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

[Boost] formal release dates for 1.34
by Andy Stevenson 15 Dec '06

15 Dec '06

Hi, Got any firm dates for when 1.34 will be released formally as the main boost distribution and what will/will not be in it in terms of libraries? Andy Stevenson Andy Stevenson CEO Fujitsu Telecommunications Europe Limited Tel: +44 121 717 6440 P Please consider the environment - do you really need to print this email?

1 0

composite key, want to use compatible key on part of it
by bert hubert 15 Dec '06

15 Dec '06

Hi everybody, I have this boost::multi_index container (I'm very very happy with boost::multi_index!): typedef multi_index_container< CacheEntry, indexed_by < ordered_unique< composite_key< CacheEntry, member<CacheEntry,string,&CacheEntry::d_qname>, member<CacheEntry,uint16_t,&CacheEntry::d_qtype> >, composite_key_compare<CIStringCompare, std::less<uint16_t> > >, sequenced<> > > cache_t; For details on CacheEntry, see http://wiki.powerdns.com/cgi-bin/trac.fcgi/browser/trunk/pdns/pdns/recursor… What I want to do is use a compatible key ('char *') to search only on d_qname. I've tried: struct string_char_comp { bool operator()(const char* x, const tuple<string,uint16_t>& )const { return false; // bogus, just compile testing } bool operator()(const tuple<string,uint16_t>&, const char* x )const { return false; } }; Followed by: cache_t d_cache; // ... d_cachecache=d_cache.equal_range(qname.c_str(), string_char_comp()); But this doesn't compile with famously huge template errors (I've shortened boost::multi_index to bi::): ord_index_ops.hpp:87: error: no match for call to `(const string_char_comp) (bi::composite_key_result<bi::composite_key<MemRecursorCache::CacheEntry, bi::member<MemRecursorCache::CacheEntry, std::basic_string<char, std::char_traits<char>, std::allocator<char> >, &MemRecursorCache::CacheEntry::d_qname>, bi::member<MemRecursorCache::CacheEntry, short unsigned int, &MemRecursorCache::CacheEntry::d_qtype>, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type> >, const char* const&)' These errors make me wonder if what I want is even possible - unless I make my comparison function accept a composite_key_result with lots of bi internals in it. Joaquin? -- http://www.PowerDNS.com Open source, database driven DNS Software http://netherlabs.nl Open and Closed source services

2 2

[Graph] clear_vertex() bug?
by Silvia Richter 15 Dec '06

15 Dec '06

Hi, I'm new to Boost, and hope you can help me. When the attached source file is compiled with: gcc version 3.2.3 20030502 (Red Hat Linux 3.2.3-47.fc4) or gcc version 3.3.6 (Ubuntu 1:3.3.6-13ubuntu2) or gcc version 4.0.4 20060630 (prerelease) (Ubuntu 4.0.3-4), then running the program gives a segfault on line 27. When compiled with: gcc version 4.0.2 20051125 (Red Hat 4.0.2-8), no error occurs. To my understanding, no error should occur: I am using "listS" for both edges and vertices in adjacency_list. According to the documentation, clear_vertex() should not invalidate the adjacency_iterator in this case. However, in my code that's exactly what happens. I call adjacent_vertices() on a vertex A, which gives me an adjacency_iterator. Then I remove all edges from A by calling clear_vertex(). Then I want to use the adjacency_iterator obtained previously to check the new degrees of all (previous) neighbours of the removed vertex v. But the program crashes with a segmentation fault, when I try to use the iterator. Has anyone experienced this before? Is that a bug of the library? Cheers, Silvia -------------------------------------------------------------------------- This email and any attachments may be confidential. They may contain legally privileged information or copyright material. You should not read, copy, use or disclose them without authorisation. If you are not an intended recipient, please contact us at once by return email and then delete both messages. We do not accept liability in connection with computer virus, data corruption, delay, interruption, unauthorised access or unauthorised amendment. This notice should not be removed.

1 0

Re: [Boost-users] [bjam]BorlandC++BuilderX problem staging Boost
by Jerry Schreiber 14 Dec '06

14 Dec '06

Rene Rivera wrote: >Use: > bjam -sTOOLS=borland stage >Or possibly: > bjam -sTOOLS=borland-5_6_4 stage >Or perhaps you just made a mistake in typing in "Borland" instead of >"borland" above? I tried "borland", worked the same as "Borland" which I used before, but "borland-5_6_4" made a substantial difference. Rene Rivera wrote: >You need to add that directory to your library search path when linking >your own programs. I've added the C:\Boost\boost_1_33_1\libs to the "-L" library search path (only had the parent directory previously). But I still get: Fatal: Unable to open file 'LIBBOOST_FILESYSTEM-BCB-S-1_33_1.LIB' In examining the output from the bjam using borland-5_6_4, I now see references to "HardLink stage\lib\libboost_filesystem-bcb56-s.lib", but no references to "libboost_filesystem-bcb-s-1_1_33_1.lib". I also see several "...skipped <@boost!lib>boost_filesystem-bcb56.lib for lack of <@boost!lib>boost_filesystem-bcb56-1_33_1.lib..." I'll attach all of the bjam output which I was able to trap. Jerry Schreiber

2 1

Boost: issue with Threads and Mutexes
by Séverin Lemaignan 14 Dec '06

14 Dec '06

Hello, I'm Severin Lemaignan, software engineer at the INRIA french public research center. We are working on automation in public transportation. All our project is currently a pretty big C++/CLI-based piece of software, and we are relying more and more on Boost for low-level operations like threading. We ran very recently into a problem with Boost mutexes: We compile the project with Visual Studio 2005 (with the /clr flag), and everything was fine until we (re-)enabled precompiled headers. After that, the program still compiled and linked but it crashed as soon as started with a 0xc000007b error (namely "The application failed to initialize properly"). After futher investigation, we discovered it's due to the way we use "#pragma unmanaged" and "#pragma managed". The Boost headers were simply included in a "managed" part of the code while Boost methods (especially mutexes) were called in a "unmanaged" part. Hence causing the crash. The Boost library never complained at compile time. Maybe it should detect inconsistent use of managed/unmanaged pragma and warn the user ? We found another pratical (but unexpected) solution to the issue: instanciate a Boost::thread before the mutex. It solved the run-time crash. If someone has an explaination... Best regards, Séverin Lemaignan

3 2

in-depth iterator
by Guillaume Chereau 14 Dec '06

14 Dec '06

Hello everyone, I have a question concerning the boost::iterator library. Here is my situation : I have a Tree structure that looks like that : template<class T> struct Node : public std::vector<Node<T> > { T value; }; I'd like to create an iterator that would yield all the values of such a structure. And also being able to prevent the iterator from going into some branches. The interface should look like this : template<class T> struct Policy { bool go_inside(const Node& node) { return True; }; }; Node<int> my_tree; ... depth_iter<int, Policy> iter(my_tree, Policy()); for (; depth_iter++; depth_iter != my_tree.end()) { .... } I can't find an easy way to create this "depht_iter" class. Doesn't the boost::iterator library gives some way to create in-depth iterators ? Is there a way to use the boost::graph library ? Thanks in advance Guillaume

2 1

updating mpl::map to 1.34
by Ovanes Markarian 14 Dec '06

14 Dec '06

Hi! I would like to update mpl::map and mpl algorithms to 1.34 release candidate libs. What is the best way to do it? If there an archive of currently tested boost libs available or which CVS label should I use for new mpl version. Is it sufficient to replace boost/mpl directory in the installed boost libs? Currently I have problems when iterating (mpl::for_each, mpl::max_element) through mpl::map sequences. With Kind Regards, Ovanes Markarian

2 11

threads / scoped_lock
by Goroll, Torsten 14 Dec '06

14 Dec '06

Hi, as I understand the scoped_lock, it locks every ressource in a code block. What if I dont want to lock a ressource all the time the method is running? Do I have to create a new method? Example: boost::mutex a_mutex; ... { boost::mutex::scoped_lock lock(a_mutex); cou<<"bla"; //much of code where cout dont has to be locked ... } Torsten -----Ursprüngliche Nachricht----- Von: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org]Im Auftrag von boost-users-request(a)lists.boost.org Gesendet: Donnerstag, 14. Dezember 2006 14:55 An: boost-users(a)lists.boost.org Betreff: Boost-users Digest, Vol 1117, Issue 1 Send Boost-users mailing list submissions to boost-users(a)lists.boost.org To subscribe or unsubscribe via the World Wide Web, visit http://lists.boost.org/mailman/listinfo.cgi/boost-users or, via email, send a message with subject or body 'help' to boost-users-request(a)lists.boost.org You can reach the person managing the list at boost-users-owner(a)lists.boost.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Boost-users digest..."

5 5

Re: [Boost-users] [bjam]BorlandC++BuilderX problem staging Boost
by Jerry Schreiber 14 Dec '06

14 Dec '06

I have experimented further with bjam input syntax, trying to follow the examples in http://www.boost.org/tools/build/v1/build_system.htm. Since I am currently interested in the filesystem library only, I added "--with-filesystem". This reduced the amount of output so that the CMD window's capacity was not exceeded. I discovered that the first thing going wrong was: borland-C++-action bin\boost\libs\filesystem\build\boost_filesystem.dll\borland- 5_6_4\debug\exception.obj C:\Boost\boost_1_33_1\libs\filesystem\build\../src/exception.cpp: Fatal F1003 C:\Boost\boost_1_33_1\boost/filesystem/config.hpp 23: Error directive: Dynamic linking Boost.Filesystem does not work for Borland; use static linking instead *** 1 errors in Compile *** I have tried variations on -sBUILD=runtime-link-static and -sBUILD=<runtime-link>static, but to no avail; I keep getting the message that the borland-C++-action is requesting Dynamic linking, which does not work with Borland. What is the precise syntax to get only static linking from bjam? Jerry Schreiber

1 0

building 64-bit Windows using MVC 8.0
by Rusty Jones 14 Dec '06

14 Dec '06

Hi, I have found some information on building 64-bit versions of boost along with people's experience, but I don't seem to have a clear understanding of the state of 64-bit support. 1. How do you build boost 1.33.1 to build as a 64-bit version in MVC 8.0? Can you use bjam or do you have to do something else? 2. Will version 1.34 support 64-bit windows? The regression tests for RC_1_34_0 used to have columns for 64-bit windows but I don't see them there anymore ( http://engineering.meta-comm.com/boost-regression/CVS-RC_1_34_0/developer/s… ). Any information would be appreciated. I need to make our software 64-bit compatible and getting 64-bit compatible boost is essential to do this. Cheers, Rusty

1 0

mapped_file_source
by Victor Morales-Duarte 14 Dec '06

14 Dec '06

Hi all, I'm trying to use boost::iostream::mapped_file_source directly, but I can't close and then re-open the file without getting an exception. Here is an example. ********************************************************************************************** #include <boost/iostreams/device/mapped_file.hpp> int main( void ) { std::string fileName( "SomeFile.ext" ); int osAlignment = boost::iostreams::mapped_file_source::alignment(); boost::iostreams::mapped_file_source fileMap( fileName, 2 * osAlignment, 0 ); // Load data to application memory using fileMap.data(); fileMap.close(); // The following line throws because fileMap.is_open() returns true fileMap.open( fileName, 1024 * osAlignment, 1024 * osAlignment ); // Load more data to application memory using fileMap.data(); return 0; } ***************************************************************************************** What I truly don't understand is why a call to is_open() woud return true immediately after a call to close(). What do I need to do to be able to map different parts of the same file? I can't map the whole file because the number of files that I need to process is large enough that I'd run out of allocation space for my process. Thanks guys. Victor I'm using boost 1.33.1 with Visual Studio 2003 on WinXP SP2. --------------------------------- Cheap Talk? Check out Yahoo! Messenger's low PC-to-Phone call rates.

