- Boost-users - lists.boost.org

Problems with (de)serialization from/through pointer to base class
by Rafael Arco Arredondo 24 Oct '11

24 Oct '11

Hello: I am encountering problems with boost::serialization, particularly when objects are sent with boost::mpi and contain pointers to a base class. I am getting segmentation faults, but not always, only when the transmitted objects exceed a particular size. I have two abstract classes, A and B, two instantiable classes, D and E, and a fifth class C, which is the class eventually serialized and contains one pointer to an object of class A and another to an object of class B. In the example attached, when n (which determines the size of E objects) is 1000, everything works fine. If, for instance, n=2000, the segmentation fault appears when the object is deserialized after calling comm.recv. The output for this last case is also attached in file output.log. I am also using boost::shared_ptr as well to wrap class C objects, but the error is also thrown when no shared pointers are used. When types and object sizes are known a priori there are no problems, but I would need something more general and dynamic right now. It is possible to do this, isn't it? Any ideas of what is happening? Am I doing something wrong? Did I miss something in the documentation? Is there any size limitation for serializing/deserializing objects? I imagine MPI_Unpack, which appears in the stack trace, is restricted by the size of the input buffer, indicated by an int, but I believe this is not the problem here. Is it perhaps a bug? Any help or comments will be appreciated. Thank you in advance, Rafa

1 2

[optional] Working around gcc 4.6 bug which produces error "‘*((void*)& aNewItem +8)’ may be used uninitialized in this function"
by Rob Desbois 24 Oct '11

24 Oct '11

I've just begun having issues with some use of boost::optional producing the error shown in the subject. This is a gcc bug, filed at http://gcc.gnu.org/bugzilla/show_bug.cgi?id=47679 The corresponding downstream bug for RedHat/Fedora is at https://bugzilla.redhat.com/show_bug.cgi?id=676361 and has a code listing to reproduce. Sending this partially as an FYI, and also to ask if anyone has any workaround? Altering client code to construct with boost::none or a false condition, then assigning a value if required, still yields the error on the assign operation. My current workaround is to initialize the optional<T> in question with a dummy T instance, then call reset() with no params if I want an 'uninitialized' optional. It looks rather bizarre, but if it must be then it must be. -- Rob Desbois Eml: rob.desbois(a)gmail.com Blog: http://theotherbranch.wordpress.com/

1 0

[regex] Grouping behaviour
by U.Mutlu 24 Oct '11

24 Oct '11

Hi, I need to split a regex pattern into two groups like from this blafoo to (bla)(foo) But the second variant matches also input like this one: blaABCfoo There is no .* between the groups, so I wonder why it still matches. How to fix this?

2 2

Re: [Boost-users] [locale] IOStream encoding conversions?
by Artyom Beilis 24 Oct '11

24 Oct '11

> Currently I'm messing about with Boost.Locale, I'm intending to use it > for my project, but I don't know how to do something along the lines of > specifying the external encoding (usually the user's locale, en_US.UTF-8 > in my case), and the internal coding of my program (UTF-16), > to have input and output converted, possibly using a codecvt. A possible example > would be translating UTF-16-≥POSIX UTF-8 paths, or UTF-8-≥Windows UTF-16 > paths. I assume you use Windows and wide characters. It is enough to imbue the locale to the output **wide** stream and write to it. boost::locale::generator gen; // stdout std::wcout.imbue(gen("")); std::wcout << L"שלום" << std::endl; // UTF-8 file I/O std::wfstream stream; stream.imbue(gen("")); stream.open(name,std::wfstream::out); stream << L"שלום" << std::endl; Would actually output the text in narrow encoding. Also there is an example how to create custom wide streams that would convert wide to narrow encoding http://svn.boost.org/svn/boost/trunk/libs/locale/doc/html/charset_handling.… Also for converting things like path UTF-8/UTF-16 you can use utf_to_utf functions that work just on chunks of text. http://svn.boost.org/svn/boost/trunk/libs/locale/doc/html/group__codepage.h… Artyom Beilis -------------- CppCMS - C++ Web Framework: http://cppcms.sf.net/ CppDB - C++ SQL Connectivity: http://cppcms.sf.net/sql/cppdb/

