -I. -I.. -I/opt/local/include -Wall -Wextra -Wpointer- arith -Wcast-align -Wwrite-strings -Wconversion -Woverloaded-virtual -Wno-sign-compare -Wno- unused -fno-nonansi-builtins" \ "LD_FLAGS = -g -m64" g++ -g -fno-inline -m64 -I. -I.. -I/opt/local/include -Wall -Wextra -Wpointer-arith -Wcast-align - Wwrite-strings -Wconversion -Woverloaded-virtual -Wno-sign-compare -Wno-unused -fno- nonansi-builtins -c test.cc -o test.o test.cc: In function 'int main(int, char**)': test.cc:25: error: no match for 'operator<<' in 'ar << a1p' /opt/local/include/boost/archive/detail/interface_oarchive.hpp:63: note: candidates are: Archive& boost::archive::detail::interface_oarchive<Archive>::operator<<(T&) [with T = A*, Archive = boost::archive::binary_oarchive] test.cc:33: error: no match for 'operator>>' in 'ar >> a2p' /opt/local/include/boost/archive/detail/interface_iarchive.hpp:60: note: candidates are: Archive& boost::archive::detail::interface_iarchive<Archive>::operator>>(T&) [with T = A*, Archive = boost::archive::binary_iarchive] make[1]: *** [test.o] Error 1 make[1]: Target `all' not remade because of errors. make: *** [debug] Error 2 Compilation exited abnormally with code 2 at Fri Oct 23 14:31:18 This completely boggles my mind. My limited understanding suggests the compiler should not be able to distinguish between the two programs! I can put the cast on either one of the two pointers and it fails, and using an old style cast to A* also fails. I seem to not understand what boost::serialization is doing generally, so I was hoping the answer to this question (which is obviously a theoretical curiosity and of no practical value) would lead to further insight. Best regards, Ben

3 2

boost.thread query
by Sean Farrow 23 Oct '09

23 Oct '09

Hi: I am considering using boost.threadfir a project and was wondering whether the function passed to boost.thread has to be of any pargicular type i.e void return. Secondly, if I don't call thread.join, will the thread continue executing? I'm terminating the thread with a windows event. Finally, is there a terminate function? Any help appreciated. Sean.

2 1

[serialization] Special case of abstract Interface serialization.
by Germán Diago 23 Oct '09

23 Oct '09

Hello. I have a problem which I think it's not documented, and I don't know how to solve it. The problem is the following: I have a framework with an Object base class. I have an interface IDrawable. Interfaces don't inherit Object. Now I have a container, which holds IDrawables. class IDrawable { virtual void draw() const = 0; } Container<IDrawable *> drawables; //holds IDrawables that I know they are Objects. I do know that inside the drawables, objects are held, so now I want to serialize that container: ar << drawables; IDrawable does not hold any data, but Object does. So I want to serialize these objects in a polymorphic way, so that I can save anything derived from Object which is pointed by an IDrawable. Thanks for your time.

1 0

[serialization] Cannot serialize multimap.
by Germán Diago 23 Oct '09

23 Oct '09

Hi. I have a multimap<A, B> where A and B are not default constructible. I provided load_construct_data and save_construct_data for both, but the library doesn't allow me to serialize this multimap. It fails at compile time: It seems to be the library trying to use default constructors. Does anyone know any way or a workaround or whatever to make this compile correctly? Thanks in advance.

2 2

[asio] icmp example not working on Windows 2000
by Wolfgang Fertsak 23 Oct '09

23 Oct '09

Hello! The icmp example provided with the asio-library throws a system error with code 10044 "The support for the specified socket type does not exist in this address family." when compiled with _WIN32_WINNT 0x0500 and executed on an Win2000 machine. The error comes from the gai_echeck()-function which is called with family AF_INET and socktype SOCK_RAW: inline int gai_echeck(const char* host, const char* service, int flags, int family, int socktype, int protocol) { .... switch (family) { case AF_UNSPEC: break; case AF_INET: case AF_INET6: if (socktype != 0 && socktype != SOCK_STREAM && socktype != SOCK_DGRAM) return EAI_SOCKTYPE; ... I have attached the callstack for call to gai_echeck() I'm using Visual Studio 2005 and boost 1.39. I changed the return value of gai_echeck() back to 0 in the debugger and then the rest of the icmp(ping) example ran without any apparent problems. Best regards, Wolfgang boost::asio::detail::socket_ops::gai_echeck(const char * host=0x0012fad0, const char * service=0x0012faa8, int flags=0, int family=2, int socktype=3, int protocol=1) Line 1380 C++ boost::asio::detail::socket_ops::getaddrinfo_emulation(const char * host=0x0012fad0, const char * service=0x0012faa8, const addrinfo * hintsp=0x0012fa78, addrinfo * * result=0x0012faf0) Line 1441 + 0x1d bytes C++ boost::asio::detail::socket_ops::WspiapiGetAddrInfo(const char * host=0x0012fad0, const char * service=0x0012faa8, const addrinfo * hints=0x0012fa78, addrinfo * * result=0x0012faf0, boost::system::error_code & ec={...}) Line 1801 + 0x15 bytes C++ boost::asio::detail::resolver_service<boost::asio::ip::icmp>::resolve(boost::shared_ptr<void> & __formal={...}, const boost::asio::ip::basic_resolver_query<boost::asio::ip::icmp> & query={...}, boost::system::error_code & ec={...}) Line 147 + 0x51 bytes C++ boost::asio::ip::resolver_service<boost::asio::ip::icmp>::resolve(boost::shared_ptr<void> & impl={...}, const boost::asio::ip::basic_resolver_query<boost::asio::ip::icmp> & query={...}, boost::system::error_code & ec={...}) Line 105 + 0x1b bytes C++ boost::asio::ip::basic_resolver<boost::asio::ip::icmp,boost::asio::ip::resolver_service<boost::asio::ip::icmp> >::resolve(const boost::asio::ip::basic_resolver_query<boost::asio::ip::icmp> & q={...}) Line 98 + 0x1d bytes C++ pinger::pinger(boost::asio::io_service & io_service={...}, const char * destination=0x00324199) Line 32 + 0x1a bytes C++ main(int argc=2, char * * argv=0x00324170) Line 155 + 0x16 bytes C++