1 0

Re: [Boost-users] threads / scoped_lock
by Goroll, Torsten 14 Dec '06

14 Dec '06

But why so simple if I could do it intricately too? ;o) Thank you -----Ursprüngliche Nachricht----- Von: Ovanes Markarian [mailto:om_boost@keywallet.com] Gesendet: Donnerstag, 14. Dezember 2006 15:20 An: boost-users(a)lists.boost.org Betreff: Re: [Boost-users] threads / scoped_lock Hi! you could define a scope where to lock ;) Just like this: boost::mutex a_mutex; ... { { boost::mutex::scoped_lock lock(a_mutex); //lock cout<<"bla"; }//end of scope => unlock //much of code where cout dont has to be locked ... } With Kind Regards, Ovanes Markarian On Thu, December 14, 2006 15:12, Goroll, Torsten wrote: > Hi, > > as I understand the scoped_lock, it locks every ressource in a code block. What if I dont want to > lock a ressource all the time the method is running? Do I have to create a new method? > > Example: > > boost::mutex a_mutex; > ... > { > boost::mutex::scoped_lock lock(a_mutex); > cou<<"bla"; > > //much of code where cout dont has to be locked > ... > } > > Torsten > > -----Ursprüngliche Nachricht----- > Von: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org]Im Auftrag von > boost-users-request(a)lists.boost.org > Gesendet: Donnerstag, 14. Dezember 2006 14:55 > An: boost-users(a)lists.boost.org > Betreff: Boost-users Digest, Vol 1117, Issue 1 > > > Send Boost-users mailing list submissions to > boost-users(a)lists.boost.org > > To subscribe or unsubscribe via the World Wide Web, visit > http://lists.boost.org/mailman/listinfo.cgi/boost-users > or, via email, send a message with subject or body 'help' to > boost-users-request(a)lists.boost.org > > You can reach the person managing the list at > boost-users-owner(a)lists.boost.org > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Boost-users digest..." > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

BOOST_ASSERT and others
by abir basak 14 Dec '06

14 Dec '06

Hi, Hi, if any assert is there like BOOST_ASSERT or assert where I can pass a custom message to print on assert failure ? i.e. something like BOOST_ASSERT(condition, message) Also, is there any debug print version for boost? Something like, I want a message should be printed only when the application is running in DEBUG version (whatever flag the system provides for that). The message can be cout or printf kind of thing. A little safety is better sometimes like debug print can't change the value of the variable etc. Thanks -- Abir Basak, Member IEEE Software Engineer, Read Ink Technologies B. Tech, IIT Kharagpur email: abir(a)abirbasak.com homepage: www.abirbasak.com

2 1

A default constructor for Boost.MPI environment?
by Bo Peng 14 Dec '06

14 Dec '06

Dear boost users, I am using Boost.MPI for a python module so there is no argc, argv that can be provided to boost::parallel::mpi::environment(int & argc, char ** & argv) Because an environment object has to exist during the life time of the module, I have to define a global variable in an awkward way: int fake_argc = 0; char * fake_argv = ""; char ** ptr_argv = &fake_argv; boost::parallel::mpi::environment env(fake_argc, ptr_argv); Can a default constructor be provided? Bo

5 5

[bjam]BorlandC++BuilderX problems staging Boost
by Jerry Schreiber 14 Dec '06

14 Dec '06

I am entirely new to Boost. I used the following command: bjam -sTOOLS=Borland stage I got a large amount of output with lots of errors indicated, but I did get quite a bit created in C:\Boost\boost_1_33_1\bin\boost\libs. As I am currently interested in using the filesystem library, I attempted to build the simple_ls.cpp example file and got the following error on the link step: ilink32 -D -ap -Tpe -x -Gn -v -LC:\CBuilderX\lib;C:\Boost\boost_1_33_1 c0x32.objwindows\Debug_Build\simple_ls.obj,C:\Boost\boost_1_33_1\libs\filesystem\windows\Debug_Build\simple_ls.exe,, cw32.lib import32.lib,, Turbo Incremental Link 5.65 Copyright (c) 1997-2002 Borland Fatal: Unable to open file 'LIBBOOST_FILESYSTEM-BCB-SD-1_33_1.LIB' Is C++BuilderX supported by Boost, and if so, what is the proper command for bjam? Is there anything else I need to do in order to get the link to work? Thanks. Jerry Schreiber

2 1

Re: [Boost-users] HowTo Pass arguments to boost threads?
by Gabriel Ki 13 Dec '06

13 Dec '06

Hi, I am just trying the boost threads example from the link. #include <boost/thread/thread.hpp> #include <boost/thread/mutex.hpp> #include <boost/thread/tss.hpp> #include <iostream> boost::mutex io_mutex; boost::thread_specific_ptr<int> ptr; struct count { count(int id) : id(id) { } void operator()() { if (ptr.get() == 0) ptr.reset(new int(0)); for (int i = 0; i < 10; ++i) { (*ptr)++; boost::mutex::scoped_lock lock(io_mutex); std::cout << id << ": " << *ptr << std::endl; } } int id; }; int main(int argc, char* argv[]) { boost::thread thrd1(count(1)); boost::thread thrd2(count(2)); thrd1.join(); thrd2.join(); return 0; }// Output $ ./testme 10591400: 1 10591400: 2 10591400: 3 10591400: 4 10591400: 5 10591400: 6 10591400: 7 10591400: 8 10591400: 9 10591400: 10 Segmentation fault I have a few questions. 1) Why is id copied incorrectly in operator()? 2) Why does the count object get copied so many times? (i tried to write a copy constructor into count to see the id value) 3) I tried to add setstacksize in boost code and recompile. Is it a better way to do it? Thanks in advance. Gabriel ----- Original Message ---- From: Moshe Matitya <Moshe.Matitya(a)Kayote.com> To: boost-users(a)lists.boost.org Sent: Tuesday, December 12, 2006 3:40:03 AM Subject: Re: [Boost-users] HowTo Pass arguments to boost threads? Forever Kid wrote: > > What is the method boost prefers to pass arguments to boost > threads? (such as a pointer to some class > object) > > I saw the class thread_specific_ptr. > Is thread_specific_ptr designed for this purpose? > > Is there a recommended site with boost examples to check out? > Boost has good documentation, although I prefer to learn by > example and use strict documentation as I become more experienced. > > Thanks much, > Graham Take a look this introductory article on Boost.Threads by Bill Kempf: http://www.ddj.com/184401518 It addresses the question you're asking in a very clear, straightforward manner, with examples. Moshe _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

2 1

program_options variable_value as cast explodes
by Edward Z. Yang 13 Dec '06

13 Dec '06

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 I'm using program_options in a few programs that I'm doing, and I'd like to clarify what's going on at certain points. vm.count("key") appears to tell us the number of times a flag is specified, but the documentation here: <http://www.boost.org/doc/html/variables_map.html> doesn't seem to say anything about it. Where is it specified? Does abstract_variables_map actually inherit off of std map? Why don't we use !vm["key"].empty() instead? (slightly more verbose, but it's clearer about what it's doing). The vm["term"].as<int>() template seems to combust spectacularly (as in cause a stack dump) when it turns out that vm["term"] doesn't actually exist. This doesn't seem very robust, and according to the documentation, an exception ought to be thrown. Am I just misunderstanding the docs? What is the preferred method determining that no parameters were passed? I'm currently using argc == 1, which seems a bit hacky, and doesn't catch errant parameters. Thanks. This might seem a little pedantic, but I'm trying to do things right. :-) -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.1 (MingW32) Comment: Using GnuPG with Mozilla - http://enigmail.mozdev.org iD8DBQFFf2sIqTO+fYacSNoRAqz5AJ9nBOEz87v8H4HjRB6mM2NSvNSllQCaA+SE s0kcqXiQXy7+ID7ugotbDKc= =NvIj -----END PGP SIGNATURE-----

2 1

thread: scoped_try_lock
by Robert Marion 12 Dec '06

12 Dec '06

Hello all, I have many good examples of how to use a scoped_lock, but I can find almost none on a scoped_try_lock. Can anyone point me to some good examples of this? Is there any decent documentation for this? I found the boost docs, by the author, a bit difficult to follow. Thanks, Rob ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

2 1

[ANN] Boost.Time_series: feedback requested
by Eric Niebler 12 Dec '06

12 Dec '06

I'm pleased to announce the availability of a new library for computing with time series (http://en.wikipedia.org/wiki/Time_series) From the documentation: > The purpose of the Boost.Time_series library is to provide data structures, numerical operators and algorithms to operate on time series. A time series is a series of data points, sampled at regular intervals. The library provides numerous time series containers, each with different time/space trade-offs, and a hierarchy of concepts which allow the time series to be manipulated generically. The library also provides operators and algorithms which use the generic interfaces to perform calculations on time series and accumulate various statistics about them. > > Boost.Time_series does not yet contain all the algorithms one might want in order to perform full time series analysis. However, the key contribution of Boost.Time_series is the framework and the rich hierarchy of concepts with which such algorithms can be written to efficiently and generically process series of data with widely divergent in-memory representations and performance characteristics. Boost.Time_series provides several such series containers, as well as mechanisms for defining additional series types and algorithms that fit within the framework. Some examples of series types that are provided are: dense, sparse, piecewise constant, heaviside and others, as well as adaptors for providing shifted, scaled and clipped series views. The documentation can be found here: http://boost-sandbox.sf.net/libs/time_series/doc/html/index.html The code is in time_series.zip, which can be downloaded here: http://boost-consulting.com/vault/index.php?directory=Math%20-%20Numerics Any suggestions for improvements are most welcome. -- Eric Niebler Boost Consulting www.boost-consulting.com

2 2

bjam problems under cygwin/mingw
by Bo Peng 12 Dec '06

12 Dec '06

Dear boost-users, I compiled boost-1.33.1 under cygwin/mingw and experienced two problems: 1. using the .exe installer, build/jam_src/build.sh uses dos newlines. The latest cygwin can not execute this script because of this. dos2unix has to be used. 2. Before I figure out the dox2unix solution, I compiled bjam using msvc. The generated bjam.exe ignores -sTOOLSET=mingw and use msvc even when cl is not in the path. Cheers, Bo

