- Boost-users - lists.boost.org

[boost::shared_ptr] Memory is not freed
by Panagiotis Foteinos 28 Mar '11

28 Mar '11

Hello users. In the past, I wrote a program using the C++ built-in pointers. I made sure that there was not any memory leak: indeed, valgrind reported no memory leaks. I decided to switch to boost::shared_ptr, and replaced all the C++ built-in pointers with boost:shared_ptr ones. Interestingly, valgrind now reports a huge memory leak. I thought that by just replacing the built-in pointers with the boost::shared_ptr pointers (forgetting about explicit deletes of memory) would be enough. Note that I use boost::make_shared<> for memory allocations. I wish I could include the code in this email, but it is very long and complicated. Therefore, I am kindly asking you to guess what might have been wrong with the boost:shared_ptr. Could you think of any scenarios that would cause the observed bahavior? Any ideas would be welcome. OS: Ubuntu 64 Compiler: gcc version 4.4.3 Best Regards, Panagiotis Foteinos

3 3

[serialization] empty xml output
by Olivier Tournaire 27 Mar '11

27 Mar '11

Hi all, I am trying to introduce XML serialization in my project, but even with the simplest test, I cannot get something to work. Here is what I did: #define PARAMETERS_HPP #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/xml_oarchive.hpp> #include <iostream> class rss_feeds_to_parse { public: bool all; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { ar & BOOST_SERIALIZATION_NVP(all); } }; class parameters { public: rss_feeds_to_parse feeds; template<class Archive> void serialize(Archive & ar, const unsigned int file_version) { ar & BOOST_SERIALIZATION_NVP(feeds); } }; #endif // PARAMETERS_HPP Later, in my main, I have: parameters p; p.feeds.all = true; std::ofstream ofs("test_parameters.xml"); assert(ofs.good()); boost::archive::xml_oarchive oa(ofs); oa << BOOST_SERIALIZATION_NVP(p); It compiles, links and executes well, the file test_parameters.xml is created ... but empty ! What am I doing wrong? Hope you could help. Best regards, Olivier

2 4

Re: [Boost-users] Boost-users Digest, Vol 2674, Issue 3
by Jim Hodapp 27 Mar '11

27 Mar '11

Ian, That's excellent, thanks! Is g++ the compiler with the most of the c++0x specification implemented and the best to experiment with the new features? Thanks, Jim On Mar 27, 2011, at 12:00 PM, boost-users-request(a)lists.boost.org wrote: > > > When replying, please edit your Subject line so it is more specific > than "Re: Contents of Boost-users digest..." > Today's Topics: > > 1. C++0x resources (Jim Hodapp) > 2. Re: C++0x resources (Ian Bruntlett) > > From: Jim Hodapp <james.hodapp(a)gmail.com> > Date: March 27, 2011 10:53:09 AM EDT > To: boost-users(a)lists.boost.org > Subject: [Boost-users] C++0x resources > Reply-To: boost-users(a)lists.boost.org > > > Hi all, > > I just saw in another post that C++0x FDIS has been voted complete and ready to ship. Is there a good website (or book) that I can start reading to get familiar with the new features? I've seen a few resources here and there on one or two topics, such as "auto", but no good place to read and see example for every new feature. Is the GNU g++ compiler the best implementation to experiment with these new language features? > > -Jim > > > > > > > From: Ian Bruntlett <ian.bruntlett(a)gmail.com> > Date: March 27, 2011 11:27:42 AM EDT > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] C++0x resources > Reply-To: boost-users(a)lists.boost.org > > > Hi Jim, > > On 27 March 2011 15:53, Jim Hodapp <james.hodapp(a)gmail.com> wrote: > I just saw in another post that C++0x FDIS has been voted complete and ready to ship. Is there a good website (or book) that I can start reading to get familiar with the new features? I've seen a few resources here and there on one or two topics, such as "auto", but no good place to read and see example for every new feature. Is the GNU g++ compiler the best implementation to experiment with these new language features? > > Take a look at Bjarne Stroustrup's C++ 0x FAQ:- > http://www2.research.att.com/~bs/C++0xFAQ.html > > HTH, > > > Ian > > -- > -- ACCU - Professionalism in programming - http://www.accu.org/ > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users