1 1

[Program Options] first.cpp compile problem
by philip tucker 23 Oct '09

23 Oct '09

Hello, I'm an enthusiastic and grateful user of boost (multi_index_set, accumulators, much else), so first thanks a lot to you all. But I can't get program_options/example/first.cpp to compile. I am sure it's something simple I've missed, but I can't tell what. I am running boost 1.40, gcc version 4.0.1, on a new iMac. output: cc -g -o test obj/pgmopts.o -lstdc++ Undefined symbols: "boost::program_options::detail::cmdline::run()", referenced from: boost::program_options::basic_command_line_parser<char>::run () in pgmopts .o "boost::program_options::options_description_easy_init::operator() (char const* , boost::program_options::value_semantic const*, char const*)", referenced from: _main in pgmopts.o "boost::program_options::validation_error::what() const", referenced from: vtable for boost::exception_detail::error_info_injector<boost::program_opt ions::invalid_option_value>in pgmopts.o vtable for boost::program_options::invalid_option_valuein pgmopts.o vtable for boost::exception_detail::clone_impl<boost::exception_detail::er ror_info_injector<boost::program_options::invalid_option_value> >in pgmopts.o ... Help! and thanks again.

2 2

[BGL] seg fault
by John Robertson 23 Oct '09

23 Oct '09

Replacing vector < set< Colour > > eColours; with std::map < int, set< Colour > > emColours; in edge properties. Using property_map < graph_t, map < int, <set< Colour > > > edge_properties::*>::type emColours_map = get(&edge_properties::emColours, g); But get compiler error: template argument 2 invalid. Thanks -JR

2 2

[BGL] seg fault
by John Robertson 23 Oct '09

23 Oct '09

Something is goofy. Sometimes only part of my mail message is shown in the boost users list. Part is deleted. Using yahoo mail. Maybe its their new version thats messed up. -JR