3 2

Re: [Boost-users] shared_ptr with custom free function
by james.jones＠firstinvestors.com 12 Dec '06

12 Dec '06

From: "Phil Endecott" <spam_from_boost_users(a)chezphil.org> > I have a C library that returns pointers to structs, and has special > free functions that I must call to finalise and free these structs. Is > there any way that I can create a shared_ptr that will call this > special library free function when the last copy is destroyed? I was > hoping that there would be some way of achieving this by overloading > operator delete, but it seems not. Any ideas? I would think one solution would be simply to wrap your C struct in a C++ class with a constructor and a destructor, then use shared_ptr on the C++ struct. Example: C: struct blob { ... }; blob * create_blob (); void free_blob(blob *); C++: class foo { public: foo() : data(create_blob()) {} ~foo() { free_blob(data); } private: blob * data; }; Now use shared_ptr< foo > as you usually would. This also gives you the flexibility NOT to use shared_ptr if you choose. - James Jones Administrative Data Mgmt. Webmaster 375 Raritan Center Pkwy, Suite A Data Architect Edison, NJ 08837

2 1

[iostreams] : bug in <boost/iostreams/copy.hpp> ?
by Mathieu Peyréga 12 Dec '06

12 Dec '06

Hello, i'm a new user to these great boost libraries... I think i found a bug in <boost/iostreams/copy.hpp> (version 1.33.1) the current code starting line 80 is template<typename Source, typename Sink> std::streamsize copy_impl( Source& src, Sink& snk, std::streamsize buffer_size, mpl::false_, mpl::true_ ) { // Copy from an indirect Source to a direct Sink. using namespace std; typedef typename char_type_of<Source>::type char_type; typedef pair<char_type*, char_type*> pair_type; detail::basic_buffer<char_type> buf(buffer_size); pair_type p = snk.output_sequence(); streamsize total = 0; bool done = false; while (!done) { streamsize amt; done = (amt = iostreams::read(src, buf.data(), buffer_size)) == -1; std::copy(buf.data(), buf.data() + amt, p.first + total); if (amt != -1) total += amt; } return total; } I think it should be somthing like template<typename Source, typename Sink> std::streamsize copy_impl( Source& src, Sink& snk, std::streamsize buffer_size, mpl::false_, mpl::true_ ) { // Copy from an indirect Source to a direct Sink. using namespace std; typedef typename char_type_of<Source>::type char_type; typedef pair<char_type*, char_type*> pair_type; detail::basic_buffer<char_type> buf(buffer_size); pair_type p = snk.output_sequence(); streamsize total = 0; streamsize amt = 0; //bool done = false; while ((amt = iostreams::read(src, buf.data(),buffer_size))!= -1) { std::copy(buf.data(), buf.data() + amt, p.first + total); total += amt; } return total; } The current code happens to crash whereas the proposed version works properly... Did I miss something or a I right ? Regards, Mathieu -- http://matioupi.free.fr/

2 7

[random] Four-Tap Shift Register RNG
by Sebastian Weber 12 Dec '06

12 Dec '06

Hello everyone! I'm using quite a lot of Boost libraries and have by the time implemented some bits which enhance some parts of the library. E.g. I ported a four-tap shift register random number generator to the random number library. I herby post the code as it might be interesting for others since this type of random number generator is known to produce random numbers with very low 2 and 3 point correlations (see http://arxiv.org/abs/cond-mat/9710104 ). The code is partially taken from the gsl-library and extended to 64bit unsigned integers which gives random numbers of very high precision. However, you can use this code only with the latest random number library from Boost CVS as the numbers returned use the full range of 32bit or 64bit unsigned ints, which causes overflows with the code in boost 1.33.1 and 1.34_RC. So please have look at the code and submit it into CVS if the code is any good. Greetings, Sebastian Weber

1 1

Re: [Boost-users] HowTo Pass arguments to boost threads?
by Forever Kid 12 Dec '06

12 Dec '06

Thanks Moshe! This pretty much sums it up from that article: "An overloaded constructor takes a function object called with no arguments and returning nothing. This constructor starts a new thread of execution, which in turn calls the function object. " Very nice article too. I must have missed that when I searched Dobbs. Graham ----- Original Message ---- From: Moshe Matitya <Moshe.Matitya(a)Kayote.com> To: boost-users(a)lists.boost.org Sent: Tuesday, December 12, 2006 5:40:03 AM Subject: Re: [Boost-users] HowTo Pass arguments to boost threads? Forever Kid wrote: > > What is the method boost prefers to pass arguments to boost > threads? (such as a pointer to some class > object) > > I saw the class thread_specific_ptr. > Is thread_specific_ptr designed for this purpose? > > Is there a recommended site with boost examples to check out? > Boost has good documentation, although I prefer to learn by > example and use strict documentation as I become more experienced. > > Thanks much, > Graham Take a look this introductory article on Boost.Threads by Bill Kempf: http://www.ddj.com/184401518 It addresses the question you're asking in a very clear, straightforward manner, with examples. Moshe _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

1 0

java wrapper
by Andreas Fabri 12 Dec '06

12 Dec '06

Hello, are there any efforts underway to provide a wrapper similar to the Python wrapper ? andreas

1 0

HowTo Pass arguments to boost threads?
by Forever Kid 12 Dec '06

12 Dec '06

What is the method boost prefers to pass arguments to boost threads? (such as a pointer to some class object) I saw the class thread_specific_ptr. Is thread_specific_ptr designed for this purpose? Is there a recommended site with boost examples to check out? Boost has good documentation, although I prefer to learn by example and use strict documentation as I become more experienced. Thanks much, Graham ____________________________________________________________________________________ Yahoo! Music Unlimited Access over 1 million songs. http://music.yahoo.com/unlimited

2 1

Issue linking to static boost thread library (Newbie)
by Forever Kid 12 Dec '06

12 Dec '06

I am having a little trouble getting boost to load the static thread library. For some reason it wants to load the dynamic library instead of the static boost_thread library. Everything compiles and builds, but I get a message that it can't find the shared library, when I really want it linked statically. (see below) What did I miss? Thanks much for any help, Graham Boost version: 1.33.1 OS: Linux 2.6 Compiler: GCC 4.1.1 Compiler command line (debug): g++ -I/home/quirk/dev/libraries/boost_1_33_1/ -O0 -g3 -Wall -c -fmessage-length=0 -pthread -MMD -MP -MF"main.d" -MT"main.d" -o"main.o" "../main.cpp" Linker command line (debug): g++ -L/home/quirk/dev/libraries/boost_1_33_1/stage/lib -o"ManagedMakeProject" ./main.o -lpthread -lboost_thread-gcc-mt-d-1_33_1 Code: #include <cstdio> #include "boost/thread/thread.hpp" void foo() { printf("Hello MT World\n"); } int main(int argc, char* argv[]) { printf("Hello World\n"); boost::thread aThread(foo); aThread.join(); return 0; } Output: error while loading shared libraries: libboost_thread-gcc-mt-d-1_33_1.so.1.33.1: cannot open shared object file: No such file or directory ____________________________________________________________________________________ Have a burning question? Go to www.Answers.yahoo.com and get answers from real people who know.

3 4

[tuple] extend to more than 10 elements
by Gabriel Ki 11 Dec '06

11 Dec '06

Hi, Has anyone successfully extended tuple to contain more than 10 elements? Any pointer to do that would be appreciated. I found the combination of Loki and boost tuple is very amazing. I would love to match up their capacity. Thanks for the great library. Regards, Gabriel Ki ____________________________________________________________________________________ Any questions? Get answers on any topic at www.Answers.yahoo.com. Try it now.

2 1

using the fusion library
by craigp 11 Dec '06

11 Dec '06

i've been looking at the fusion library for a few days now, and i feel like i'm missing something (or just trying to make this fit for a use-case it's simply not suited for). i'm looking as using it for compile-time introspection for, let's say, serialization. one thing i was curious about was whether i can statically map each of my persistent types to a string (which could be used as the name of an xml element/attribute). i quickly determined that a map (of fusion-pairs) with more than just a few elements quickly overwhelms the compiler (msvc 7.1; raising the limit of 20 to ~50 just makes the compiler barf). any thoughts on this? fusion provides what appears to be a lot of infrastructure, but i can't seem to take advantage of most of it ('out of the box', anyway). the best i've come up with so far is to create a for_each_it algorithm which, instead of calling back the visitor function with the value, calls it back with the corresponding iterator instead. this seems to work ok, but it appears i'd need to hand-write each algorithm (as well as each sequence), and i don't think that's how it was intended to be used. ;-) any thoughts? am i just barking up the wrong tree? what kinds of applications/use-cases were in mind when fusion was designed? most things i've thought of would (i'd guess) be impractical with today's compilers. i think there's some connection with fusion and spirit... how does spirit avoid the compiler limits i seem to run into? thanks! --craig __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

3 2

Get Your BoostCon Session Proposals In!
by David Abrahams 11 Dec '06

11 Dec '06

In case you hadn't heard... The inaugural Boost conference takes place in Aspen, Colorado May 14-18, 2007! BoostCon 2007 will be hosted at Aspen Center for Physics, one of the most beautiful meeting sites in the world, and a great venue for collaboration and discovery. We've already developed an exciting lineup of speakers and topics with such notables as Sean Parent of Adobe and Scott Meyers on the agenda. Still, we'd like to bring in a broader range of ideas from the community, so we're extending the deadline for session proposals through Monday, December 17th. We're especially interested in seeing proposals from Boost library authors for tutorial sessions on their libraries. We'd also like to see proposals for experience reports from the user community. That said, don't limit yourself: there are more ideas on the session proposal page, and as it says there, "don't hold back a proposal just because it doesn't fit into a pigeonhole." Call for sessions: http://www.boost.org/more/BoostCon07_session_call.html Submit session proposals to the program committee at: boostcon-program(a)lists.boost-consulting.com See you in May! -- Dave Abrahams Boost Consulting www.boost-consulting.com

1 0

Is Boost Threads missing "pthread_cancel"-like Functionality?
by Offlein 11 Dec '06

11 Dec '06

Hi all, i'm missing something like PThread's pthread_cancel. Is there a Way to use Boost::Threads in a similar Way. I looked in the Docu and didn't find anything similar to pthread_cancel. Is there another Functionality which kind of "implement" this Behaviour of PThreads or am i blind and don't see where it's written in the Docu? I don't want to use anything else because Boost is a very good Collection of Libraries, which are very well written.