1 0

Portable stack trace mechanism?
by Chris Cleeland 24 Oct '11

24 Oct '11

I've written this code several times and thought boost might have something. No, it doesn't, but I found this old bit thread: http://lists.boost.org/Archives/boost/2010/10/172303.php >From the evidence I can conclude that it never made it in, but the thread gives no indication as to why. I don't want to resurrect something that's already been fully hashed and excluded, but if it never made it in simply because nobody carried the torch forward, perhaps I could be the new bearer.

6 5

[Review] Boost.Atomic review period extended
by Tim Blechmann 24 Oct '11

24 Oct '11

Hi all, Since there has been only rather limited discussion about the Boost.Atomic library and only one full review, the review period will be extended until next Sunday, October 30th. I encourage everyone with an interest in this library to participate in the remaining days of the review. If someone wants to do a review, but won't find the time during the review period, please let me know how much time you need. Cheers, Tim Review Manager

1 0

[serialization] how can load the nth element using binary_archive?
by gongyiling 24 Oct '11

24 Oct '11

Hello, Everyone: boost::serialization is a cool library. I use it in my project. I encounter a problem now. I serialized a array of objects to file as following: std::ofstream ofs(filename, std::ios_base::binary); boost::binary_oarchive oar(ofs); oar << obj[0] << obj[1] << ... << obj[n-1]; Now I want to load the obj[i] (i>=0 && i<n) from file. I tried recording obj[i]'s position in file, then set file stream's pointer to that position and trying to load it, but failed with exception input_stream_error, can I by any means doing that? here is my minimun code segments: struct word_block { std::string word; template<typename Archive> void serialize(Archive & ar, const unsigned int file_version ) { ar & word; } }; int main(int argc, char* argv[]) { const char* filename = "words.idx"; word_block wb1, wb2, wb2_r; std::ofstream ofs(filename, std::ios_base::binary); boost::archive::binary_oarchive oar(ofs); oar << wb1; size_t wb2_pos = ofs.tellp(); //record the position of wb2. oar << wb2; ofs.close(); std::ifstream ifs(filename, std::ios_base::binary); boost::archive::binary_iarchive iar(ifs); ifs.seekg(wb2_pos); //set file pointer to wb2. iar >> wb2_r; //failed with exception input_stream_error. return 0; } Any idea is appreciated!

4 3

ublas speed and multi-core processors
by petros 23 Oct '11

23 Oct '11

Hi, Is there any way to use ublas taking advantage of the multi-core facility of current processors? Even naive “strategies”, like breaking the number of rows of a targeted matrix expression across different threads would improve performance significantly, even if not optimally. It would be nice to be able to use the matrix expression templates to perform operations in a multi-threaded way. Is there any provision for this? Thank you for your help, Petros

3 4

ublas matrix_range resize
by petros 22 Oct '11

22 Oct '11

HI, I wanted to ask how I can resize a matrix_range object. I understand that is is a very light/disposable object but here is the problem I want to solve: I need to wrap some Lapack functions (Linear equation solvers) using ublas. The package I am using (mkl) asks for specific memory requirements that translate to having a –possibly- larger lda (basically the hosting matrix number of rows). So my way to address this was to contain a general matrix (ublas::matrix) and derive from a matrix_range which references it. Alas, I cannot resize! Is there a way? – Yes I could contain a pointer to the matrix_range and reconstruct, but this is rather silly. There should be a way to assign a matrix_range to an existing one, overwriting its contents. Thank you, in advance, for your help, Petros

1 0

boost serialization- handle endianness
by ohad barta 22 Oct '11

22 Oct '11

Hi I am serializing my data in a one computer, transfer it to another computer and desirialize the data there,.Does the data can get messed because of the 2 endian option, or is boost lobrary take care of it automatically? If I do need to handle this myself, are there any functions/MACROS in the boost serialization library that will help me? Many Thanks Ohad Barta

2 1