1 0

[BGL]seg fault : resend
by John Robertson 23 Oct '09

23 Oct '09

Trying to replace vector with std::map as follows. Replace typedef string Colour; struct edge_properties { vector < set< Colour > > eColours; } with typedef string Colour; struct edge_properties { std::map < int, set< Colour > > emColours; } Using property_map < graph_t, map < int, <set< Colour > > > edge_properties::*>::type emColours_map =get(&edge_properties::emColours,g); getting compiler errors error: template argument 2 is invalid error: template argument 4 is invalid Thanks -JR ReplyReply AllMove...InboxBGLGo to Previous message | Go to Next message | Back to Messages Select Message EncodingASCII (ASCII)Greek (ISO-8859-7)Greek (Windows-1253)Latin-10 (ISO-8859-16)Latin-3 (ISO-8859-3)Latin-6 (ISO-8859-10)Latin-7 (ISO-8859-13)Latin-8 (ISO-8859-14)Latin-9 (ISO-8859-15)W. European (850)W. European (CP858)W. European (HPROMAN8)W. European (MACROMAN8)W. European (Windows-1252)Armenia (ARMSCII-8)Baltic Rim (ISO-8859-4)Baltic Rim (WINDOWS-1257)Cyrillic (866)Cyrillic (ISO-8859-5)Cyrillic (KOI8-R)Cyrillic (KOI8-RU)Cyrillic (KOI8-T)Cyrillic (KOI8-U)Cyrillic (WINDOWS-1251)Latin-2 (852)Latin-2 (ISO-8859-2)Latin-2 (WINDOWS-1250)Turkish (ISO-8859-9)Turkish (WINDOWS-1254)Arabic (ISO-8859-6, ASMO-708)Arabic (WINDOWS-1256)Hebrew (856)Hebrew (862)Hebrew (WINDOWS-1255)Chinese Simplified (GB-2312-80)Chinese Simplified (GB18030)Chinese Simplified (HZ-GB-2312)Chinese Simplified (ISO-2022-CN)Chinese Simplified (WINDOWS-936)Chinese Trad.-Hong Kong (BIG5-HKSCS)Chinese Traditional (BIG5)Chinese Traditional (EUC-TW)Japanese (SHIFT_JIS)Japanese (EUC-JP)Japanese (ISO-2022-JP)Korean (ISO-2022-KR)Korean (EUC-KR)Thai (TIS-620-2533)Thai (WINDOWS-874)Vietnamese (TCVN-5712)Vietnamese (VISCII)Vietnamese (WINDOWS-1258)Unicode (UTF-7)Unicode (UTF-8)Unicode (UTF-16)Unicode (UTF-32)| Full Headers Mail Search #yui-main { background-color:#D6DEEC }gLaunchProfile.stop('RT_RP', false);gLaunchProfile.start('RT_BP');gLaunchProfile.start('LD_JS_BP');YAHOO.mc.intl = "us";YAHOO.mc.IM_TOGGLE_STRINGS = {"im_settings_helpurl" : "http://help.yahoo.com/l/us/yahoo/mail/classic/chat/","im_settings_help_toggle" : "1","im_enable_mobile_l10n": "1","im_sign_in": "true","im_time_format_am" : "AM","im_time_format_pm" : "PM","im_date_time_format" : "{M}/{d}/{yy} {h}:{mm} {A}"};if (YAHOO.mc.page.generic) {YAHOO.mc.page.generic.initComposeMenu();}YAHOO.mc.page.BasieSpaceId= "398300973";YAHOO.mc.page.TriggerULT=true;YAHOO.util.Event.onDOMReady(function () {sessionMgr.initSession();sessionMgr.loadFromBookmarks();});var adPrefetch = true;var isAdPrefetched = false;if (document.getElementById("JSMailOptionsLink")) {document.getElementById("JSMailOptionsLink").href = optionObj["JSMailOptionsLink"];}if (document.getElementById("chk_pop_options") || document.getElementById("chk_pop_error") || document.getElementById("option_bad_folder") || document.getElementById("folders_options_spam") || document.getElementById("folders_options_sent") || document.getElementById("spam_edit_options")) {YAHOO.util.Event.addListener(YAHOO.mc.optionsID, "click", displayOptions, optionObj);}YAHOO.namespace("fcue"); YAHOO.mc.FCUE_METADATA = {imInit:"",imSent:"",fcElap:"1254813133500",fcEcap:"2",fcNag:"",inEcap:"",inElap:""};YAHOO.mc.noCues = false;YAHOO.mc.msgrLoaded = false;YAHOO.mc.loadMsgr = function () {clearTimeout(YAHOO.mc.msgrLoadID);if (hasIM && !YAHOO.mc.msgrLoaded) {YAHOO.mc.msgrLoaded = true;YAHOO.util.Get.script("http://mail.yimg.com/d/combo?/mg/8_0_4/js/im_blue_all.js&/mg/8_0_4/js/us_st…");}};YAHOO.util.Event.on(window, "load", YAHOO.mc.loadMsgr);YAHOO.mc.msgrLoadID = setTimeout(YAHOO.mc.loadMsgr, 5000); if (YAHOO.mc.page.galaxy) { YAHOO.mc.page.galaxy.cleanup(); } YAHOO.mc.page.galaxy = new GalaxyHandler(); function myMsgHandler(fid, mid) { YAHOO.mc.page.galaxy.evtCallBack("showMessage?fid=" + fid + "&mid=" + mid + "&.rand=557745628"); } YAHOO.galaxy.setMsgListener(myMsgHandler); function myFolderHandler(fid, filterBy) { var filterUrl = ""; if (filterBy) { filterUrl += "&filterBy=" + filterBy; } YAHOO.mc.page.galaxy.evtCallBack("showFolder?fid=" + fid + filterUrl + "&.rand=557745628"); } YAHOO.galaxy.setFolderListener(myFolderHandler); gLaunchProfile.stop('LD_JS_BP', false);resizeLeftPane();gLaunchProfile.start('RT_AD_FOOT');

1 0

[BGL] seg fault
by John Robertson 23 Oct '09

23 Oct '09

Trying to replace vector with std::map as follows. Replace typedef string Colour; struct edge_properties { vector < set< Colour > > eColours; } with typedef string Colour; struct edge_properties { std::map < int, set< Colour > > emColours; } Using property_map < graph_t, map < int, <set< Colour > > > edge_properties::*>::type emColours_map =get(&edge_properties::emColours,g); getting compiler errors error: template argument 2 is invalid error: template argument 4 is invalid Thanks -JR