4 5

#include <boost/lambda/lambda.hpp>
by Andreas Wißmeier 11 Dec '06

11 Dec '06

Dear all, if I include this file I get the following error-message: d:\cvsatSixt\sixt_offer_src\boost\boost\lambda\detail\arity_code.hpp(66): error C2065: 'T' : undeclared identifier I am using Microsoft Visual C++ .NET version 1.0.3705 I am probably missing something very basic. Kind regards Andreas

3 2

boost function with lambda bind
by Santosh Joseph 11 Dec '06

11 Dec '06

I'm trying to use a boost function object within a lambda bind. This is my test program and the commented line does not compile. What i'm trying to do is iterate over a container of test_str structs which has one data member (a string) and get that string and insert that string into another string. #include <string> #include <set> #include <vector> #include <iostream> #include "boost/lambda/lambda.hpp" #include "boost/lambda/bind.hpp" #include <algorithm> #include "boost/function.hpp" using namespace std; struct test_str { test_str(string s): str(s){} string str; }; int main(int argc, char* argv[]) { using namespace boost::lambda; boost::lambda::placeholder1_type Arg; typedef pair<set<string>::iterator,bool> (set<string>::*INSERT)(const string&); set<string> str_con; INSERT insert_func = &set<string>::insert; vector<test_str> v2; v2.push_back( test_str("one") ); v2.push_back( test_str("two") ); boost::function<string&(test_str&)> f = bind(&test_str::str, Arg); //for_each( v2.begin(), v2.end(), bind( insert_func, var(str_con), var(f))); for_each( v2.begin(), v2.end(), bind( insert_func, var(str_con), bind(&test_str::str, Arg))); return 0; } This is part of the error message i get from a vs2003.net compiler. Any help would be appreciated. c:\santosh\tools\boost_1_33_1\boost\lambda\detail\function_adaptors.hpp(250): error C2664: 'std::pair<_Ty1,_Ty2> (Arg1)' : cannot convert parameter 1 from 'boost::function<Signature>' to 'const std::basic_string<_Elem,_Traits,_Ax> ' with [ _Ty1=std::_Tree<std::_Tset_traits<std::string,std::less<std::string>,std::allocator<std::string>,false>>::iterator, _Ty2=bool, Arg1=const std::string & ] and [ Signature=std::string &(test_str &) ] and [ _Elem=char, _Traits=std::char_traits<char>, _Ax=std::allocator<char> ] __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

boost::array as part of container.
by chun ping wang 11 Dec '06

11 Dec '06

hi i would like to know how to insert boost::array into containers such as vectors, list, etc, in an effective and easy way.. tuples don't work cause you can't get to random access index. Thanks. AES word example. typedef boost::array<int, 4> word; std::vector vecOfWord; // perform some word insertion.

5 8

Serialization/cvs can not serialize vector<bool>
by Bo Peng 11 Dec '06

11 Dec '06

Dear boost users, To test Boost.MPI, I upgrade from boost 1.33.1 to boost/cvs. I get error messages like: /usr/include/boost-1_35/boost/archive/detail/oserializer.hpp:151: instantiated from `void boost::archive::detail::oserializer<Archive, T>::save_object_data(boost::archive::detail::basic_oarchive&, const void*) const [with Archive = boost::archive::binary_oarchive, T = simuPOP::population]' /home/bpeng/simuPOP/release/build/population_ba.cpp:1567: instantiated from here /usr/include/boost-1_35/boost/serialization/detail/get_data.hpp:31: warning: taking address of temporary /usr/include/boost-1_35/boost/serialization/detail/get_data.hpp:31: error: cannot convert `std::_Bit_reference*' to `bool*' in return which looks like the old vector<bool>::reference problem. Is this a known problem? Does anyone know a solution? Many thanks in advance. Bo

2 3

Problem with multi_index
by Sunil Chomal 11 Dec '06

11 Dec '06

Dear All, I am a newbie with multi_index container. I need to store the following data structure in the multi_index_container: ************************************* #if !defined(NDEBUG) #define BOOST_MULTI_INDEX_ENABLE_INVARIANT_CHECKING #define BOOST_MULTI_INDEX_ENABLE_SAFE_MODE #endif #include <boost/multi_index_container.hpp> #include <boost/multi_index/member.hpp> #include <boost/multi_index/ordered_index.hpp> #include <algorithm> #include <iostream> #include <iterator> #include <string> using boost::multi_index_container; using namespace boost::multi_index; struct base { int base_data; }; struct child : public base { int child_data; }; struct base_data{ }; typedef multi_index_container < child, indexed_by< ordered_non_unique< tag<base_data>, BOOST_MULTI_INDEX_MEMBER(child, int, base_data) > > > child_set; int main() { child_set cs; return 0; } *************************************** If i try and compile the above code, I get the following error test.cpp:33: error: could not convert template argument &base::base_data to int child::* test.cpp:33: error: template argument 2 is invalid test.cpp:34: error: template argument 1 is invalid test.cpp:35: error: template argument 2 is invalid test.cpp:35: error: invalid type in declaration before ; token I have looked up the examples, but I can find any example for structs which are inherited. Can someone help? Regards, Sunil

2 1

Re: [Boost-users] Issue linking to static boost thread library (Newbie)
by Nat Goodspeed 11 Dec '06

11 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Forever Kid > Sent: Monday, December 11, 2006 12:54 AM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Issue linking to static boost thread library > (Newbie) > > I am having a little trouble getting boost to load the > static thread library. > > For some reason it wants to load the dynamic library > instead of the static boost_thread library. [Nat] I don't know if this is still applicable, but have you tried BOOST_THREAD_USE_LIB? This is the most pertinent message I've turned up: http://archives.free.net.ph/message/20040629.155158.2084b216.en.html

2 1

ptime construct from string
by Delvaille Florent 11 Dec '06

11 Dec '06

Hi, I ve a problem with ptime. I'm doing a a project on Eclipse, and I need to create a ptime object from a string. Here's my code: std::string ts("2002-01-20 23:59:59.000"); ptime t(time_from_string(ts)) This code is exactly the same in the boost documentation. On compilation, no error is detected. But there's a problem with the linker. I paste here the error that Eclipse gives to me: Invoking: GCC C++ Linker g++ -o"brol" ./initmain.o ./initmain.o: dans la fonction « unsigned short boost::date_time::month_str_to_ushort<boost::gregorian::greg_month>(std::basic_string<char, std::char_traits<char>, std::allocator<char> >)»: /usr/include/boost/date_time/date_parsing.hpp:67: référence indéfinie vers « boost::gregorian::greg_month::get_month_map_ptr() » collect2: ld returned 1 exit status Do you have an idea?

2 1

Re: [Boost-users] shared_ptr with custom free function
by Nat Goodspeed 11 Dec '06

11 Dec '06

From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Sohail Somani Sent: Sunday, December 10, 2006 9:17 PM To: boost-users(a)lists.boost.org; boost-users(a)lists.boost.org Subject: Re: [Boost-users] shared_ptr with custom free function From: boost-users-bounces(a)lists.boost.org on behalf of Nat Goodspeed > > [Nat] It might be worth binding that knowledge into a specific > > shared-pointer-to-blob type so that the required deleter function is > > implicitly captured EVERY time you construct such a pointer. > > ------------- > > How do you encode that knowledge into the type? [Nat] I admit I haven't tried coding this. My thought is to derive your "blob pointer" type from shared_ptr<blob>, implementing a subset of the constructor variants that will be useful to you, and ensuring that the constructor that accepts a dumb blob* pointer implicitly passes free_blob to the base-class constructor. Something like this: class blob; void free_blob(blob*); class blob_ptr: public shared_ptr<blob> { typedef shared_ptr<blob> super; public: // should only need to implement (a subset of) constructors // ... // This constructor is the reason blob_ptr is a class rather // than a typedef. Since we MUST use free_blob() to free a // heap blob instance, always bind free_blob() as shared_ptr's // custom deleter. blob_ptr(blob* ptr): super(ptr, free_blob) {} // ... };

1 0

Boost fusion: Bug in sample?
by Sohail Somani 11 Dec '06

11 Dec '06

Warning: MPL noob, still waiting for my book to arrive I'm reading the fusion library's documentation that was in the zip file for the announcement. If this is outdated, please ignore this email. The section in the quickstart: Print Only Pointers =================== Let's get a little cleverer. Say we wish to write a generic function that takes in an arbitrary sequence and XML prints only those elements which are pointers. Ah, easy. First, let's include the is_pointer boost type trait: #include <boost/type_traits/is_pointer.hpp> Then, simply: template <typename Sequence> void xml_print_pointers(Sequence const& seq) { for_each(filter_if<boost::is_pointer<Sequence> >(seq), print_xml()); } ==================== This doesn't seem to compile as is. I needed to do the following instead of the above: template <typename Sequence> void xml_print_pointers(Sequence const& seq) { using namespace mpl::placeholders; typedef mpl::lambda<boost::is_pointer<_1> >::type is_pointer; fusion::for_each(fusion::filter_if<is_pointer>(seq), print_xml()); } This is using boost 1.33.1. Did I miss something? Thanks! Sohail

2 3

How can I find what I didn't install
by Bo Yang 11 Dec '06

11 Dec '06

Hi, I complile and installed boost 1.31 library today, I followd the instructions in the website and I got part of it installed. I have seen some messages in the process of compile which said , something is skiped and something is failed. Now , I think part of boost have already installed. And I want to know , which part I didn't install, How can I get such information? Thank you !

3 3

Re: [Boost-users] shared_ptr with custom free function
by Nat Goodspeed 11 Dec '06

11 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Peter Dimov > Sent: Sunday, December 10, 2006 7:00 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] shared_ptr with custom free function > > > Is > > there any way that I can create a shared_ptr that will call this > > special library free function when the last copy is destroyed? > > Yes, there is: > > shared_ptr< blob > pb( create_blob(), free_blob ); [Nat] It might be worth binding that knowledge into a specific shared-pointer-to-blob type so that the required deleter function is implicitly captured EVERY time you construct such a pointer.

2 1

shared_ptr with custom free function
by Phil Endecott 10 Dec '06

10 Dec '06

Dear Boost experts, I have a C library that returns pointers to structs, and has special free functions that I must call to finalise and free these structs. Is there any way that I can create a shared_ptr that will call this special library free function when the last copy is destroyed? I was hoping that there would be some way of achieving this by overloading operator delete, but it seems not. Any ideas? Here's a more concrete example, in case it's not obvious: class Foo { public: Foo(struct blob* b): my_blob(b) {} ~Foo() {free_blob(my_blob);} private: struct blob* my_blob; }; Foo is not copyable, because the both copies will free the blob when they are destroyed. So: class Foo2 { public: Foo(struct blob* b): my_blob(b) {} private: boost::shared_ptr<struct blob> my_blob; }; Foo2 can be copied and the blob will be deleted exactly once when the last copy is destroyed, but the library's free_blob() function has not been used. inline void operator delete(struct blob* b) { free_blob(b); } Doesn't work; that's not how overloading of operator delete works. Thanks for any suggestions. Phil.

