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

Compilation problems on Linux
by mayankg＠iiitb.ac.in 31 Mar '04

31 Mar '04

Hi, We are trying to use the libboost_python.so library in Linux with our project. When we try to compile the C++ file of our project having all the boost interfaces, we get the following compilation error only on Linux. Although on Windows it works fine. We are using the Python 2.2.3 and Boost 1.30.2 versions. /usr/include/boost/python/converter/as_to_python_function.hpp:28: sorry, unimplemented: ` method_call_expr' not supported by dump_expr /usr/include/boost/python/converter/as_to_python_function.hpp:28: sorry, unimplemented: ` method_call_expr' not supported by dump_expr /usr/include/boost/python/converter/as_to_python_function.hpp:28: sorry, unimplemented: ` method_call_expr' not supported by dump_expr /usr/include/boost/python/converter/as_to_python_function.hpp:28: sorry, unimplemented: ` method_call_expr' not supported by dump_expr Please help, Regards, Mayank

2 1

Location and namespace of a boost library (say, iterators)
by Deane Yang 31 Mar '04

31 Mar '04

I'm sorry if I'm being dumb about this, but I've run into the following difficulty a couple of times with boost libraries. I've decided to try using the new boost iterator library. Based on the documentation, I think I want to use iterator_adaptor. But scanning through the documentation, I don't immediately see - what namespace iterator_adaptor lives in. I assume it's just boost, but it probably doesn't hurt to mention this explicitly, since some boost libraries use sub-namespaces. - what header file to include. I couldn't find this in the documentation. I can figure all of this out, but it is helpful to have it specified in the documentation.

2 1

Installing Boost 1_30_0 on HPUX
by Albert Gome 31 Mar '04

31 Mar '04

Hi, I am new to BOOST and i like to install that on HPUX11.11. I have downloaded BOOST_1_30_0 and have gcc installed on HPUX. What is the efficient way of installing it. ie the parameters with the bjam. Thanks Albert __________________________________ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html

2 1

[BGL] Finding out what properties exist in a graph?
by Steve Juranich 30 Mar '04

30 Mar '04

Is there an easy way to find out all of, for instance, the vertex properties associated with a graph? What I'm looking for is something like the following: <pseduocode> vector<properties> my_vertex_properties; property_iterator pi, last; for(tie(pi, last) = vertex_property_list(Graph); pi != last; ++pi) my_vertex_properties.push_back(*pi); </pseudocode> How about it? Does anything like this exist or am I just daydreaming again? Thanks. -- Stephen W. Juranich Science Applications Intl. Corp. (SAIC) (520) 570-7706

2 2

Serialization library
by Steve Folly 30 Mar '04

30 Mar '04

Hi, I'm looking using the Robert Ramey's serialization library for our own projects and would like to know it's current status. Obviously, it's not in 1.31.0, so what are the plans for it's future inclusion? Thanks. Steve.

2 1

graph, dfs search
by Peter Aronsson 30 Mar '04

30 Mar '04

Hi. I am having a problem with depth_first_search. The thing I do different is that I do not put any info in a property map and I do not want to bother myself with colormaps etc. The application code does: Graph g; pass2_visitor vis; depth_first_search(g,boost::visitor(vis)); But I get a compilation error, saying that it can not find a suitable operator+ in the property_map... See attached files for complete example. Grateful for all help! (I am running from HEAD on sourceforge cvs ) Regards Peter Aronsson -- _________________________________________________________________ / Peter Aronsson, Phd Student at PELAB (Programming Environments \ | Laboratory ) Department for Computer & Information Science | | Linköping University, Sweden | |=================================================================| | petar(a)ida.liu.se , phone +46 (0)13-28 1737 Room 3B:490 | \_________________________________________________________________/ #include "tg.hpp" int main() { TaskGraph g; pass2_visitor vis; depth_first_search(g,boost::visitor(vis)); } #ifndef _TASKGRAPH_H #define _TASKGRAPH_H #include <boost/config.hpp> #include <iostream> // for std::cout #include <utility> // for std::pair #include <algorithm> // for std::for_each #include <boost/utility.hpp> // for boost::tie #include <boost/graph/graph_traits.hpp> // for boost::graph_traits #include <boost/graph/adjacency_list.hpp> #include <boost/graph/adjacency_iterator.hpp> #include <boost/graph/graphviz.hpp> #include <boost/graph/depth_first_search.hpp> //Task graph types using namespace std; using boost::tie; struct vertex_execcost_t { typedef boost::vertex_property_tag kind; }; struct vertex_unique_id_t { typedef boost::vertex_property_tag kind; }; struct edge_result_set_t { typedef boost::edge_property_tag kind; }; typedef boost::property<boost::vertex_name_t,string, boost::property<vertex_execcost_t,float, boost::property<vertex_unique_id_t, int > > > VertexProperty; typedef boost::property<boost::edge_weight_t, int > EdgeProperty; typedef boost::adjacency_list<boost::listS, boost::listS, boost::bidirectionalS, VertexProperty, EdgeProperty> TaskGraph; typedef boost::graph_traits<TaskGraph>::vertex_descriptor VertexID; class pass2_visitor : public boost::default_dfs_visitor { public: pass2_visitor(/* initialize containers to store stuff in*/) { }; template <typename Vertex, typename Graph> void discover_vertex(Vertex u, const Graph & g) const { // Do some stuff }; }; #endif