1 0

[BGL] seg fault
by John Robertson 22 Oct '09

22 Oct '09

Could I set the size of a set< String> at instantiation time? -JR

1 0

[BGL] seg fault : more readable
by John Robertson 22 Oct '09

22 Oct '09

The following compiles and runs. typedef string Colour; struct edge_properties { vector < set< Colour > > eColours; } //get a reference to the vector of the sets of colours of some edge. vector < set<Colour> > & v_colour_outset = get(m_eColours_map,*out_iter); // Here *out_iter is some edge. //Add a set<Colour> to v_colour_outset ie. to get(m_eColours_map,*out_iter) . set<Colour> colour_set; v_colour_outset.push_back(colour_set); // get a reference to one of the set<colour> in get(m_eColours_map,*out_iter) . set<Colour> & outcset = v_colour_outset[1]; //or [0], or any elt.. But when I try to add a new colour to this set outcset.insert("Blue"); I get segfault. My guess is that there is a predetermined size for colour_set, and when it is push_backed into v_colour_outset, a new colour cannot be added into it. I tried to create a vector of set<Colour> references as follows. struct edge_properties { vector < set< Colour > & > eColours; } But it did not compile. ( shouldn't it? ) Whats the best solution to this problem. Thanks - JR ReplyReply AllMove...InboxBGLGo to Previous message | Go to Next message | Back to Messages Select Message EncodingASCII (ASCII)Greek (ISO-8859-7)Greek (Windows-1253)Latin-10 (ISO-8859-16)Latin-3 (ISO-8859-3)Latin-6 (ISO-8859-10)Latin-7 (ISO-8859-13)Latin-8 (ISO-8859-14)Latin-9 (ISO-8859-15)W. European (850)W. European (CP858)W. European (HPROMAN8)W. European (MACROMAN8)W. European (Windows-1252)Armenia (ARMSCII-8)Baltic Rim (ISO-8859-4)Baltic Rim (WINDOWS-1257)Cyrillic (866)Cyrillic (ISO-8859-5)Cyrillic (KOI8-R)Cyrillic (KOI8-RU)Cyrillic (KOI8-T)Cyrillic (KOI8-U)Cyrillic (WINDOWS-1251)Latin-2 (852)Latin-2 (ISO-8859-2)Latin-2 (WINDOWS-1250)Turkish (ISO-8859-9)Turkish (WINDOWS-1254)Arabic (ISO-8859-6, ASMO-708)Arabic (WINDOWS-1256)Hebrew (856)Hebrew (862)Hebrew (WINDOWS-1255)Chinese Simplified (GB-2312-80)Chinese Simplified (GB18030)Chinese Simplified (HZ-GB-2312)Chinese Simplified (ISO-2022-CN)Chinese Simplified (WINDOWS-936)Chinese Trad.-Hong Kong (BIG5-HKSCS)Chinese Traditional (BIG5)Chinese Traditional (EUC-TW)Japanese (SHIFT_JIS)Japanese (EUC-JP)Japanese (ISO-2022-JP)Korean (ISO-2022-KR)Korean (EUC-KR)Thai (TIS-620-2533)Thai (WINDOWS-874)Vietnamese (TCVN-5712)Vietnamese (VISCII)Vietnamese (WINDOWS-1258)Unicode (UTF-7)Unicode (UTF-8)Unicode (UTF-16)Unicode (UTF-32)| Full Headers Mail Search #yui-main { background-color:#D6DEEC }gLaunchProfile.stop('RT_RP', false);gLaunchProfile.start('RT_BP');gLaunchProfile.start('LD_JS_BP');YAHOO.mc.intl = "us";YAHOO.mc.IM_TOGGLE_STRINGS = {"im_settings_helpurl" : "http://help.yahoo.com/l/us/yahoo/mail/classic/chat/","im_settings_help_toggle" : "1","im_enable_mobile_l10n": "1","im_sign_in": "true","im_time_format_am" : "AM","im_time_format_pm" : "PM","im_date_time_format" : "{M}/{d}/{yy} {h}:{mm} {A}"};if (YAHOO.mc.page..generic) {YAHOO.mc.page.generic.initComposeMenu();}YAHOO.mc.page.BasieSpaceId= "398300973";YAHOO.mc.page.TriggerULT=true;YAHOO.util.Event.onDOMReady(function () {sessionMgr.initSession();sessionMgr.loadFromBookmarks();});var adPrefetch = true;var isAdPrefetched = false;if (document.getElementById("JSMailOptionsLink")) {document.getElementById("JSMailOptionsLink").href = optionObj["JSMailOptionsLink"];}if (document.getElementById("chk_pop_options") || document.getElementById("chk_pop_error") || document.getElementById("option_bad_folder") || document.getElementById("folders_options_spam") || document.getElementById("folders_options_sent") || document.getElementById("spam_edit_options")) {YAHOO.util.Event.addListener(YAHOO.mc.optionsID, "click", displayOptions, optionObj);}YAHOO.namespace("fcue"); YAHOO.mc.FCUE_METADATA = {imInit:"",imSent:"",fcElap:"1254813133500",fcEcap:"2",fcNag:"",inEcap:"",inElap:""};YAHOO.mc.noCues = false;YAHOO.mc.msgrLoaded = false;YAHOO.mc.loadMsgr = function () {clearTimeout(YAHOO.mc.msgrLoadID);if (hasIM && !YAHOO.mc.msgrLoaded) {YAHOO.mc.msgrLoaded = true;YAHOO.util.Get.script("http://mail.yimg.com/d/combo?/mg/8_0_4/js/im_blue_all.js&/mg/8_0_4/js/us_st…");}};YAHOO.util.Event.on(window, "load", YAHOO.mc.loadMsgr);YAHOO.mc.msgrLoadID = setTimeout(YAHOO.mc.loadMsgr, 5000); if (YAHOO.mc.page.galaxy) { YAHOO.mc.page.galaxy.cleanup(); } YAHOO.mc.page.galaxy = new GalaxyHandler(); function myMsgHandler(fid, mid) { YAHOO.mc.page.galaxy.evtCallBack("showMessage?fid=" + fid + "&mid=" + mid + "&.rand=635860173"); } YAHOO.galaxy.setMsgListener(myMsgHandler); function myFolderHandler(fid, filterBy) { var filterUrl = ""; if (filterBy) { filterUrl += "&filterBy=" + filterBy; } YAHOO.mc.page.galaxy.evtCallBack("showFolder?fid=" + fid + filterUrl + "&.rand=635860173"); } YAHOO.galaxy.setFolderListener(myFolderHandler); gLaunchProfile.stop('LD_JS_BP', false);resizeLeftPane();gLaunchProfile.start('RT_AD_FOOT');