2 1

multi_index + pure abstract
by fungos 10 Dec '06

10 Dec '06

Doesn't multi_index accept hold classes that has pure abstract methods? Even if these methods aren't used by the container? Is there a way to get the container to hold objects with same interface? -- Animal Frontline Liberation

2 1

[filesystem] path to temporary file
by Chris Weed 10 Dec '06

10 Dec '06

Hi, Is there any support in filesystem to create paths to temporary files. I would like to make unit tests to test my file reader, by creating temporary files that are in some safe platform temp area that can be deleted at exit. Chris

1 0

Boost Serialization make_binary & XML/ASCII
by Terence Wilson 09 Dec '06

09 Dec '06

I need to parse an XML file generated by Boost Serialization, part of the file is a section generated by make_binary. I was wondering which scheme was used to convert binary to ASCII, uuencode, Base64? Obviously I need to know in order that I can convert back. TIA Dr. Terence Wilson Principal, Bayside Technical email: tez(a)latte.com web: http://latte.com

2 5

date-time New US daylight saving rules
by michael mathews 09 Dec '06

09 Dec '06

Hi All, Will /are the daylight savings time rules adjusted to reflect the new law taking effect in 2007? The rules in 1.33.1 are incorrect starting in 2007. thanks in advance michael __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 2

[ptr_container] ownership
by David Grüner 09 Dec '06

09 Dec '06

Hello, i need to give up the ownership of a ptr_container (vector here) over its pointers (without assignment to a new ptr_container). Two questions here: 1. Do i really have to write aPtrContainer.release().release()? It looks weird. For me, the function name release() of the container seems confusing since an unassigned aPtrContainer.release() just does the "opposite" of auto_ptr<T>::release(), namely deleting the pointers. ;o) The behavior is more like transfer_ownership() or something similar ($0.02). 2. Beside managing memory, i plan to use ptr_container primarily to propagate const. So, in some cases i need a ptr_container which just doesn't own any given memory. What's the best way to do that? Any comments or pointers to missed docs are welcome. David

2 1

Re: [Boost-users] boost::tokenizer and boost::tokenizer::iterator
by Benjamin Winkler 09 Dec '06

09 Dec '06

>I'm having a problem with tokenizer::iterator crashing the program when >I increment it. > > >tokens.assign(string("1234#4567#1##"), separator("#")); The tokenizer does not copy the sequence (the string) but only remembers its begin and end iterators. You are constructing a temporary string object which will be destroyed immediately at the end of the line, thus rendering the iterators invalid. Make sure that the string exists while you are using the tokenizer and its iterators on it, e.g. by making it a class member. Benjamin

1 0

Problem with lambda bind & overloaded functions
by Santosh Joseph 09 Dec '06

09 Dec '06

I have read threads about this and tried to follow the solutions but I still compile time errors: i'm using vs2003.net and i get this problem with g++3.4.2 using namespace boost::lambda; boost::lambda::placeholder1_type Arg; typedef pair<set<string>::iterator,bool>(set<string>::*INSERT)(const string&); std::vector<string> v; set<string> str_con; INSERT insert_func = &set<string>::insert; for_each( v.begin(), v.end(), bind( insert_func, str_con, Arg)); p:\Vendor\boost-1.33.1\boost\lambda\detail\actions.hpp(96): error C2665: 'boost::lambda::function_adaptor<boost::lambda::return_type_N<boost::lambda::function_action<3>,boost::lambda::lambda_functor_base<boost::lambda::action<3,boost::lambda::function_action<3> >,boost::tuples::tuple<const boost::lambda::detail::bind_traits<const INSERT>::type,const boost::lambda::detail::bind_traits<const set<string> >::type,const boost::lambda::detail::bind_traits<const boost::lambda::placeholder1_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type,boost::lambda::detail::bind_traits<boost::tuples::null_type>::type> >::sig<boost::tuples::tuple<const allocator<string>::value_type &> >::rets_t>::plain_Func>::apply': none of the 2 overloads can convert parameter 2 from type 'const rt1' ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

2 1

boost::tokenizer and boost::tokenizer::iterator
by Javier Estrada 08 Dec '06

08 Dec '06

I'm having a problem with tokenizer::iterator crashing the program when I increment it. I'm using it in a similar scope: In class constructor: { tokenizer tokens(...); tokenizer::iterator tokenIterator; } In a member function: { tokens.assign(string("1234#4567#1##"), separator("#")); tokenIterator = tokens.begin(); // use token iterator } In another member function: { cout << "token: " << *tokenIterator << endl; ++tokenIterator; ^^^ crash occurs in this instruction } I've verified: - Everything happens in the same thread (it's a multithreaded app) - tokenIterator != tokens.end(); Can anybody offer some insight? Regards, Javier

1 0

Signals, Format and Bind
by Memfis 08 Dec '06

08 Dec '06

I have two signals: signal<void(int)> source; signal<void(string)> target; What want is for every call: source(12345); to have a call: target("the value is 12345"); I am trying to achieve this with format and bind, but without much luck. What I came up with is the following: source.connect(bind(ref(target), bind(&format::str, bind(&format::operator%, format("the value is %1"), _1))); Unfortunately the code above treats the parameter to source as if it was empty, i.e. I get calls target("the value is "); What am I doing wrong and how can this be done correctly? -- Memfis

3 5

Boost 1.34 release
by Timm von Puttkamer 08 Dec '06

08 Dec '06

I would like to know when the next version of Boost will be released. Is there any time schedule? (At least I could not find anything on the website) Best regards Timm

1 0

need boost.subprocess
by Ilya Sokolov 08 Dec '06

08 Dec '06

Hi all! Are there any similar to python.subprocess in boost?

2 2

Re: [Boost-users] [tuple] [iterator] Multiple definition linker errors
by Mark Ruzon 08 Dec '06

08 Dec '06

>Mark Ruzon wrote: >> I have an executable with two .cpp files, each of which includes a >> header file that in turn includes boost headers. When I link them >> (gcc 3.2.3) >> I am greeted with multiple definition errors on somewhat random >> symbols like '(anonymous namespace)::_1' (through _9) from bind... > >This can be caused by precompiled headers not handling anonymous >namespaces >correctly. You can use the following patch: > >http://boost.cvs.sourceforge.net/boost/boost/boost/bind/placeholders.hpp?r1… > >to fix that. That worked great, Peter, thanks! Anyone know a similar trick for boost::detail::tuple_impl_specific::tuple_equal(boost::tuples::null_type, boost::tuples::null_type) on zip_iterator.hpp:320? This is my only other duplicate symbol. Mark

3 2

Re: [Boost-users] Auto-linking + serious Xplatform dev =>mutuallyexclusive?
by Sohail Somani 08 Dec '06

08 Dec '06

> [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Vladimir Prus > I suppose because on Linux, it's not typical to have several > variants. And > if you have just one variant, using > > -lwhatever Is it typical on Windows because of the way Microsoft set up the C++ runtime libraries? I can't recall ever seeing a single-threaded version of the g++ runtimes.

2 1

Re: [Boost-users] [BGL] read_graphviz
by Dmitry 08 Dec '06

08 Dec '06

Thank you very much to both. The life seems a little bit more sweeter now)

1 0

Re: [Boost-users] Auto-linking + serious Xplatform dev =>mutuallyexclusive?
by Sohail Somani 08 Dec '06

08 Dec '06

> [mailto:boost-users-bounces@lists.boost.org] On Behalf Of John Maddock > > It's only useful if your compiler supports it, period. > > However, those compilers that do support it typically have a > number of > different binary-incompatible build variants: figuring out > which lib variant > to link against without it in those cases is tricky, this is > a frequent > cause of problem reports caused by picking the wrong variant. I guess none of this is really news to me. Thanks for your reply.

1 0

Newbie to boost
by fungos 08 Dec '06

08 Dec '06

Hi, I need help to get started with boost, Im trying about some time to get a little project mine converted to use boost libs. My current project uses raw pointers and std::map. But I want to use shared_ptr and multi_index_container. I cant find help anywhere to my problems with this code, I hope I can get some here. Project Outline: - Strict OO design. - Filesystem emulation (a simple fake filesystem...) - I have one Filesytem interface with one implementation. - I have one Entry interface with three implementations. - Directory is a Entry. - File is a Entry. - Empty is a Entry. See the sample here: http://rafb.net/paste/results/hfHOx934.html , I have done this little sample to start understunding multi_index_container with classes using methods (mem_fun) as "selectors?". What is wrong with it? Please, I really need help to start with boost. But I doesnt want to let use best practices like good design and patterns. Thanks -- Animal Frontline Liberation

3 6

Auto-linking + serious Xplatform dev => mutually exclusive?
by Sohail Somani 08 Dec '06

08 Dec '06

Hi, For serious cross-platform development, have boost users found the auto-linking to be useful? I think it only works on Windows, so if you're doing any serious cross-platform development, the feature isn't really that useful. Am I missing something? Is this only useful on Windows? I'd love for it to work in general, but I think I have to replicate the logic in my build system which means I shouldn't really use it on Windows (since my logic will already be there). Of course, thats not to say that the feature isn't generally useful. If I was a Visual C++ developer only, I'd use it, no questions asked. Thanks, Sohail

3 2

boost::interprocess - allocating objects using node_allocator
by alexmark 08 Dec '06

08 Dec '06

Hi, I'd like to allocate objects directly in shared memory (not within STL-like container) in the way like that: struct X { int a; int b; }; typedef boost::interprocess::node_allocator<X, A_POOL_SIZE, managed_shared_memory::segment_manager > AllocType; int main (...) { // create shared memory managed_shared_memory sharm (create_only, "a_name", SOME_SIZE); // create allocator object AllocType alloc_object (sharm.get_segment_manager()); // allocate memory for object instance taking it from the allocator's pool X* p = detail::get_pointer(alloc_object.allocate(1)); // invoke the constructor alloc_object.construct(p, X()); // obtain the handle sharm.get_handle_from_address(p); // now the handle is passed to another application to let it access the shared object // .... } Is it correct to do so? Will it work? TYA

2 1

Multiple definition linker errors
by Mark Ruzon 07 Dec '06

07 Dec '06

I have an executable with two .cpp files, each of which includes a header file that in turn includes boost headers. When I link them (gcc 3.2.3) I am greeted with multiple definition errors on somewhat random symbols like '(anonymous namespace)::_1' (through _9) from bind or equality operators in zip_iterator etc. All my header files are properly wrapped with #ifndefs, I have only one global variable that does not involve boost, and everything links fine for executables with only one .cpp file that includes this same header. I assume the fix lies with the gcc command line, either a missing flag or the order of arguments. Does anyone know how to fix this or what information would be useful in tracking this down? I'm out of ideas. Sincerely, Mark Ruzon

2 1

Re: [Boost-users] [BGL] read_graphviz
by Dmitry 07 Dec '06

07 Dec '06

I don't see the way you read graph property?? "Jeffrey Holle" <jeffreyholle(a)bellsouth.net> wrote in message news:<el7jds$o4$1(a)sea.gmane.org>... So true. Thought brevity was more important. Jens Muller wrote: > Jeffrey Holle schrieb: >> Since no one replied to this message, I'll answer it myself. Yes, to >> both questions! >> This is the code that demonstrates: > > There are obviously includes missing in there. ----------

3 2

[regex VC 7.1] multiple symbols in link
by Marko Bozikovic 07 Dec '06

07 Dec '06

Hi all, I'm looking into upgrading from Boost 1.32 to 1.33.1, but I have a problem with regex lib. When linking my application with the regex lib (I use automatic linking), I get multiple symbols linker error: libboost_regex-vc71-mt-gd-1_33_1.lib(usinstances.obj) : error LNK2005: "public: static unsigned int __cdecl std::char_traits<unsigned short>::length(unsigned short const *)" (?length@?$char_traits@G@std@@SAIPBG@Z) already defined in msvcprtd.lib(MSVCP71D.dll) libboost_regex-vc71-mt-gd-1_33_1.lib(usinstances.obj) : error LNK2005: "public: static bool __cdecl std::char_traits<unsigned short>::eq(unsigned short const &,unsigned short const &)" (?eq@?$char_traits@G@std@@SA_NABG0@Z) already defined in msvcprtd.lib(MSVCP71D.dll) With boost 1.32, everything works fine. In both 1.32 and 1.33.1, regex was built with bjam using VC 2003 compiler, default configuration (just ran bjam.exe "-sTOOLS=vc-7_1") STL library is the standard one, shipped with VS 2003. Any ideas? Thank you, -- Marko ICQ: 5990814 I'm not under the alkafluence of inkahol that some thinkle peep I am. It's just the drunker I sit here the longer I get.