2 4

CVS import gotcha - beware!
by Steve Folly 29 Mar '04

29 Mar '04

Hi, Most of you are probably aware of this, but I just thought I'd share my experience with trying to import the new version of boost into our CVS repository using 'cvs import': It wasn't recognising new folders - e.g. boost/spirit/core and all it's contents. I was puzzling it for a while until I realised the import output was showing 'I' next to boost/spirit/core. Aha - it's ignored. DOH! Well, core files are ignored, aren't they! To all those who have been long suffering with this : don't forget to include '-I !' in the command to include all files!! I'll be happy if it helps at least one person! :) -- Regards, Steve.

1 0

RE: [Boost-users] Defining less on member variables/functions
by Chris Jurney 29 Mar '04

29 Mar '04

In this case, I could have used pair's operator<, but if I have a class by value as the second item of the pair that doesn't have a meaningful operator<, I want to ignore it in the sort. Thanks for all the help, guys. Chris -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Victor A. Wagner Jr. Sent: Friday, March 26, 2004 8:40 AM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Defining less on member variables/functions why do you want to sort ONLY on the 1st element of the pair? At Thursday 2004-03-25 11:34, you wrote: >content-class: urn:content-classes:message >Content-Type: multipart/alternative; > boundary="----_=_NextPart_001_01C41297.D800D40A" > >I've had this same pattern come up 4 or 5 times now, and each time I've >been unable to find a stl + boost solution that allowed me to do what I >want without writing my own less function. Basically, I want something >that will let me do a sort, or define a map on structs or pairs without >writing a less that compares the member variable or ..first. My last attempt: > >vector<pair<float, float>> myVector >std::sort(myVector.begin(), myVector.end(), _1.first < _2.first); > >No such luck. It looks like I can do operator * and & on lambda objects, >but not access members through them. Is there anything in boost, stl, or >elsewhere that lets me not write this: > >class myLess >{ > bool operator() (pair<float, float> a, pair<float, float>b) > {return a.first < b.first;} >} > >Thanks for any help, >Chris > > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users Victor A. Wagner Jr. http://rudbek.com The five most dangerous words in the English language: "There oughta be a law" _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