2 1

seg fault
by John Robertson 22 Oct '09

22 Oct '09

The following compiles and runs. typedef string Colour; struct edge_properties { vector < set< Colour > > eColours; } //get a reference to the vector of the sets of colours of some edge. vector < set<Colour> > & v_colour_outset = get(m_eColours_map,*out_iter); // Here *out_iter is some edge. //Add a set<Colour> to v_colour_outset ie. to get(m_eColours_map,*out_iter) . set<Colour> colour_set; v_colour_outset.push_back(colour_set); // get a reference to one of the set<colour> in get(m_eColours_map,*out_iter) . set<Colour> & outcset = v_colour_outset[1]; //or [0], or any elt. But when I try to add a new colour to this set outcset.insert("Blue"); I get segfault. My guess is that there is a predetermined size for colour_set, and when it is push_backed into v_colour_outset, a new colour cannot be added into it. I tried to create a vector of set<Colour> references as follows. struct edge_properties { vector < set< Colour > & > eColours; } But it did not compile. ( shouldn't it? ) Whats the best solution to this problem. Thanks - JR

1 0

[fusion] is there anything resembling a joint_view of 2 maps?
by er 22 Oct '09

22 Oct '09

Hi All, I am wondering if anyone knows of anything that resembles a joint_view of a pair of maps. typedef mpl::int_<1> key1_; typedef mpl::int_<2> key2_; typedef pair<key1_,int> pair1_; typedef pair<key2_,int> pair2_; typedef map<pair1_> map1_; typedef map<pair2_> map2_; map1_ m1( make_pair<key1_>(1) ); map2_ m2( make_pair<key2_>(2) ); typedef joint_view< map1_, map2_ > jv_; // not supposed to compile (?), yet it does: jv_ view(m1,m2); // but not this: // at_key<key1_>(view) // I have no choice (?) but create a new map, // although view helps a bit for initializing it: typedef map< pair1_, pair2_ > map3_; map3_ m3 = fusion::as_map(view); int n1 = fusion::at_key<key1_>(m3); BOOST_ASSERT(n1 == 1); int n2 = fusion::at_key<key2_>(m3); BOOST_ASSERT(n2 == 2);

3 6

[conversion] Release v0.3 : Changing the order of assign_to parameters + `tie' function
by vicente.botet 22 Oct '09

22 Oct '09

Hi, I have just released the version 0.3 of Boost.Conversion , which includes since my last publicly announced version 0.1: Version 0.3.0, Oct 22, 2009 Changing the order of to and from parameters on assign_to + Added `tie' function Incopatibility: * Changing the order of to and from parameters on assign_to. * Now boost/conversion/convert_to.hpp and boost/conversion/assign_to.hpp files are separated. New Features: * Added <boost/conversion.hpp> global file. * Added mca() function. * Added convert_to_via function. Test: * Added test for the new features Version 0.2.0, Mai 16, 2009 Adding array + fusion::tuples + Adaptation to Boost 1.39 New Features: * conversion between boost::array of explicitly convertible types. * conversion between Boost.Fusion sequences of explicitly convertible types. You can found all the detail for dowloading, documentation, ... in http://svn.boost.org/trac/boost/wiki/LibrariesUnderConstruction#Boost.Conve… I'd appreciate it if people could take a look, test it on their compiler and give me some feedback. Best regards, _____________________ Vicente Juan Botet Escribá