2 1

Boost.Python windows auto-linking
by David Abrahams 07 Dec '06

07 Dec '06

Boost's CVS HEAD now contains support for the auto-linking feature provided by most windows compilers (e.g. Visual C++). I'm very eager to hear whether it works for people, especially since we're getting very close to releasing Boost 1.34 Thanks, -- Dave Abrahams Boost Consulting www.boost-consulting.com

1 0

bjam doesn't like filenames with '++' in them???
by Paul Kerchen 07 Dec '06

07 Dec '06

I'm using bjam for a project that needs to copy some third party dlls (ImageMagick) to an installation directory. One of those dlls has '++' in the filename, which, under Windows at least, is legal, but bjam can't seem to deal with it. This line: InstallInto $(BIN_INSTALL_DIR) : CORE_RL_Magick++_.dll ; results in this output: Install AppInstall\CORE_RL_Magick++_.dll The system cannot find the file specified. copy ImageMagick\CORE_RL_Magick++_.dll AppInstall\CORE_RL_Magick++_.dll ...failed Install AppInstall\CORE_RL_Magick++_.dll... ...failed updating 1 target... I've verified that the source file does exist and all of the other DLLs in the same directory are copied without a problem. I've tried several things to work around the problem: escaping the '+' characters, putting the whole filename in quotes, definining variables that I can use to construct a filename with. However, it seems to come down to InstallInto not being able to deal with '++' in filenames. And since this is a third party DLL, I can't simply rename the file. I've searched the Boost newsgroups and FAQs for an answer, but no one seems to have encountered this problem, so that leads me to believe that I'm doing something wrong. Has anyone had this problem before? Does anyone have any ideas on how to fix it? Any help with this would be greatly appreciated! Paul Kerchen

2 2

Math library on the BOOST licence.
by Seweryn Habdank-Wojewódzki 07 Dec '06

07 Dec '06

Hi Are there any similar to GSL library which is designed using templates and on the BOOST license. I would like to join to such a project - I have developed Runge-Kutta generic solver maybe it can be useful, at present I preparing other designs, but maybe somewhere other is done. Regards. -- |\/\/| Seweryn Habdank-Wojewódzki `\/\/'

2 4

boost::cref behavior
by roblin＠jlab.org 07 Dec '06

07 Dec '06

Hi all, I was trying to understand the use of boost::ref and boost::cref and wrote this little snippet: #include <iostream> #include <boost/ref.hpp> #include <string> using namespace std; int main(int argc,char** argv) { boost::reference_wrapper<string const> ref=boost::cref(string("yoh")); const string das("bla"); const string & dasref=das; cout<<"das ref is "<<dasref<<endl; cout<<"ref is "<<ref.get()<<endl; return 0; } I am confused by what I get when I run this: das ref is bla ref is bla I was expecting: das ref is bla ref is yoh What am I doing wrong there ??

2 1

Spamming through boost???
by Ovanes Markarian 07 Dec '06

07 Dec '06

Hi all! I give to all my opponents personalized email addresses, e.g. for boot my email is om_boost(a)domain_name.ext. Today I got my first spamming email to this address. Now I really think, there must be email servers inbetween, which simply log the sender and put them to the spamming lists. I don't see any other possiblity of how this can happen. I don't beleave that boost sells email addresses to spammers, after I know you guys. Would be nice to know what others think. I know that this email is off-topic, but can be useful anyway. With Kind Regards, Ovanes

3 2

[Boost.Iostreams] does filtering_stream have a binary mode?
by Maurice van Swaaij 07 Dec '06

07 Dec '06

Hi, Is there a way to set a filtering_stream to binary mode so I can use its operator>> and operator<< to read and write binary data? Thanks. -- Maurice van Swaaij - Blue Sky Studios NY

2 1

[BGL] read_graphviz
by Jeffrey Holle 06 Dec '06

06 Dec '06

I using the boost CVS version, obtained last week Tuesday. I'm interested in using read_graphviz in a fairly complex manner. I want to read multiple properties from a *.dot file, including: - 1 graph property - 2 vertex properties - 2 edge properties Can this be done with the read_graphviz? Moreover, I would like these properties to end up outside of the graph. If this is possible, how is it done?

2 3

[date_time] how to parse "1/13/2006"
by Markus Werle 06 Dec '06

06 Dec '06

Hi! I have to read in a file format where the date is given as month/day/year, but without leading '0' for day and month. >From http://boost.org/doc/html/date_time/date_time_io.html I could not guess how to declare the facet format string (No leading zeroes required) Could you please complete the following code? std::istringstream iss; local_time_input_facet* input_facet = new local_time_input_facet(); ss.imbue(locale(ss.getloc(), input_facet)); input_facet->format("???????"); regards, Markus

1 1

using source code files from boost-sandbox
by Luca Cappa 06 Dec '06

06 Dec '06

Hello, I would like to use the circularbuffer class provided by boost-sandbox, and I would be glad if someone could tell me which are the rules to include the boost-sandbox files (either the .cpp and the .h) in my building process, i.e. should I copy only the relevant files into the actual boost installation directory, or anything else? Luca

1 0

Re: [Boost-users] Binding to an overloaded member function
by Javier Estrada 05 Dec '06

05 Dec '06

Excellent. Thx. I knew it was somewhere in the documentation (Boost 1.33) but I could not get my hands around it ;-) -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Sebastian Redl Sent: Tuesday, December 05, 2006 1:20 PM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Binding to an overloaded member function Javier Estrada wrote: > How can I use boost bind to bind to an overloaded member function? > You have to cast explicitly to select the correct overload. > But when I added: > > void C::dial(char key, MyConstants c); void C::dial(string toDial); > void C::dial(string, MyConstants c); > > References to my working code break: > > for_each(begin, end, boost::bind(&C::dial, this, _1 )); > for_each(begin, end, boost::bind( static_cast<void (C::*)(char)>(&C::dial), this, _1)); Sebastian Redl _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

Binding to an overloaded member function
by Javier Estrada 05 Dec '06

05 Dec '06

How can I use boost bind to bind to an overloaded member function? I had working code for a member function: void C::dial(char key); But when I added: void C::dial(char key, MyConstants c); void C::dial(string toDial); void C::dial(string, MyConstants c); References to my working code break: for_each(begin, end, boost::bind(&C::dial, this, _1 )); Regards, Javier

2 1

Boost-users Digest, Vol 1107, Issue 3 (fwd)
by Mark Ruzon 05 Dec '06

05 Dec '06

>On Dec 4, 2006, at 6:28 PM, Mark Ruzon wrote: > >> The documentation for tuple says there are no portability problems >> with >> gcc 2.95. I am using 3.2.3 and am getting a syntax error on the >> line that >> does the work of this function object: >> >> #include <boost/tuple/tuple.hpp> >> >> struct get_0_func { >> template <typename II> // II models Input Iterator on tuple >> void operator()(II it) { it->get<0>(); } > >try : > > void operator()(II it) { it->template get<0>(); } > >Not tested, but this is likely the reason. The compiler does not how >yo parse < between get and 0, if not told >that get is a template. > > Jaakko J?rvi Thank you, Jaakko, this works fine. How do I get the documentation for tuple amended to include this form? Mark

1 0

[iostreams] How to strip off checksum at end of file in multichar input filter?
by eg 05 Dec '06

05 Dec '06

Hi all, I've written a multichar_dual_use_filter for to handle a custom file encryption. Ideally, the encrypted file format consists of 3 distinct types of data: 1) initialization vector IV (only used by encry/decryp) 2) Main file data 3) Checksum (only used to validate file decryp is correct) The output filter can easily write the above 3 items. I have the input filter reading in and handling items 1 and 2 above... But I need some advice on how to strip off the checksum (or any fixed size chunk of data) at the end of the file to validate but not pass this into the output buffer. I have thought of passing in the input file size to the filters constructor and keep track of how many bytes are being processed... then handling the final bytes separately. Does this sound like a feasible approach? It makes using the filter harder as it imposes unusual constraints on calling it. Any other ideas? My current filters Read function is of the following form: template<typename Source> std::streamsize read(Source& src, char* s, std::streamsize n) { if (!bInit) { // Handle reading IV bytes and initialization } // Read in buffer to "s" and then decrypt it count = boost::iostreams::read(src, s, n); if (count <= 0) return -1; //EOF or Error // Decrypt s in place DecryptBuf(s, count); return count; }

1 0

[regex] - merging scanners
by Frank Bergemann 05 Dec '06

05 Dec '06

Hi, is there a way to define a set (list) of regular expressions and merge them into *one* "scanner". So that i can pass some input string to this merged scanner ( with a _single_ function call ) and as a result get back which of the regular expression(s) match for that string (two modes:(1) 1st hit & (2) all hits)? Currently i use a loop to check for each expression individually/sequentially. \Frank