2 1

C++0x resources
by Jim Hodapp 27 Mar '11

27 Mar '11

Hi all, I just saw in another post that C++0x FDIS has been voted complete and ready to ship. Is there a good website (or book) that I can start reading to get familiar with the new features? I've seen a few resources here and there on one or two topics, such as "auto", but no good place to read and see example for every new feature. Is the GNU g++ compiler the best implementation to experiment with these new language features? -Jim

2 1

[C++0x] C++0x FDIS ships!
by Beman Dawes 27 Mar '11

27 Mar '11

On Friday In Madrid, Spain, the ISO C++ committee approved the Final Draft International Standard (FDIS) for C++0x. So C++0x is now complete in a technical sense. There is no possibility for further technical change. It will be a few weeks before the actual FDIS document is available. The ISO formal voting process will take until August or later. It's an up or down, yes or no, vote. All the countries actively participating voted yes for shipping the FDIS, so this ISO vote is mostly a formality. Most of the changes approved in Madrid were minor wording changes fixing mistakes in specification. Proposals to remove several features on the grounds they haven't been implemented or tested were rejected, usually by overwhelming vote margins. For example, the range-based-for-loop problem that surfaced on the Boost list caused a tweak in name lookup wording. On the library side, a lot of effort was devoted to "noexcept". A much more coherent set of usage guidelines was agreed on, and that meant going through the standard library and removing a lot of uses of "noexcept" that were over aggressive. More on that later, since Boost will probably want to use a similar set of guidelines. A straw poll for restarting work on Library Technical Report 2 passed without opposition. There was a lot of private discussion as to which Boost libraries committee members wanted to see go into TR2. --Beman

2 1

[context review] Review will end Mars 30st 2011
by Vicente BOTET 27 Mar '11

27 Mar '11

Hi all, The review of Oliver Kowalke Context library will end on Mars 30th. There were a few interesting reviews with quite constructive comments. I encourage you to make a review so you can help improving the library and also help me to take the right decision ;-) Please let me know if you needs more time so I can see if the review could be extended. Vicente Review Manager Boost.Context is a basic building block library that is used for more higher abstractions. It is used by Boost.Fiber, Boost.Tasklet and Boost.Task from the same author. I really hope to see your vote and your participation in the discussions on the Boost DEVELOPMENT mailing lists! WARNING !!!!!!!!!!!!! Only on the Boost DEVELOPMENT mailing lists Please use [context review] as prefix of your post to make easier the review follow up. You can download a frozen version for the review from the Vault, TinyURL: http://tinyurl.com/5r77rel or The documentation is available directly here: http://ok73.ok.funpic.de/boost/libs/context/doc/html/ --------------------------------------------------- About the library: Boost.Context provides a possibility to switch between different user-level context and is intended to be the basis for a higher abstraction like coroutine and fiber. A user-level context represents the current execution state, including all registers and CPU flags, the instruction pointer, the stack pointer. boost::context encapsulates such a user-level context and is able to store/restore its associated user-level context. This allows multiple execution paths running on a single thread using a sort of cooperative scheduling (in contrast a thread is preemptively scheduled) - the running boost::context decides explicitly when its yields to allow another boost::context to run (user-level context switching). A context can only run on a single thread at any point in time but may be migrated between thread. A context switch between threads costs usually thousands of CPU cycles on x86 compared to a user-level context switch with few hundreds of cycles. Boost.Context differes to setjmp()/longjmp() in the fact that the C99 standard does not require that longjmp() preserves the current stack frame, e. g. jumping into a function which was exited via a call to longjmp() is undefined. Boost.Context allows to select the mechanism provided by the operating system (ucontext on UNIX, Windows Fibers on Windows) or a faster implementation provided by this library (fcontext using assembler). --------------------------------------------------- Please always state in your review, whether you think the library should be accepted as a Boost library! Additionally please consider giving feedback on the following general topics: - What is your evaluation of the design? - What is your evaluation of the implementation? - What is your evaluation of the documentation? - What is your evaluation of the potential usefulness of the library? - Did you try to use the library? With what compiler? Did you have any problems? - How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? - Are you knowledgeable about the problem domain? Have a good review, Vicente Review Manager

1 0

filesystem::copy_directory for filesystem3 in 1.46.1 missing?
by Ákos Maróy 27 Mar '11

