- Boost-users - lists.boost.org

[iterator] lazy container
by alfC 23 Jun '10

23 Jun '10

Hi, I am implementing a lazy container/iterator that represents the results of certain iterative algorithm. I have a class (call it solver) that should look like an interator. First I was using boost.operators to make it look like a pointer but then I found boost.Iterator. Because of the nature of the algorithm it can be only traversed forward. I am using it in this way: class solver : //public boost::dereferenceable<solver, result const*>, //replaced with boost::iterator public boost::iterator_facade<solver, result const, boost::forward_traversal_tag, result const /*noref*/ >{ ... result dereference() const{ //was result operator*() const ... } void increment() {...} //was operator++ } note that tipically the last argument is intended to be a reference but in this case, since there in no underlying data the return can't be a reference but a new element generated by the algorithm. Question: is this the intended use of boost::iterator_facade is there an alternative base class for lazy operators? Thank you, Alfredo

1 0

[serialization] no RTTI data error
by Noah Roberts 22 Jun '10

22 Jun '10

I'm hoping someone can help me narrow down what I might be looking for. Already spent a lot of time trying to get somewhere and still trying to step through but using VS2010 and it's slower than ANYTHING ELSE THAT EXISTS!! The problem is that when I load my file I get an, "Access violation - no RTTI data," error. Yes, RTTI is turned on and the application actually relies upon it very heavily throughout its operation (acyclic visitor is a major component). The problem seems to have just cropped up and is within code that has previously passed tests. Best I can do at this point is describe the problem and hope someone has an idea of what to look for because I've been inch by inch stepping through serialization template code with a debugger that takes 30 seconds each damn step. I won't be done for hours. (anyone know how to speed this bitch up?) The problem seems to involve a type that I'm using save/load_construct_data on. There is a collection of these objects that is stored and loaded previously in the file and then I try to load a vector of objects that have shared pointers to these previously loaded objects on an individual basis. It was working fine and then I attached more data after the vector. Now it throws this exception when I try to load the pointer; before it reaches the new code that is otherwise entirely unrelated (it's just an object that happens to be stored in the same parent). If and when I get more information I'll post but for now I'm hoping someone can save me some time and frustration. This debugger is the slowest damn thing I've ever seen in my life. -- http://crazyeddiecpp.blogspot.com/

2 3

asio: tcp::iostream::async_connect with specific io_service instance
by Walther Zwart 22 Jun '10

22 Jun '10

I have a question about boost asio iostreams. I'm trying to create a server and client with tcp::iostreams that use async_connect and async_accept and use a specific tcp::io_service instance. My code currently looks somewhat like this. Server: ------------ io_service my_service; tcp::endpoint my_endpoint; ... ... tcp::acceptor my_acceptor(my_service, my_endpoint); tcp::iostream my_stream; my_acceptor.async_accept(my_stream.rdbuf(), boost::bind(connectedHandler, placeholders::error)); ------------ This works the way I want to. Another thread calls run on the io_service and thus makes sure the accept is handled. Client: ------------ io_service my_service; tcp::endpoint my_endpoint; ... ... tcp::iostream my_stream; my_stream.rdbuf()->async_connect(my_endpoint, boost::bind(connectedHandler, placeholders::error)); ------------ This doesn't work with my io_service. I don't see a way to tell the tcp::iostream or the underlying basic_socket_streambuf which io_service to use. This means I have to call run on a new io_service every time I connect to a socket somewhere, right? Is there a way to do an async_connect with your own service? regards, Walther

1 0

Providing a string-to-X feature
by John Dlugosz 22 Jun '10

22 Jun '10

Suppose I wanted to expose a "parse from string" feature for my data type. For discussion, I'll suppose it's a date/time type since everyone knows that and it illustrates the worst problems. What is the current thought on the "best practices" way to expose this? Maybe I'll start with a stream operator>>, and make sure it works with lexical_cast<X>(string). But, where do I add the extra options that tells it what to allow or assume when parsing? Dates for example might have different formats and it can recognize any legal format, but you can tell it to ignore some since that clashes with other input data. It contains spaces and trailing components are optional, so it doesn't know when to stop until it runs into something that doesn't fit. In some sense, that ought to be a constructor. But it should take any kind of string-like thing or stream, and the declaration might clash with other more normal constructors, so a purpose-named "take from human-readable form" is better. An earlier discussion touched on whether to use lexical_cast or a to_string function. This is the other direction, but might have similar issues. Only from_string doesn't seem to be as omni-present as an expectation. Any suggestions? --John (big footer ahead) TradeStation Group, Inc. is a publicly-traded holding company (NASDAQ GS: TRAD) of three operating subsidiaries, TradeStation Securities, Inc. (Member NYSE, FINRA, SIPC and NFA), TradeStation Technologies, Inc., a trading software and subscription company, and TradeStation Europe Limited, a United Kingdom, FSA-authorized introducing brokerage firm. None of these companies provides trading or investment advice, recommendations or endorsements of any kind. The information transmitted is intended only for the person or entity to which it is addressed and may contain confidential and/or privileged material. Any review, retransmission, dissemination or other use of, or taking of any action in reliance upon, this information by persons or entities other than the intended recipient is prohibited. If you received this in error, please contact the sender and delete the material from any computer.