1 0

Microsoft Managed C++
by Roger Wilco 29 Mar '04

29 Mar '04

I have a somewhat large (~100K lines) C++ server application. It uses Boost (lots of smart pointer use, thread use, regex), Xerces XML libraries, some MFC for GUI, and ADO for database access. I tried to build a test build of my application using Microsoft's Managed C++. I got all my code to compile, but I can't link to either Boost or Xerces XML. Is this possible? Has anyone done this? If I ultimately go the .NET route, I would replace Boost and Xerces with functionality provided by the native .NET runtime libraries. However, that would be a massive time consuming undertaking. At this point, I just want to dip my toes in the water and observe pros/cons of .NET, and have the option of migrating to .NET piecemeal. I have solid experience with C# and Java and I am extremely experienced with C++ (10 years). My main problems with C++ are: - Compile times. On my 2.8GHz dev system, a rebuild takes ~20 minutes and even a build after making a minor change takes 45 seconds. It's amazing how VC++ build speeds really have remained so stagnant despite increases in hardware speed. I realize VC++ is the slower of the IDEs for build times. - Debugging memory corruption or thread concurrency issues. This is much easier in Java/C# and the tools for assisting are much easier to use. I spent many days trying tools such as Purify for C++ and couldn't produce any fruitful results. - Diagnosing memory usage. My server application uses lots of memory. After running for several hours under production conditions it will reach around 400MB-1.3GB depending on the tasks its doing. I'd like to get a general overview of where this memory is being used. I don't suspect blatant memory leaks (new with no delete) but instead runaway caches or legitimate usage. These issues are of course still present in C#/Java but the diagnostic tools to address these issues seem to be much better. I haven't found any usable C++ tool for this issue and I rely on manually logging data points and manually trying to infer an idea of what is happening. - Java/C# is overall a better (easier/faster to develop/maintain) development platform for this type of server application work. However, I will happily stick with C++ if I can resolve these other issues. And I am of course, concerned about any performance hit when migrating to Java/C#. We max out CPU and RAM use, so these factors aren't important. I couldn't take more than a 25% drop in efficiency. Any other suggestions on my situation?

2 1

SQLite
by Witz 28 Mar '04

28 Mar '04

I don't know if this is the forum for bringing this up but ... I was wondering if anyone thinks SQLite might make a good addition to Boost - www.sqlite.org - if the author would agree to it. Ian Whittley

2 1

boost::format with non-standard string class
by Scott Lamb 27 Mar '04

27 Mar '04

Hi, I'm not real happy with std::string, largely because it seems poorly suited to UTF-8. I'm happier with other classes such as Glib::ustring. [1,2] Is it possible to use boost::format with such a class? I'd previously thought that boost as a whole expected you to use std::string, but the recent discussion of boost::regex + Glib::ustring gave me some hope. It seems the requirements for boost::regex are more simple - a bidirectional, read-only character (not byte) iterator. That should be easy enough to provide for any UTF-8 string class. I do see explicit mention of std::basic_string in the boost::format headers, but I thought I'd ask anyway. (I likely will find or make another class to avoid pulling in all of glib. But it would be similar to Glib::ustring.) [1] - more in-depth discussion of Glib::ustring vs. std::string at <http://www.gtkmm.org/gtkmm2/docs/reference/html/ classGlib_1_1ustring.html#_details> [2] - additional reasons for disliking std::string at <http://www.sgi.com/tech/stl/string_discussion.html> Thanks, Scott

1 0

boost/graph/push_relabel_max_flow::is_flow() bug?
by Arun Bhalla 27 Mar '04

27 Mar '04

Hi, I'm not really a user of Boost, but I've been referring to the Boost Graph Library code for another project. In particular, I've been looking at push_relabel_max_flow.hpp. I was writing a unit test which followed the same ideas of the is_flow() test, and it was failing on what should be a flow network with a valid maximum flow. The code in question: // a is an edge_descriptor if (capacity[a] > 0) if ((residual_capacity[a] + residual_capacity[reverse_edge[a]] != capacity[a]) || (residual_capacity[a] < 0) || (residual_capacity[reverse_edge[a]] < 0)) return false; I contend the test should actually be: // a is an edge_descriptor if (capacity[a] > 0) if ((residual_capacity[a] + residual_capacity[reverse_edge[a]] ! != capacity[a] + capacity[reverse_edge[a]]) || (residual_capacity[a] < 0) || (residual_capacity[reverse_edge[a]] < 0)) return false; I refer to an example network in CLR (1st ed.), p.589. There are several vertices, but I focus on vertices v1 and v2. In this graph, we have capacities c(v1,v2) = 10 c(v2,v1) = 4 and residual capacities cf(v1,v2) = 11 cf(v2,v1) = 3 With the above (original) test, it would fail because 11 + 3 (the sum of residual capacities) != the capacity of either edge. This would hold true if the sum of residual capacities were compared with the sum of capacities. Ordinarily this test would work if reverse_edge[a] has a capacity of 0, which would be the case if the edge (a) is in the flow network, but its reverse is not. I'm not a graph theory expert, though, so I defer to anyone who is. :) Thanks, Arun