1 0

Another sub_range problem
by abir basak 05 Dec '06

05 Dec '06

Hi, I am facing some more range problem from boost::range; According to the CVS RC_1_34 works. Revision 1.21 version specified by Thorsten Ottosen , most of the classes work. What is my interest is to have a get_range function inside some of the container class like std::vector and boost::circular_buffer (from boost CVS) Here I have test class to show the problem I am facing. First, when I inherit from a vector , and specify a range function just like iterator functions ( begin() and end() ) template<typename T> class range_vector : public std::vector<T>{ public: typedef boost::sub_range<range_vector> range; typedef boost::sub_range<const range_vector> const_range; range get_range(size_type from,size_type to){ std::cout<<"range"<<std::endl; return range(begin()+from,begin()+to); } const_range get_range(size_type from,size_type to)const{ std::cout<<"const range"<<std::endl; return const_range(begin()+from,begin()+to); } }; int main(){ range_vector<int> v; int x[] = {0,1,2,3,4,5,6,7,8}; v.insert(v.end(),x,x+9); std::copy(v.begin(),v.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; range_vector<int>::range r = v.get_range(1,5); r[0] = 100; copy(r.begin(),r.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; range_vector<int>::const_range cr = v.get_range(1,5); copy(cr.begin(),cr.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; const range_vector<int>& cv = v; range_vector<int>::const_range cr1 = cv.get_range(1,5); copy(cr.begin(),cr.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; } This code compiles perfectly and gives result as expected. It also works with deque . However when I compile the same code for boost::circular_buffer,i.e template<typename T> class range_vector : public boost::circular_buffer<T>{ public: typedef boost::sub_range<range_vector> range; typedef boost::sub_range<const range_vector> const_range; range get_range(size_type from,size_type to){ std::cout<<"range"<<std::endl; return range(begin()+from,begin()+to); } const_range get_range(size_type from,size_type to)const{ std::cout<<"const range"<<std::endl; return const_range(begin()+from,begin()+to); } }; It has a same interface that of a vector, it fails to compile when I call the non-const get_range function. i.e //Initialized the circular_buffer with a few elements. range_vector<int> v; int x[] = {0,1,2,3,4,5,6,7,8}; v.insert(v.end(),x,x+9); std::copy(v.begin(),v.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; ///got a const reference and invoking const version of get_range. works fine. const range_vector<int>& cv = v; range_vector<int>::const_range cr1 = cv.get_range(1,5); copy(cr1.begin(),cr1.end(),std::ostream_iterator<int>(std::cout," ")); std::cout<<std::endl; ///now trying to get a non-const version range_vector<int>::range r = v.get_range(1,5); Here it gives error in Visual Studio e:\boost_1_33_1\boost\range\iterator_range.hpp(60): error C2440: 'type cast' : cannot convert from 'boost::range_iterator<C>::type' to 'boost::range_iterator<C>::type' with [ C=const boost::sub_range<range_vector<int>> ] and [ C=range_vector<int> ] ///now trying to get a non-version and assign it to const version range_vector<int>::const_range r = v.get_range(1,5); It gives the same error. Here C (container) is passed as correct one. begin() and end() are too executes correct version (non-const one). Not getting why it won't work. Any kind of help is highly appreciated. Thanks to boost community in advance. abir basak -- Abir Basak, Member IEEE Software Engineer, Read Ink Technologies B. Tech, IIT Kharagpur email: abir(a)abirbasak.com homepage: www.abirbasak.com

3 4

[tuple] get not compiling with gcc 3.2.3
by Mark Ruzon 05 Dec '06

05 Dec '06

The documentation for tuple says there are no portability problems with gcc 2.95. I am using 3.2.3 and am getting a syntax error on the line that does the work of this function object: #include <boost/tuple/tuple.hpp> struct get_0_func { template <typename II> // II models Input Iterator on tuple void operator()(II it) { it->get<0>(); } }; The boost::get<0>(*it) form works fine. Does anyone know why a newer version of gcc would not compile this? Mark

2 1

Why BOOST_CURRENT_FUNCTION not in docs?
by Pavel Syomin 04 Dec '06

04 Dec '06

Hi all! Why BOOST_CURRENT_FUNCTION not in docs? Is it a part of API or not?

3 6

functional/hash: hash_value with context
by Boris Mansencal 04 Dec '06

04 Dec '06

Hi, A newbie question regarding functional/hash. I would like to use TR1 unordered_map for a 2D point custom data type. For this, I use hash_value from Boost. But I would like to give/pass a context to this function : here, width & height of domain in which are my 2D points. Currently, I declare theses width and height static. My code (in my cpp file) looks like : static int P_width; static int P_height; std::size_t hash_value(point const& p) { return ((size_t)p.x()<<1)*P_width + (size_t)(p.y()<<1); } But this solution is not satisfactory as I can not have several instances of my class including the unordered_map, i.e., several domains of different width & height. Is there a straightforward solution that I have overlooked ? Regards, Boris.

3 6

[License] OSI certification progress
by Ole Christian Weidner 04 Dec '06

04 Dec '06

Hi, I just read through the Boost License page at http:// www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl? Boost_License. It states that an OSI license approval process has been initiated in January 2005 - now I'm curious about the status quo, since I can't find any further information on this effort... It would be much easier to release software under the Boost License if one could directly refer to OSI compliance since they seem to be well known and have a broad acceptance. Does anyone have any information on this topic? Thanks a lot! Cheers, Ole

1 0

Re: [Boost-users] Boost-users Digest, Vol 1106, Issue 1
by craigp 04 Dec '06

04 Dec '06

> Message: 7 > Date: 02 Dec 2006 17:51:00 -0600 > From: Aleksey Gurtovoy <agurtovoy(a)meta-comm.com> > Subject: Re: [Boost-users] odd msvc 7.1 compilation errors with mpl > To: boost-users(a)lists.boost.org > Cc: dave(a)boost-consulting.com, craigp <craigp98072(a)yahoo.com> > Message-ID: <m2y7pp3li3.fsf(a)meta-comm.com> > Content-Type: text/plain; charset=us-ascii > > craigp writes: > > For some reason, I can't get the IDE to compile BOOST_MPL_ASSERT. I tried <snip> > This is a known issue: the errors you see are specific to MSVC "Edit & > Continue" mode, in which the __LINE__ macro (used internally in MPL > asserts) stops being a preprosessing-time constant and becomes a > reference to a global run-time variable > (http://support.microsoft.com/kb/199057) > > To alleviate this problem, Microsoft has introduced a proprietary > predefined macro __COUNTER__ > (http://msdn2.microsoft.com/en-us/library/b0084kay.aspx) intended to > serve as a replacement for __LINE__ in situations where the latter is > used to generate a unique identifier rather than to actually refer to > a line number (which is the case with MPL asserts). Unfortunately, > because __COUNTER__ is incremented every time it's used, fixing this > is a bit more complicated then just a search & replace for __LINE__, > and therefore it has to wait for 1.35 (or you can grab the latest > version of "boost/mpl/assert.hpp" from Boost CVS's HEAD trunk). > > HTH, > -- > Aleksey Gurtovoy > MetaCommunications Engineering thanks Aleksey!! works fine now. cheers, --craig ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

2 1

Re: [Boost-users] circular_buffer doc
by Pfligersdorffer, Christian 04 Dec '06

04 Dec '06

http://boost.cvs.sourceforge.net/*checkout*/boost-sandbox/boost-sandbox/lib… Under the above link you should find the documentation. Greetings, -- Christian Pfligersdorffer Dipl.-Ing. Bakk.rer.nat. Software Engineering EOS GmbH - Electro Optical Systems Robert-Stirling-Ring 1 D-82152 Krailling / Munich Phone +49 89 893 36-229 Fax +49 89 893 36-287 christian.pfligersdorffer(a)eos.info http://www.eos.info

1 0

odd msvc 7.1 compilation errors with mpl
by craigp 02 Dec '06

02 Dec '06

For some reason, I can't get the IDE to compile BOOST_MPL_ASSERT. I tried this: #include <boost/type_traits/is_same.hpp> #include <boost/mpl/assert.hpp> int main() { BOOST_MPL_ASSERT(( boost::is_same< char, char > )); } and it gives the following errors: d:\dev\boost.cvs\boost\libs\fusion\example\extension\test_example.cpp(31) : error C2143: syntax error : missing ';' before '=' d:\dev\boost.cvs\boost\libs\fusion\example\extension\test_example.cpp(31) : error C2143: syntax error : missing ';' before '=' what's even stranger is that i copied/pasted the pre-processed file, and that compiled just fine: static const std::size_t mpl_assertion_in_line_26 = sizeof( boost::mpl::assertion_failed<false>( boost::mpl::assert_arg( (void (*) ( boost::is_same< char, char > ))0, 1 ) ) ); anyone else seen something like this before?? all the mpl regression tests compile fine via bjam (using msvc 7.1), so it appears ide-specific, somehow... thanks, --craig ____________________________________________________________________________________ Do you Yahoo!? Everyone is raving about the all-new Yahoo! Mail beta. http://new.mail.yahoo.com

3 2

Can't build Boost on Windows
by Ronald Levine 02 Dec '06

02 Dec '06

This is my second attempt to post this problem without having seen it on the list. and I'm really desperate for some help on this. Having moved to my new Compaq Presario V6000 ( AMD Turion 64 X2), with Windows XP Media Center Edition, as my ONLY Windows machine, I find that I can't build Boost on it, so all of my Windows C++ development is stalled. Following the simple instructions, produces the following output, and abortion. HELP! C:\kits\boost_1_33_1>bjam "-sTOOLS=vc-8_0" install ****************************************************** Building Boost.Iostreams with bzip2 support disabled. To enable bzip2, consult the Boost.Iostreams documentation ****************************************************** ****************************************************** Building Boost.Iostreams with zlib and gzip support disabled. To enable zlib and gzip, consult the Boost.Iostreams documentation ****************************************************** Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). ...patience... ...patience... ...found 13371 targets... ...updating 4941 targets... MkDir1 C:\Boost spawn: Invalid argument

3 3

Boost Lamdba: Binding to overloaded operator and templated functions???
by Haroon Khan 02 Dec '06

02 Dec '06

Is it possible to bind to the following functions using boost.lamdba and if so how? class some_class { public: some_class operator % (int x) { return this } template<typename T> void foo(T x){ } }; I use the examples given in online-documentation but it gives me a lot of errors. I'm using VC8 on windows XP. Thanks, Haroon

2 4

newbie: shared_ptr member function customer deleter???
by Travis Parks 02 Dec '06

02 Dec '06

Hello: I just started using some boost today. I am trying to use a non-static member function as the deleter in a shared_ptr. I was looking at the source code and I don't think it is supported. This is what I had: shared_ptr<Environment> environment(Environment::createEnvironment(), Environment::terminateEnvironment); string userName = "monkey_bottoms"; string password = "blahblahblah" shared_ptr<Connection> connection( environment->createConnection( userName, password), &Environment::terminateConnection); This didn't work at all! When I looked at the source I saw why. Note that Environment::terminateConnection is a static method. However, this source works fine for the static Environment::createEnvironment method. I am working in Visual Studio 8 with the OCCI. For the time being, I have made a temporary solution. I am hoping someone can help me find a less customized solution. Plus, I am thinking having a pointer to an object that I am trying to manage kind of defeats the point. Here you go: #include "occi.h" #include <boost/shared_ptr.hpp> #include <string> #include <functional> // retarded temporary solution template <class TCLASS, class TRETURN, class TARG> class call_redirect_t : public std::unary_function<TARG, TRETURN> { TCLASS * instance; // barf I have a pointer to a managed resource! TRETURN (TCLASS::*func) (TARG); public: call_redirect_t(TCLASS * instance, TRETURN (TCLASS::*func) (TARG)) : instance(instance), func(func) { } TRETURN operator()(TARG arg) { return (instance->*func)(arg); } }; // convenience function template <class TCLASS, class TRETURN, class TARG> call_redirect_t<TCLASS, TRETURN, TARG> call_redirect(TCLASS * instance, TRETURN (TCLASS::*func)(TARG)) { return call_redirect_t<TCLASS, TRETURN, TARG>(instance, func); } int main(int argc, char * argv[]) { // lazy using namespace std; using namespace oracle::occi; using namespace boost; shared_ptr<Environment> environment( Environment::createEnvironment(), // gets Environment * Environment::terminateEnvironment); // static method string userName = "flipflop"; string password = "memoryrizor"; shared_ptr<Connection> connection( environment->createConnection(userName, password), // gets Connection * call_redirect( environment.get(), // pull the pointer out; instant poo &Environment::terminateConnection) // non-static method ); } Well, now that I think about it; since the temporary call_redirect will only exist in the scope of the shared_ptr (who handles my resource) it doesn't matter that I let it go, and parameter evaluation order is not an issue. Take a look at the operator() in call_redirect_t. I had to put parenthesis around: (instance->*func) I am think I have forgotten the syntax. Hmph. Weep, I am not perfect. Thanks for any input, folks. Travis Parks