1 0

matrix_slice access, assignment, issues
by Eric Shamay 22 Jun '10

22 Jun '10

I've been trying to form a class based on the matrix_slice to ease some operations. Thus far I have: http://codepad.org/EepVDx2C The compiler complains about the line where I form a matrix_slice (line 21 in the paste), and specifically with the first argument. I'm lost as to why I can't submit a matrix<double> as the first argument in the manner I've tried. Any help would be greatly appreciated. ~Eric

2 1

[MPI] error in hybrid OpenMP + Boost.MPI application
by Riccardo Murri 22 Jun '10

22 Jun '10

Hello, I have an OpenMP+MPI application, which crashes with an exception on some inputs. The backtrace shows that the error originates in the Boost.MPI code:: terminate called after throwing an instance of 'boost::archive::archive_exception' what(): unregistered class ... [compute-0-7:23208] [ 0] /lib64/libpthread.so.0 [0x3110c0e4c0] [compute-0-7:23208] [ 1] /lib64/libc.so.6(gsignal+0x35) [0x3110430215] [compute-0-7:23208] [ 2] /lib64/libc.so.6(abort+0x110) [0x3110431cc0] [compute-0-7:23208] [ 3] /usr/lib64/libstdc++.so.6(_ZN9__gnu_cxx27__verbose_terminate_handlerEv+0x114) [0x31114bec44] [compute-0-7:23208] [ 4] /usr/lib64/libstdc++.so.6 [0x31114bcdb6] [compute-0-7:23208] [ 5] /usr/lib64/libstdc++.so.6 [0x31114bcde3] [compute-0-7:23208] [ 6] /usr/lib64/libstdc++.so.6 [0x31114bceca] [compute-0-7:23208] [ 7] /home/oci/murri/sw/lib/libboost_serialization.so.1.43.0(_ZN5boost7archive6detail19basic_iarchive_impl12load_pointerERNS1_14basic_iarchiveERPvPKNS1_25basic_pointer_iserializerEPFS9_RKNS_13serialization18extended_type_infoEE+0x23b) [0x2b3d485348ab] [compute-0-7:23208] [ 8] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost7archive6detail17load_pointer_typeINS_3mpi15packed_iarchiveEE6invokeIPN9Waterfall9Processor9SparseRowEEEvRS4_RT_+0x49) [0x4d7989] [compute-0-7:23208] [ 9] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost7archive4loadINS_3mpi15packed_iarchiveEPN9Waterfall9Processor9SparseRowEEEvRT_RT0_+0x22) [0x4d79f6] [compute-0-7:23208] [10] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost7archive6detail15common_iarchiveINS_3mpi15packed_iarchiveEE13load_overrideIPN9Waterfall9Processor9SparseRowEEEvRT_i+0x28) [0x4d7a20] [compute-0-7:23208] [11] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost7archive21basic_binary_iarchiveINS_3mpi15packed_iarchiveEE13load_overrideIPN9Waterfall9Processor9SparseRowEEEvRT_i+0x23) [0x4d7a45] [compute-0-7:23208] [12] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost3mpi15packed_iarchive13load_overrideIPN9Waterfall9Processor9SparseRowEEEvRT_iN4mpl_5bool_ILb0EEE+0x23) [0x4d7a6b] [compute-0-7:23208] [13] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost3mpi15packed_iarchive13load_overrideIPN9Waterfall9Processor9SparseRowEEEvRT_i+0x2a) [0x4d7a98] [compute-0-7:23208] [14] /home/oci/murri/rank/rank_wf_dbg(_ZN5boost7archive6detail18interface_iarchiveINS_3mpi15packed_iarchiveEErsIPN9Waterfall9Processor9SparseRowEEERS4_RT_+0x2a) [0x4d7ac4] [compute-0-7:23208] [15] /home/oci/murri/rank/rank_wf_dbg(_ZNK5boost3mpi12communicator9recv_implIPN9Waterfall9Processor9SparseRowEEENS0_6statusEiiRT_N4mpl_5bool_ILb0EEE+0x93) [0x4f41ef] [compute-0-7:23208] [16] /home/oci/murri/rank/rank_wf_dbg(_ZNK5boost3mpi12communicator4recvIPN9Waterfall9Processor9SparseRowEEENS0_6statusEiiRT_+0x3f) [0x4f427b] [compute-0-7:23208] [17] /home/oci/murri/rank/rank_wf_dbg(_ZN9Waterfall4rankEv+0x291) [0x4acf27] [compute-0-7:23208] [18] /home/oci/murri/rank/rank_wf_dbg(main+0x698) [0x4ad9b6] [compute-0-7:23208] [19] /lib64/libc.so.6(__libc_start_main+0xf4) [0x311041d974] [compute-0-7:23208] [20] /home/oci/murri/rank/rank_wf_dbg [0x4aa219] [compute-0-7:23208] *** End of error message *** One MPI rank is started per compute node; all OpenMP threads may call mpi::isend(); only one will do mpi::iprobe()/mpi::recv(). Although the above error is in the mpi::communicator::recv(), serializing the mpi::isend() calls apparently solves the issue; similarly, the program runs fine if I run it on one node only; with some other (smaller) inputs, it runs fine as well. This leads me to think that it is a thread-safety issue with the MPI part. I have checked that the MPI library (OpenMPI 1.4.2) is initialized with MPI_Init_threads() and provides the threading level MPI_THREAD_MULTIPLE. So, question: is there a (known) thread-safety issue with Boost.MPI, or should I definitely look somewhere else? Thanks for any help! Riccardo