1 0

Defining less on member variables/functions
by Chris Jurney 26 Mar '04

26 Mar '04

I've had this same pattern come up 4 or 5 times now, and each time I've been unable to find a stl + boost solution that allowed me to do what I want without writing my own less function. Basically, I want something that will let me do a sort, or define a map on structs or pairs without writing a less that compares the member variable or .first. My last attempt: vector<pair<float, float>> myVector std::sort(myVector.begin(), myVector.end(), _1.first < _2.first); No such luck. It looks like I can do operator * and & on lambda objects, but not access members through them. Is there anything in boost, stl, or elsewhere that lets me not write this: class myLess { bool operator() (pair<float, float> a, pair<float, float>b) {return a.first < b.first;} } Thanks for any help, Chris

6 8

Re: [Boost-users] MAC OS X installation of boost libs
by Michael Short 26 Mar '04

26 Mar '04

Hello I am using MAC OS X 10.3.3. First I done find the bjam tool? Second when I build the bjam tool myself I get the following error. Jamfile:71: in module scope rule unless unknown in module. I think if you could please just give me a quick step by step of what you did to setup the libs with the correct directorys you used then that would probably fix the problems. I did try following the instructions on the page you stated but just did not help. thanks Michael On Tuesday, March 23, 2004, at 09:25PM, James Mastro <jmastro(a)rochester.rr.com> wrote: >On Mar 22, 2004, at 5:12 PM, Michael Short wrote: > >> The only problem is that I am having problems installing and building >> the libs on my MAC OS X system. > >What kind of problems? What version of OS X? I built boost 1.31.0 on >OS X 10.3.2. I think the regex library gave me some problems, but I >believe everything else built OK. > >I just followed the directions here: > >http://www.boost.org/more/getting_started.html > >-jim > > >

4 7

MultiArray questions
by Dr Mark H Phillips 24 Mar '04

24 Mar '04

Hi, I've recently found out about the existence of MultiArray and I am quite impressed! I was thinking of implementing something similar myself but now hopefully I don't have to! However I have a few questions that I'm hoping someone can answer. Question 1 ---------- In the user documentation there is an example which has some example code: for(index i = 0; i != 3; ++i) for(index j = 0; j != 4; ++j) for(index k = 0; k != 2; ++k) A[i][j][k] = values++; Why can't you change the != into <? Ie, do: for(index i = 0; i < 3; ++i) for(index j = 0; j < 4; ++j) for(index k = 0; k < 2; ++k) A[i][j][k] = values++; or can you? This would be more efficient I should think, and requiring that index types be ordered shouldn't be too onerous should it? Question 2 ---------- If I use the built-in "no-frills" C array type to represent a 2-dimensional array in row-major order, then I would access an element (i, j) by doing: a[ &a[0] + i*n + j ] where a is my m by n array. And if I wish to traverse a, in row-major order, I can do: for (int* ip=&(a[0]); ip<&(a[n*m]); ip+=n) { for (int* ijp=&(ip[0]); ijp<&(ip[n]); ++ijp) { visit(ijp); } } These are a very efficient ways of doing things, but it lacks somewhat in ease-of-generalization and in readbility. This is where the MultiArray library is useful. But do I lose in efficiency by going to MultiArray? Ie, in the simple case where base is 0, stride is 1, and when we traverse in row-major order, is the MultiArray library smart enough to do compile-time optimizations to generate code which is as efficient as the above? I am guessing that this sort of approach is what the Blitz++ library does?? Question 3 ---------- Is it possible to construct a new multi_array/multi_array_ref/const_multi_array_ref from an array_view by doing a deep copy of the view? A view is a nice way of looking at different parts of a MultiArray object in a readable way, but sometimes you might wish to turn this view into a more-efficiently represented object. Is this easy to do in the MultiArray library? Question 4 ---------- Sometimes you might have an object starts off being mostly accessed in row-major order, so you want a representation which works efficiently for this use. But down the track you might wish to change to mostly column-major access. Does MultiArray provide easy ways to transform from one underlying representation to another? Thanks, Mark Phillips.

3 2

RE: [Boost-users] boost::filesystem::exists() & GetfileAttributesA
by Ben Hutchings 24 Mar '04

24 Mar '04

Sebastian Orfino <sorfino(a)lanacion.com.ar> wrote: > I am experiencig a problem with exists() and its Windows > implementation with GetFileAttributes API. I am using exists() prior > to the scan of a directory to see if it exists; the fact is that on > W2K (just tested it on that OS) after a day or two > GetFileAttributesA() returns INVALID_FILE_ATTRIBUTES but the path > does exists (it is a UNC-style path). > After a restart of the app, all works fine. <snip> It couldn't possibly be related to this bug, could it: <http://support.microsoft.com/default.aspx?scid=kb;en-us;325929>? I know that that talks about Windows XP clients and NT 4 servers but MS does not always test thoroughly whether a bug found in one version of a product exists in other versions. The bug is supposed to be fixed in Windows XP SP2, by the way.