2 1

[Asio] Any port with an endpoint
by Ryan McConnehey 22 Oct '09

22 Oct '09

In windows there is the capability to create a socket with an address but leave the port number up to windows to select. Is this possible with the asio endpoint or will I always need to define "some" port number? Ryan

2 2

Re: [Boost-users] [signals2][review] The review of the signals2 library (formerly thread_safe_signals) begins today, Nov 1st
by Nat Goodspeed 22 Oct '09

22 Oct '09

[Sorry, I know I'm replying to a very old post, but this IS continuing the same exchange.] Frank Mori Hess wrote: > > On Wednesday 18 February 2009, Nat Goodspeed wrote: >> [snip - talking about boost::signals2::signal::connect() accepting a boost::bind expression containing a boost::shared_ptr] >> the >> shared_ptr copy stored in the boost::bind() result makes the referenced >> object effectively immortal. >> >> To my surprise, I find that it's not destroyed even when I explicitly >> disconnect the resulting connection. > It will get destroyed eventually. The signal cleans up its slot list little > by little during connect/invoke. It doesn't immediately remove disconnected > slots from the slot list since other threads might be using the same slot > list concurrently. It might be possible to make it immediately reset the > shared_ptr owning the slot though, leaving an empty shared_ptr in the slot > list, since that wouldn't invalidate any iterators. Frank, has this logic changed at all since Boost 1.39? I've bumped into a situation where the lazy cleanup is causing a crash. We've got a DLL containing a slot stored in a signal used by the main program. Before attempting to unload the DLL, we disconnect the slot -- but the signal retains a reference to memory invalidated by the departure of the DLL, resulting in badness. Would updating to Boost 1.40 solve this problem? Or is there some explicit cleanup call we could/should make with a 1.39 boost::signals2::signal? Or could you suggest how to patch the logic as you describe above? Many thanks.

4 6

shared_ptr and inheritance
by Daniele Barzotti 22 Oct '09

22 Oct '09

Maybe this is a simple question but... Suppose to have: template<typename T> class Base {...}; template<typename T> class Derived : public Base<T> {...}; and shared_ptr< Base<T> > BasePtr; shared_ptr< Derived<T> > DerivedPtr; How can assign BasePtr to DerivedPtr? Thanks Daniele.

2 2

using pairs with boost.foreach
by Sean Farrow 22 Oct '09

22 Oct '09

Hi: Is it possible to use an std::pair with boost.foreach? I'm getting all sorts of errors when trying to do this. Any help appreciated.

4 4

need help using boost bzip2 filter on Windows (or zlib filter)
by ChristinaDRS 22 Oct '09

22 Oct '09

Hello all, I have the following code simple code to compress/uncompress a string: #include <iostream> #include <boost/iostreams/copy.hpp> #include <boost/iostreams/device/back_inserter.hpp> #include <boost/iostreams/filter/bzip2.hpp> #include <boost/iostreams/filtering_stream.hpp> using namespace std; int main() { string message = "this is a test message!"; cout << "message: " << message << endl; string compressed; boost::iostreams::filtering_streambuf<boost::iostreams::output> out; out.push(boost::iostreams::zlib_compressor()); out.push(boost::iostreams::back_inserter(compressed)); boost::iostreams::copy(boost::make_iterator_range(message), out); cout << "compressed message: " << compressed << endl; boost::iostreams::filtering_streambuf<boost::iostreams::input> in; in.push(boost::iostreams::zlib_decompressor()); in.push(boost::make_iterator_range(compressed)); string decompressed; boost::iostreams::copy(in, boost::iostreams::back_inserter(decompressed)); cout << "decompressed message: " << decompressed << endl; } but I get the following linking error: 1>------ Build started: Project: bzip2_compression_test, Configuration: Debug Win32 ------ 1>Compiling... 1>bzip2_test.cpp 1>Linking... 1>LINK : fatal error LNK1104: cannot open file 'libboost_bzip2-vc90-mt-gd-1_39.lib' 1>Build log was saved at "file://d:\compression\BoostCompression2\Debug\BuildLog.htm" 1>bzip2_compression_test - 1 error(s), 0 warning(s) ========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ========== I tried following the instructions on this page: http://www.boost.org/doc/libs/1_40_0/libs/iostreams/doc/installation.html so I defined these environment variables: BZIP2_BINARY = libbz2.lib BZIP2_INCLUDE = D:\bzip2-1.0.5 BZIP2_LIBPATH = D:\bzip2-1.0.5\Debug BZIP2_SOURCE = D:\bzip2-1.0.5 I also tried a similair example with zlib and I get the same error: 1>------ Build started: Project: libz_compression_test, Configuration: Debug Win32 ------ 1>Compiling... 1>zlib_test.cpp 1>Linking... 1>LINK : D:\Debug\libz_compression_test.exe not found or not built by the last incremental link; performing full link 1>LINK : fatal error LNK1104: cannot open file 'libboost_zlib-vc90-mt-gd-1_39.lib' 1>Build log was saved at "file://d:\\compression\libz_compression_test\Debug\BuildLog.htm" 1>libz_compression_test - 1 error(s), 0 warning(s) ========== Build: 0 succeeded, 1 failed, 0 up-to-date, 0 skipped ========== Any ideas how I can fix this? Thanks! Christina -- View this message in context: http://www.nabble.com/need-help-using-boost-bzip2-filter-on-Windows-%28or-z… Sent from the Boost - Users mailing list archive at Nabble.com.