2 1

[BGL] algorithms and listS as vertex list container
by Christophe Goessen 22 Jun '10

22 Jun '10

Hello, I recently tried to use some (almost all) shortest path algorithms from BGL. But since my program remove a lot of vertices, I opted for an adjacency list graph with list as vertex list container instead of vector. > boost::adjacency_list <boost::setS, boost::listS, boost::bidirectionalS,[...] This seems to trigger some troubles at compilation level. If I try anything else than vector for vertex list I'm rewarded by similar errors: > boost/graph/relax.hpp:55: error: no matching function for call to ‘put(long unsigned int*&, void*&, long unsigned int)’ My question is as follow, do algorithms handle graphs with list (or set) as vertex list containers? I searched for information about this on the internet and in the documentation but failed to find any clues. Thanks -- Christophe

2 5

Erorr compiling application with boost 1.43.0 on WinXp/VS2008
by Ovidiu MacSiniuc 22 Jun '10

22 Jun '10

Hello everybody, I am developing some application on Visual Studio 2008. I am working on Windows XP (sp3) and our minimim requirement platform for the applcation is also WinXp. I downloaded and compiled the lasta boost version (1.43.0). However, when compiling the application I'm getting this error: C:\Boost\1.43.0\boost/asio/detail/win_fenced_block.hpp(48) : error C3861: 'MemoryBarrier': identifier not found It looks like the MSVC wants to build it as a Vista application since according to http://msdn.microsoft.com/en-us/library/ms684208%28VS.85%29.aspx minimum supported client for MemoryBarrier(void) is Windows Vista. What would be the proper fix for this issue? Thanks, Ovidiu

2 4

[Proto] Building custom expression and new domain deduction not working
by Joel Falcou 22 Jun '10

22 Jun '10

Here's an update on my last version. I have now added a new function building a custom expression node in a given domain. http://codepad.org/ddjFylti This code fails to compile under current trunk by saying that there is no proper operator+ to find. I guess it's because it can't see t and iota(0,8) to be in the same , compatible domain. Did I do something wrong ?

3 2

[boost][lambda] How to adapt container objects?
by Peter Barnes 22 Jun '10

22 Jun '10

I need to create an index using a std:vector as the sort key. boost::lamda and std::sort seem just the thing: // Data vectors // // Goal is to sort index based on values of random // std::vector<int> index(10); // 0, 1, 2, 3, 4, 5, 6, 7, 8, 9 std::vector<int> random(10); // random values, not in order namespace bll = boost::lambda ; // Sorted through vector pointer // // This compiles // std::vector<int> * ranp = & random; std::sort(index.begin(), index.end(), bll::var(*ranp)[bll::_1] < bll::var(*ranp)[bll::_2] ); In reality, my std::vector is wrapped in another object (for reasons given below, but unrelated to this question), with an operator[] to retrieve contents: // Container wrapper class Container { std::vector<int> _v; public: Container(std::vector<int> v) : _v(v) { } ; const int operator[](const int i) { return _v[i]; } }; // Sorted through container object // // This *does not* compile // Container * cvp = new Container(random); std::sort(index.begin(), index.end(), bll::var(*cvp)[bll::_1] < bll::var(*cvp)[bll::_2] ); /* Compiler error (g++ 4.01 (Apple)): obj-indexer.cc:95: instantiated from here boost/lambda/detail/operator_lambda_func_base.hpp:137: error: conversion from 'const int' to non-scalar type 'boost::lambda::detail::unspecified' requested */ I don't understand the error message, and can't make out from the documentation why this doesn't work. Do I need to extend the return type deduction system somehow? But operator[] is not one of the action types. Thanks for your help, Peter Aside: Why do I wrap std::vector inside Container? I'm memory constrained, so I combine a std::vector<uint16_t> with a std::map to catch the overflows. The operator[i] then computes int ret = vector[i] + ( (o = map.find(i)) != map.end() ? o->second * 2^16 : 0); ____________ Peter Barnes pdbj(a)me.com ____________ Peter Barnes pdbj(a)me.com

2 1