1 0

It always show compiler limitation when I use boost.python to wrap a class under MSVisualStudio .Net
by Shunming Fang 24 Mar '04

24 Mar '04

Hello, guys, I try build my c++ lib as python modules using boost.python, when compile one big class, it always show compiler limit( my ram already update to 1.5 GB) while the class can be smoothly build on other platforms like SGI, Linux(RAM 512 MB+). I don't know why? I already upgrade my MSVS .net to 2003Version, Anybody know How I can get around this problems? Thank you very much __________________________________ Do you Yahoo!? Yahoo! Finance Tax Center - File online. File on time. http://taxes.yahoo.com/filing.html

1 0

Boost error on Red Hat Linux 7.4
by Jonathan Arnold 24 Mar '04

24 Mar '04

I'm trying to build the latest (1.31.0) Boost libraries for the first time on a Red Hat Linux 7.4, gcc 2.96 system. It seems to build all the libs and such correctly, but I get the following error trying to build the test programs: gcc-C++-action bin/boost/libs/test/build/libboost_test_exec_monitor.so/gcc/debug/shared-linkable-true/test_tools.o /home/jdarnold/boost_1_31_0/libs/test/build/../src/test_tools.cpp: In method `void boost::test_toolbox::detail::print_log_value<char>::operator() (ostream &, char)': /home/jdarnold/boost_1_31_0/libs/test/build/../src/test_tools.cpp:194: parse error before `__ctype_b' /home/jdarnold/boost_1_31_0/libs/test/build/../src/test_tools.cpp:205: confused by earlier errors, bailing out set -e g++ -c -Wall -ftemplate-depth-100 -DBOOST_TEST_NO_AUTO_LINK=1 -g -O0 -fno-inline -fPIC -I"bin/boost/libs/test/build" -I "/home/jdarnold/boost_1_31_0" -o "bin/boost/libs/test/build/libboost_test_exec_monitor.so/gcc/debug/shared-linkable-true/test_tools.o" "/home/jdarnold/boost_1_31_0/libs/test/build/../src/test_tools.cpp" /usr/bin/objcopy --set-section-flags .debug_str=contents,debug "bin/boost/libs/test/build/libboost_test_exec_monitor.so/gcc/debug/shared-linkable-true/test_tools.o" What is going wrong here? Is my version of F++ too old? Thanks in advance. -- Jonathan Arnold (mailto:jdarnold@buddydog.org) Daemon Dancing in the Dark, a FreeBSD weblog: http://freebsd.amazingdev.com/blog/

1 0

Cannot use BGL & boost::format together under VS2003
by Alexeev Gleb 24 Mar '04

24 Mar '04

Hi all! The following simple sample doesn't compile on VC++ .NET 2003: #include <iostream> #include <boost/format.hpp> // without the following line everything's ok #include <boost/graph/adjacency_list.hpp> int main(){ std::cout << boost::format("%d, %s") % 5 % "some text"; return 0; } -- Best regards, Alexeev Gleb mailto:alexeev_gleb@hotbox.ru