2 2

Boost Graph - Sub Graph Bundled Properties Error
by John Paulus 02 Dec '06

02 Dec '06

Could someone explain with the code below generates an error? The code worked fine with a standard boost graph (ie not a subgraph<>)...... I can create the sub graph, but cannot access the bundled properties data: typedef adjacency_list<listS, listS, undirectedS, VertexData, property<edge_index_t, int, EdgeData> > GraphType; typedef subgraph<GraphType> SubGraphType; SubGraphType SubGraph; SubGraphType::vertex_descriptor V = add_vertex(SubGraph); SubGraph[V].ID=1; ////ERROR!!!!!! Sincerely, John

3 3

Building boost 1.33.1 64-bit using VC-8
by Jason Nye 02 Dec '06

02 Dec '06

Hello all, First of all, I want to say boost has been totally awesome for the programming team in which I work. Kudos to everyone contributing! For skimmers, #3 is the one that has a question related to it. I'm trying to build boost 1.33.1 64-bit using VC-8 (team suite) and I've run into some (minor) issues. They are: 1. The vc-8_0-x86_amd64-tools.jam does not refer to the correct vcvars batch file. It specifies $(VC_ROOT)\bin\x86_amd64\VCVARSamd64.bat but there isn't any such file. The proper file is $(VC_ROOT)\bin\x86_amd64\vcvarsx86_amd64.bat 2. The thread library doesn't link because of the ice_wrapper used in once.cpp (in compare_exchange line 119). Replacing the compare_exchange with an actual call to InterlockedCompareExchange did the trick. I believe this is because it is an intrinsic function. 3. This one is the most confusing to me: the abi prefix headers perform #pragma pack(push, 8). No warnings pop up in 32-bit mode because 8 bytes is the default alignment. However, in 64-bit mode, the default alignment is 16 bytes. Why would boost ever want to enforce a specific packing? Is it safe for me to simply comment these out? I *really* don't like the idea of enforcing an 8 byte alignment when the default is 16 bytes. Any comments here? Thanks! Jason.

3 3

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 02 Dec '06

02 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Sebastian Redl > Sent: Friday, December 01, 2006 1:23 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > The use case here is this: > > NodePtr base = make_tree(); > NodePtr child = base.get("path/to/child"); > base.reset(); > Now, unless the parent knows that it owns the child and notifies it, the > parent pointer of child is dangling. A weak pointer can detect this > situation. [Nat] Okay. > The reason I see for there being no weak_intrusive_ptr is that it is > quite simply impossible to implement while staying intrusive. Let's > remember what a weak pointer does: after the death of the object it > points to, it knows that it is now invalid. There being no notification > of pointers when other pointers get destructed, there's only one way to > do this: to keep the refcount for the object alive, so that the weak > pointer can check it. > But if the refcount is part of the object, as is the case for > intrusive_ptr, it's impossible to keep the refcount alive while > destroying the object. > > In other words, the appeal of intrusive_ptr is that it doesn't require > any resources outside the object it points to. However, this restriction > makes it impossible to implement an efficient weak pointer. [Nat] Not to be argumentative, but... It seems possible to implement weak_ptr notification by building a list of weak_ptr instances referencing a given object. If we don't want that list to consume additional heap memory, the list could itself be intrusive in the weak_ptr objects. If we want it to be efficient, we build a doubly-linked list. (This may call for a policy-based implementation so the consumer can decide which overhead is least noxious.)

6 5

Boost::wave, sample driver and how to use it
by mosfet 02 Dec '06

02 Dec '06

Hi, I don't undertand how wave works. I am trying to use it on a simple C header file. I have included System and user include paths and some defines but when I start wave sample driver wave.exe, nothing happens. What I want is to convert all the defines into native types, for instance. By the way the config file doesnt' seem to like windows path while it works on the command line. foo.h //-----// VOID foo(BYTE bTest, CHAR szTest); ULONG foo(LPBYTE bTest, CHAR szTest); I would like to get void foo(unsigned char bTest, char szTest); unsigned long foo(BYTE * bTest, char szTest); but maybe I haven't understood what a preprocessor was. If someone could help, it's really urgent.

3 4

interprocess::list and destructors of stored objects
by alexmark 02 Dec '06

02 Dec '06

Hi to all! For the container type boost::interprocess::list<T, A> I need the destructor of stored object T be called when erase() is called for a corresponding list item. By default the T destructor is not called, memory only is deallocated. What is the correct way to force boost::interprocess::list to invoke destructors for contained objects while list item or the container itself is deleted? TYA.

3 2

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

________________________________________ From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Gottlob Frege Sent: Friday, December 01, 2006 3:25 PM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > On 12/1/06, Nat Goodspeed <ngoodspeed at solidworks dot com> wrote: > > It seems possible to implement weak_ptr notification by building a list > > of weak_ptr instances referencing a given object. If we don't want that > > list to consume additional heap memory, the list could itself be > > intrusive in the weak_ptr objects. If we want it to be efficient, we > > build a doubly-linked list. (This may call for a policy-based > > implementation so the consumer can decide which overhead is least > > noxious.) > > Interestingly, since you define the intrusive_ptr_add_ref and > release, you can probably implement the list management without > changing intrusive_ptr. [Nat] That's true, you could. But that would impose the requirement on everyone who wanted to use intrusive_ptr with this new variety of weak_ptr. Perhaps intrusive_ptr could do something implicit before calling intrusive_ptr_release. Maybe that could be implemented as an intrusive_ptr policy. > Would you prefer to have the list of weak_ptrs kept on the > intrusive_ptr (on the actual 'intruded' object, actually), or > have a global map of intrusive_ptr object to weak_ptr list? [Nat] I was thinking of storing the head pointer in the object itself. But it might be more consistent with the present intrusive_ptr implementation to call an intrusive_ptr_get_first_weak_ptr function, requiring the consumer to decide where to put it.

1 0

Re: [Boost-users] Boost::wave, sample driver and how to use it
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Sebastian Redl > Sent: Friday, December 01, 2006 3:37 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Boost::wave, sample driver and how to use it > > mosfet wrote: > > VOID foo(BYTE bTest, CHAR szTest); > > ULONG foo(LPBYTE bTest, CHAR szTest); > > > If this is Windows, those are typedefs, not macros. The preprocessor > does not affect them. [Nat] <off-topic> Might gccxml be a better answer for the OP? Caveat: AFAIK, gccxml doesn't deal well with MFC headers and doesn't yet have its own equivalents. </off-topic>

1 0

Re: [Boost-users] Boost::wave, sample driver and how to use it
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of mosfet > Sent: Thursday, November 30, 2006 3:07 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] Boost::wave, sample driver and how to use it > > foo.h > //-----// > > VOID foo(BYTE bTest, CHAR szTest); > ULONG foo(LPBYTE bTest, CHAR szTest); > > I would like to get > > void foo(unsigned char bTest, char szTest); > unsigned long foo(BYTE * bTest, char szTest); > > but maybe I haven't understood what a preprocessor was. > If someone could help, it's really urgent. [Nat] You mentioned Windows. Does that mean you're using Microsoft C++? Wouldn't it work to use something like cl /E ... or cl /P ... ?

1 0

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Ovanes Markarian > Sent: Friday, December 01, 2006 2:23 PM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Just a small question: > > Isn't it so, that linked lists (double or single), can be spreaded in the > memory, so that cache > misses and page faults are subject to happen? I think deque is a better > choice (which can be > better optimized in terms of locality)... [Nat] My suggestion was based on the (implied) intrusive_ptr constraint of no additional heap resources. But if we were to use a policy-based approach, a deque implementation could be one of the policies.

1 0

Problem building Boost on Windows
by Ronald Levine 01 Dec '06

01 Dec '06

C:\kits\boost_1_33_1>bjam "-sTOOLS=vc-8_0" install ****************************************************** Building Boost.Iostreams with bzip2 support disabled. To enable bzip2, consult the Boost.Iostreams documentation ****************************************************** ****************************************************** Building Boost.Iostreams with zlib and gzip support disabled. To enable zlib and gzip, consult the Boost.Iostreams documentation ****************************************************** Building Boost.Regex with the optional Unicode/ICU support disabled. Please refer to the Boost.Regex documentation for more information (and if you don't know what ICU is then you probably don't need it). ...patience... ...patience... ...found 13371 targets... ...updating 4941 targets... MkDir1 C:\Boost spawn: Invalid argument

1 0

Re: [Boost-users] boost::weak_ptr and boost::intrusive_ptr
by Nat Goodspeed 01 Dec '06

01 Dec '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org [mailto:boost-users- > bounces(a)lists.boost.org] On Behalf Of Steven T. Hatton > Sent: Friday, December 01, 2006 7:03 AM > To: Boost-users(a)lists.boost.org > Subject: [Boost-users] boost::weak_ptr and boost::intrusive_ptr > > Intrusive pointers are, IMO, the correct mechanism to hold a > tree together. They are efficient, effective, and easy to > understand. They work better than shared pointers for > recursive data structures. Unfortunately the lack of a > corresponding weak_ptr to go with boost::intrusive_ptr means I > either use a raw pointer to the parent, devise a strategy for > decrementing the pointers myself, or create my own weak > pointer. [Nat] I've thought about this problem to some extent. This is the way I see it so far. The one big advantage of a weak_ptr over a raw pointer is that the weak_ptr can tell you when its referent becomes invalid. If your data structure requires storing a pointer in a place which itself becomes invalid as soon as the referent becomes invalid (e.g. recursive deletion of a subtree), then I'm not convinced that weak_ptr is significantly better than a raw pointer. (Even "snipping loose" a subtree, but preserving it, doesn't seem to argue for weak_ptr as a parent pointer. In such a case, you'd want to clear the parent pointer anyway, wouldn't you?) Does this rationale make sense, or am I overlooking something? I'm not rejecting the idea of a weak_ptr that works with intrusive_ptr. I'm just saying that for certain use cases, raw pointers seem to address the requirements equally well -- with less runtime overhead. This case looks like one of them.

2 1