4 6

[Boost::Interprocess] access violation when function returns a pointer to shared memory object
by BaDBuTz 22 Oct '09

22 Oct '09

hi, I want to synchronize multiple Processes with boost::Interprocess. For the start I implemented a simple Synchronisation Class (with some mutexes) in my test program that will be stored in the Shared Memory. The first call of print() (in function getSyncObject) returns the correct output. But why do I get an access violation, when I call the function again (the pointer is the same!) in main()? I am using WINXP, msvc 7-1 and mingw44 (with both compilers I get the same error) Thanks in advance Martin (compilable) Code: #include <iostream> #include <boost/interprocess/managed_shared_memory.hpp> #include <boost/interprocess/allocators/allocator.hpp> #include <boost/interprocess/containers/string.hpp> typedef unsigned long U32; class SyncObject { public: SyncObject(std::string aId) : id(aId), locAddrSp0Range(0), locAddrSp0Remap(0), locAddrSp1Range(0), locAddrSp1Remap(0), usbInterfaceOwner(0), useCounter(0) {}; ~SyncObject() {} ; void print() { std::cout << id << " locAddrSp0Range = " << locAddrSp0Range << " locAddrSp0Remap = " << locAddrSp0Remap << " locAddrSp1Range = " << locAddrSp1Range << " locAddrSp1Remap = " << locAddrSp1Remap << " usbInterfaceOwner = " << usbInterfaceOwner << " useCounter = " << useCounter << "\n"; }; std::string id; boost::interprocess::interprocess_mutex mtxReadWrite; boost::interprocess::interprocess_mutex mtxFunction; U32 locAddrSp0Range; U32 locAddrSp0Remap; U32 locAddrSp1Range; U32 locAddrSp1Remap; U32 usbInterfaceOwner; U32 useCounter; }; class SharedMemory { public: SharedMemory(); ~SharedMemory(); SyncObject* getSyncObject(std::string id); }; using namespace boost::interprocess; const char* MANAGED_SHARED_MEMORY_NAME = "sharedMemoryMapTest17"; SharedMemory::SharedMemory() {} SharedMemory::~SharedMemory() {} SyncObject* SharedMemory::getSyncObject(std::string id) { managed_shared_memory segment(open_or_create, MANAGED_SHARED_MEMORY_NAME, 65536); SyncObject* ptr = segment.construct<SyncObject>(anonymous_instance)(id); ptr->print(); //works! return ptr; } int main() { SharedMemory shm; SyncObject* ptr = shm.getSyncObject("idName"); ptr->print(); //Application crashes here with an access violation return 0; }

2 1

multiple or single joins on asio socket
by Lee MacDonald 22 Oct '09

22 Oct '09

Guys, I have a application running on Linux using boost 1.35, that receives packets to several different multicast addresses, but on the same port and interface. For each different multicast address I perform the following: // create a new socket socket_(ioService) // set the listen end point to that of the multicast address and port boost::asio::ip::udp::endpoint listenEndPoint(multicastAddress.to_v4(), port); // we need to be able to reuse the underlyling socket created as more than one asio socket will share it. socket_.set_option(boost::asio::ip::udp::socket::reuse_address(true)); socket_.bind(listenEndPoint); // perform the actual join to the multicast group socket_.set_option(boost::asio::ip::multicast::join_group(destAddress.to _v4(), interfaceAddress.to_v4())); So I bind to a seperate asio socket instance for each multicast address and then perform a join to that address, hence each asio socket only has a single join associated with it. Is a better approach to have a single asio socket instance and perform a join for each multicast address that shares the port & interface? If so, how would I set listenEndPoint? Is there any performance differences between the two? Are there better approaches? Thanks. Lee MacDonald Schneider Trading 4th Floor, 25 Copthall Avenue London EC2R 7BP -------------------------------------------------------- Message from: lee.macdonald(a)schneidertrading.com Message to: boost-users(a)lists.boost.org Attached files: 0 This e-mail and any files transmitted with it are confidential, may be legally privileged, and are for the sole use of the intended recipient. Copyright in this e-mail and any accompanying document created by us is owned by us. If you are not the intended recipient of this e-mail or any part of it please telephone our IT Department at the number shown above or by e-mail at info(a)schneidertrading.com. You should not use or disclose to any other person the contents of this e-mail or its attachments (if any), nor take copies. This e-mail is not a representation or warranty and is not intended nor should it be taken to create any legal relations, contractual or otherwise. Schneider Trading Associates Ltd (Registration No. 3692131) is incorporated in England and Wales and the registered office is at 25 Copthall Avenue, London EC2R 7BP. The company is authorised and regulated by the Financial Services Authority