2 2

boost/graph - induced subgraph
by Matthew Galati 24 Mar '04

24 Mar '04

Is it possible to find the minimum spanning tree of an induced subgraph? vector< graph_traits<Graph>::vertex_descriptor > vd; typedef subgraph< Graph > SubGraph; SubGraph & inducedG = subG.create_subgraph(vd.begin(), vd.end()); Here is the example I am looking at: induced verts: 1 2 3 4 5 7 Original Graph: 0 <--> 1 2 5 7 1 <--> 0 3 5 7 2 <--> 0 3 4 5 3 <--> 2 1 4 5 4 <--> 2 3 5 6 5 <--> 0 1 2 3 4 6 7 6 <--> 4 5 7 7 <--> 0 1 5 6 Induced Graph: 1 <--> 3 5 7 2 <--> 3 4 5 3 <--> 2 1 4 5 4 <--> 2 3 5 5 <--> 1 2 3 4 7 7 <--> 1 5 So, the "create_subgraph" seems to be working fine. But, when I run kruskals with: vector<graph_traits < SubGraph >::edge_descriptor> spanning_tree; kruskal_minimum_spanning_tree(inducedG, back_inserter(spanning_tree)); for (vei = spanning_tree.begin(); vei != spanning_tree.end(); ++vei) { pair<int,int> ij = incident(*vei, inducedG); int edge_index = e_index[*vei]; cout << "e_index: " << edge_index << " : " << ij.first << " " << ij.second << endl; I get a spanning tree of just 4 edges (while the subgraph has 6 vertices). Also, how does the edge_index of the subgraph correspond to the original graph? For example, edge (4,1) is not in the original graph, but it shows up in my spanning tree. Spanning Tree: e_index: 3 : 2 0 e_index: 7 : 4 0 e_index: 4 : 3 1 e_index: 8 : 4 1 Thanks, Matthew -- Matthew Galati ISE Lehigh University IBM Service Parts Solutions 610.758.4042 (Office) 610.882.0779 (Home) magh(a)lehigh.edu, magal11(a)us.ibm.com http://sagan.ie.lehigh.edu/mgalati/

4 3

boost-sandbox socket and boost 1.31.0?
by Robert Helgesson 23 Mar '04

23 Mar '04

Howdy, I'm wondering regarding the socket library found in boost-sandbox. Will it be updated to function nicely with the latest Boost release? The problem seems to be that boost::default_iterator_policies and boost::iterator_adaptor no longer work like the socket library expect. Regards -- / Robert Helgesson |-- rycee(a)home.se --\ \-- GPG Key ID: 41E95A5B --| http://www.rycee.cx/ /

3 2

Re: [Boost-users] obtaining vertex descriptor from property_map (newbie)
by Gustavo Sánchez 23 Mar '04

23 Mar '04

I tend to use a map<string, vertex_descriptor> for this. Doug Thanks for the idea, I was just thinking something like this, but I have a problem. I need the definition of the Graph typedef property<edge_name_t, std::string> EdgeProperty; typedef property<vertex_name_t, std::map<string,vertex_property??> > NameProperty; typedef property<vertex_index_t, std::size_t, NameProperty> VertexProperty; typedef adjacency_list<vecS, vecS, undirectedS, VertexProperty, EdgeProperty> UGraph; in order to get the vertex_descriptor, but I need the map<string,vertex_descriptor> for the VertexPorperty so I drop in a circular type dependency. Can you please be more verbose? thanks --------------------------------- Antivirus Filtros antispam 6 MB gratis ¿Todavía no tienes un correo inteligente?

2 1

Solaris and thread_resource_error
by Peter Jones 23 Mar '04

23 Mar '04

On Feb 1 of this year Baptiste Lepilleur posted a status report for Boost 1.31.0 Release candidate 2 on Solaris and AIX. It was mentioned that "boost.thread is now systematically failing with 'thread resource error'". I am experiencing the same thing on Solaris 8, with GCC 3.3.2 (GNU binutils) and Boost 1.31.0. Has anyone looked into this yet? If not, I am willing to track down what line in Boost.Threads is throwing thread_resource_error(). -- Peter Jones [unix;vi;crypto;c++;4wd] (http://pmade.org) PGP [5137 173E 5BFA 8C63 8982 F035 FDBB 1636 7D69 C63F]

2 3

linking to boost::thread
by Jerry 23 Mar '04

23 Mar '04

Hi, I've just built boost1.31.0 using bjam under Linux, apparently successfully. There don't appear to be any static libraries for boost::thread. Is this correct? If so, could someone please explain why I'm getting linker errors under g++ if I don't mention the _dynamic_ thread libs on the g++ command line, and "cannot open shared object file" at runtime if I do. Any pointers appreciated. TIA, Jerry.