27 Mar '11

Hi, I'm trying to use the fileystem::copy_directory() function, in boost 1.46.1, using v3 of the filesystem library. I see that I have to reference this function by specifying the namespace boost::filesystem3 explicitly, instead of just boost::filesystem: #include <boost/filsystem.hpp> boost::filesystem::copy_directory(..) // does not compile boost::filesystem3::copy_directory(..) // compiles fine looking at boost/filystem/v3/operations.hpp, it seems that the copy_directory and the copy function is not 'exported' into the boost:filesystem namespace for some reason: namespace boost { namespace filesystem { using filesystem3::absolute; using filesystem3::block_file; using filesystem3::character_file; // using filesystem3::copy; using filesystem3::copy_file; using filesystem3::copy_option; using filesystem3::copy_symlink; ... I wonder if this is on purpose, or if this is some sort of omission... Akos

1 0

[interprocess] set doesn't work when key_type is pointer.
by skywalker 26 Mar '11

26 Mar '11

Hello all, It seems that boost::interprocess::set won't compile if the key_type is pointer. However, bip::vector or std::set works fine. I'm using g++ 4.5.2 and boost 1.46.0 with linux 2.6.37. Here is my code: // compile command: g++ a.cpp -lboost_thread namespace bip = boost::interprocess; struct comp { template<class Pointer> bool operator()(Pointer a, Pointer b) const { return *a < *b; } bool operator()(int a, int b) const { return a < b; } }; typedef bip::managed_mapped_file::segment_manager segment_manager_t; #define fset(value_t) \ bip::set<value_t, comp, bip::allocator<value_t, segment_manager_t> > bip::managed_mapped_file managed_shm(bip::open_or_create, "/tmp/bip_test", 1024); void bip_example() { fset(int*) s(comp(), managed_shm.get_segment_manager()); s.insert(new int (7)); ///< error: key_type can't be pointer in bip. fset(int) s2(comp(), managed_shm.get_segment_manager()); s2.insert(int (7)); ///< ok: non-pointer is accepted. std::set<int*, comp> s3; s3.insert(new int (7)); ///< ok: int* works with STL, as well as shared_ptr, offset_ptr. } void more_example() { typedef bip::offset_ptr<int> optr; fset(optr) s3(comp(), managed_shm.get_segment_manager()); bip::allocator<int, segment_manager_t> alloc_int = managed_shm.get_segment_manager(); optr p = alloc_int.allocate(1); alloc_int.construct(p, 7); s3.insert(p); ///< error: bip doesn't work with offset_ptr too. typedef boost::shared_ptr<int> sptr; fset(sptr) s4(comp(), managed_shm.get_segment_manager()); s4.insert(sptr(new int(7))); ///< ok: shared_ptr works fine. } Won't that fail to compile, or am I missing something? Regards, Tianjiao

2 1

[Graph] Difference between DF search and visit / Graph Printing
by Marcus Riemer 26 Mar '11

26 Mar '11

Hi there! I am struggling a little understanding the difference between the DF search and visit. I thought that the DF visit would revisit "back" nodes while the DF search would not. But as I just found out that does seem to be the difference. What I want: I would like to print all following nodes for every node of a acylic graph. Lets say I have this very simple graph, I hope spaces are preserved: -> Long \ Begin / -> End \ -> -> Short / I would like to print that graph in the following Fashion: Begin .Long ..End .Short ..End I don't mind the exact order of "Long" and "Short", as long as they are both printed. So far I did this by implementing the following dfs_visitor<> [0]. I then apply the visitor with a call to boost::depth_first_visit() and my graph. However, the problem is, that the "End" node is only visited, and therefore printed, once. Of course I could roll out the DF visit myself, but as I will be demonstrating this at my unversity, I would prefer a clean solution. So what am I probably missing to print the graph in the fasion I want it? [0] http://pastebin.com/u43TtU7F

3 4

encryption with boost.iostreams
by Sean Farrow 26 Mar '11

26 Mar '11

Hi: I'm currently writing a serialization routine that needs to encrypt data. Are there any examples of using encryption with boost.iostreams. Also, given an xml file I need to determine that a file is encrypted and then decrypt the file. Anyone got any experience with this? Cheers Sean.

1 0