1 0

[Thread] join_all and interrupt_all playing together
by Ryan McConnehey 22 Oct '09

22 Oct '09

I'm trying to set several threads running and have the main thread signal an interrupt. I've included an example that shows what I'm trying to accomplish. The threads that are created are associated with thread_group and then a join_all is called. I would expect the main thread to be able to call interrupt_all on the thread_group. Since this wasn't working I looked at what the join_all and interrupt_all function were actually doing. It turns out that thread_group has a mutex that is being called for each function and since the mutex for join_all is in use, the lock in interrupt_all is blocking. I understand why the mutex is there but am unclear how interrupt_all is suppose to work with join_all. To accomplish the functionality I need I have to use create_thread and store the return pointer in a list of my own. Then call interrupt on each pointer I stored. Which is basically duplicating the storage of thread_group. This seem like a waste. I would appreciate any advice. Ryan class Converter { public: Converter(void) { m_Threads.reset(new boost::thread_group); } void run(void) { m_Threads->add_thread(new boost::thread(boost::bind(&Converter::processSureTrakMessage, this))); //m_Threads_v.push_back(m_Threads->create_thread(boost::bind(&Converter::processSureTrakMessage, this))); m_Threads->join_all(); m_Threads.reset(new boost::thread_group); } void stop(void) { //for (std::vector<boost::thread *>::iterator iter = m_Threads_v.begin(); iter != m_Threads_v.end(); ++iter) { // (*iter)->interrupt(); //} m_Threads->interrupt_all(); } void processMessage(void) { try { for (;;) { boost::this_thread::interruption_point(); } } catch (...) { int bob = 1; } int bobo = 2; } private: boost::scoped_ptr<boost::thread_group> m_Threads; boost::scoped_ptr<boost::thread> m_IO_Service_Thread; std::vector<boost::thread *> m_Threads_v; }; boost::function<void ()> console_ctrl_function; BOOL WINAPI console_ctrl_handler(DWORD ctrl_type) { switch (ctrl_type) { case CTRL_C_EVENT: case CTRL_BREAK_EVENT: case CTRL_CLOSE_EVENT: case CTRL_SHUTDOWN_EVENT: console_ctrl_function(); return TRUE; default: return FALSE; } } int main(int argc, char * argv[]) { Converter convert; console_ctrl_function = boost::bind(&Converter::stop, &convert); SetConsoleCtrlHandler(console_ctrl_handler, TRUE); convert.run(); return 0; }

2 1

Warnings in hash_table_impl.hpp with GCC 4.4.x
by Marco Alanen 22 Oct '09

22 Oct '09

Greetings I'm trying to compile our code in an environment where we treat warnings as errors. This worked fine when we used GCC 4.1.2 which is fairly forgiving when compared to newer versions. However, after upgrading to GCC 4.4.1 (and 4.4.2) we ran into a warning in hash_table_impl.hpp in Boost 1.40.0: .../include/boost/unordered/detail/hash_table_impl.hpp: In member function 'void boost::unordered_detail::hash_table_data_equivalent_keys<Alloc>:: node_constructor::construct_preamble() [with Alloc = std::allocator<std::pair <const unsigned int, GFL::refcounted_value<const std::basic_string<char, std:: char_traits<char>, std::allocator<char> > > > >]': .../include/boost/unordered/detail/hash_table_impl.hpp:101: error: '<anonymous>' may be used uninitialized in this function Is this problem patchable, or will it manifest itself throughout the code? One way would of course be to turn that particular warning off but I'd prefer to have all warnings enabled if possible. I also tried to compile with Boost 1.39.0 with the same results. Thank you /Marco

2 2

[function] generalization of std::plus and std::multiplies with more than one argument type
by alfC 22 Oct '09

22 Oct '09

Does Boost have a generalization of std::plus and std::multiplies (std_function.hpp)? The standard defines this with only one template parameter template <class _Tp> struct multiplies : public binary_function<_Tp, _Tp, _Tp> { _Tp operator()(const _Tp& __x, const _Tp& __y) const { return __x * __y; } }; but I need something more general, where the types are different multiplies<type1, type2, type2> (assuming that the result has the type of the rhs. of course I could just define template <class _Tp1, class _Tp2, class Tp3> struct multiplies : public binary_function<_Tp1, _Tp2, _Tp3> { _Tp3 operator()(const _Tp1& __x, const _Tp2& __y) const { return __x * __y; } }; but I don't want to clutter my code with these pieces of basic code, specially if something like this is already part of other facilities in Boost. (Or even better, it would be nice if multiples<type3(type1, type2)> were defined, maybe I am asking to much sugar). I need this specific kind of multiply<3 args> in order to provide a default multiplication in the template parameters of vector space class, i.e. unless something is explicitly defined it tries to use * to implement scalar (pre)multiplication. (multiply<3 args> is the only reasonable default I can think of). Thanks, Alfredo

5 9