2 1

shared_ptr help needed, please!
by Eric Damphousse 23 Mar '04

23 Mar '04

hi all, quick question concerning shared_ptr... not sure if things get deleted (destroyed) properly and its freaking me out. I hope this is not too long Basically, I use the BGL to store my data and that seems to be ok. The way the BGL stuff is defined is as such... Note that spObjectNode is defined as typedef boost::shared_ptr<CObject> spObjectNode // Set vertex properties typedef boost::property<boost::vertex_name_t, std::string> node_name_t; typedef boost::property<graph_node_t, spObjectNode, node_name_t> node_property_t; // Set edge properties typedef boost::property<boost::edge_name_t, std::string> connection_name_t; typedef boost::property<graph_connection_t, spObjectNode, connection_name_t> edge_property_t; // Set graph typedef adjacency_list<listS, vecS, bidirectionalS, node_property_t, edge_property_t> graph_t; // graph_t m_Graph; The function below is what I use to add a node to the BGL... template <typename T> void AddNode(const T& object) { // Create the vertex boost::shared_ptr<T> pNode(new T(object)); // Add the vertex + its name to the graph vertexdescriptor_t newNode = boost::add_vertex(node_property_t(pNode, node_name_t(pNode->GetObjectName())), m_Graph); }; The way I use the AddNode ... for example in a main.cpp: /// blablabla above CSphere sphere("sphere1"); CSceneDB::Instance().AddNode(sphere); return 0; For testing purposes, in the base class and derived classes (see below) I print a message in the destructor of the class, just to make sure things are ok. If I just have in my main the previous call then upon exit I expect and get the following message... CSceneDB cleanup CSphere cleanup CGeometry cleanup CObject cleanup Now for the source of my anxiety... I wrote a function to get back my object from the BGL and it goes like this... inline boost::shared_ptr<CObject> getNodeProperty(const vertexdescriptor_t& vertex) { boost::property_map<graph_t, graph_node_t>::type node_prop_map; node_prop_map = boost::get(graph_node_t(), m_Graph); return node_prop_map[vertex]; } boost::shared_ptr<CObject> CSceneDB::getNodebyName(const std::string& strName) { boost::shared_ptr<CObject> node_prop; boost::graph_traits<graph_t>::vertex_iterator vi, vi_end; boost::tie(vi, vi_end) = boost::vertices(m_Graph); for(; vi != vi_end; ++vi) { node_prop = getNodeProperty(*vi); if(node_prop != 0) { if(node_prop->GetObjectName() == strName) break; } } return node_prop; } template <typename T> boost::shared_ptr<T> getNode(const std::string& strName) { boost::shared_ptr<CObject> abc = getNodebyName(strName); shared_ptr<T> def = boost::dynamic_pointer_cast<T>(abc); return def; } The problem is that if I do something like this... in the main.cpp /// blablabla stuff above CSphere sphere("sphere1"); CSceneDB::Instance().AddNode(sphere); boost::shared_ptr<CSphere> abc = CSceneDB::Instance().getNode<CSphere>("sphere1"); return 0; I only get upon exit... CSceneDB cleanup what happened to CSphere cleanup, CGeometry cleanup and CObject cleanup, are they leaking? Now for a sanity check, if I do in the main.cpp ... i do NOT do the assignment boost::shared_ptr<CSphere> abc = /// blablabla stuff above CSphere sphere("sphere1"); CSceneDB::Instance().AddNode(sphere); CSceneDB::Instance().getNode<CSphere>("sphere1"); return 0; upon exit i get ... CSceneDB cleanup CSphere cleanup CGeometry cleanup CObject cleanup Can someone tell me why this is, is my assignment wrong? Any help would be greatly appreciated Eric Brief description of the classes used ofr my test // Object.h class CObject { protected: CObject(){}; protected: virtual ~CObject(){cout << "CObject cleanup" << endl;}; }; typedef boost::shared_ptr<CObject> spObjectNode; // Geometry.h class CGeometry : public CObject { protected: CGeometry(){}; protected: virtual ~CGeometry(){cout << "CGeometry cleanup" << endl;}; }; // Sphere.h class CSphere : public CGeometry { public: CSphere(){}; ~CSphere(){cout << "CSphere cleanup" << endl;}; };

3 4

dumb question: iterator dynamic_cast adapter?
by Paul Elliott 22 Mar '04

22 Mar '04

I am trying to figure out iterators and I have a dumb question: I have an iterator A_iter, that can traverse in both directions ( A_iter++ and A_iter-- ) and can do input (operator*()). Value_type = type of *A_iter is A; and A is static_cast able to B: A start; B fish = static_cast<B>(start); // OK How do I use adapters to create B_iter, which is the same as A_iter, but whose value_type= typeof *B_iter is B? I am sorry if this is a dumb question. -- Paul Elliott 1(512)837-1096 pelliott(a)io.com PMB 181, 11900 Metric Blvd Suite J http://www.io.com/~pelliott/pme/ Austin TX 78758-3117

2 1

obtaining vertex descriptor from property_map (newbie)
by Gustavo Sánchez 22 Mar '04

22 Mar '04

Dear all, We have a graph which has a property of type string associated with the vertexes. What we want to do is to obtain a vertex descriptor from a string. Could you please tell us how to implement it? Thank you in advance! --------------------------------- Antivirus Filtros antispam 6 MB gratis ¿Todavía no tienes un correo inteligente?

2 1

Controlling selection of runtime libraries
by Bob Kline 22 Mar '04

22 Mar '04

We're using Boost's regex package in a Windows environment for a server built with Microsoft Visual Studio 7.x. We'd like to be able to build our server with debugging symbols for our own code, but still use the release versions of the Microsoft runtime DLLs (because the debugging versions of the Microsoft runtime DLLs aren't redistributable, so we can't legally install them on our production server). The problem is, when we compile our own code for debugging (with -Zi and -D_DEBUG), the Boost headers detect this and do their behind-the-curtain magic to point to a version of the regex package which is linked to the debugging versions of the runtime libraries (msvcr71d and msvcp71d). We've tried using the BOOST_REGEX_NO_LIB trick, specifying the version of the regex library to which we want to link (boost_regex_vc7_mdi.lib) but this doesn't work, because although it compiles fine, the linker chokes complaining about the inability to find a bunch of severely mangled names. What's the *real* trick for using the version of the library we want? Do we need to hack the distributed makefile each time we upgrade regex? (Or worse, shuffle libraries around by renaming them?) Thanks. Bob Kline

3 4

Re: boost-sandbox socket and boost 1.31.0?
by Hugo Duncan 22 Mar '04

22 Mar '04

Robert Helgesson <rycee(a)home.se> wrote: > I'm wondering regarding the socket library found in boost-sandbox. Will > it be updated to function nicely with the latest Boost release? Hi, I have something similar that works with current boost CVS code. I have uploaded this to http://sourceforge.net/projects/giallo/ in an attempt to get this moving again. So far I have low level wrappers for sockets and named pipes. Documentation, tests are lacking. I will be adding the reactor code that I uploaded to the yahoo groups files section in due course. Hugo ps Thanks Jeff for pointing me at this post

1 0

Re: Trouble building libbgl-viz.a
by Vladimir Prus 22 Mar '04

22 Mar '04

Steve Juranich wrote: Hi Steve, sorry that it took quite some time to reply, > My version of graphviz_graph_parser.hpp has the following (lines 55-58): > > #ifndef YYSTYPE > typedef int yystype; > # define YYSTYPE yystype > #endif > > So it looks like the test is different as well as the #define arguments > being switched around. > > When I changed my lines to match your lines, I was able to get the > graphviz_graph_lex.o file, but the build chokes again when it tries to > re-run with -DGRAPHVIZ_DIRECTED=1. Could you modify yystype.h to read: #ifndef BOOST_GRAPH_YYSTYPE_H #define BOOST_GRAPH_YYSTYPE_H union yystype { int i; void* ptr; }; #define YYSTYPE yystype #endif // BOOST_GRAPH_YYSTYPE_H This works for me and it if works for you as well, I'll commit the change. - Volodya

2 2

RE: [Boost-users] regular expression searching for Glib::ustring??
by Ben Hutchings 19 Mar '04

19 Mar '04

Paul Elliott <pelliott(a)io.com> wrote: > On Thu, Mar 18, 2004 at 04:08:59PM -0000, Ben Hutchings wrote: > > Paul Elliott <pelliott(a)io.com> wrote: > > > I am using Gtkmm. I want to do boost regular expression searching > > > on Glib::ustring. <snip> > > > How would you set up boost regex to search if both the regular > > > expression and the string to be searched is a ustring? <snip> > > To create a regex from a ustring I would write something like this: > > > > boost::reg_expression<gunichar, unicode_regex_traits> re; > > In the above, you have unicode_regex_traits, which is necessary > because when I tried: > > boost::basic_regex<gunichar> regular_expression; > > I got the following error message: <snip > So this means that regex_traits<gunichar> does not work, and I > must code my own substitute for regex_traits<gunichar>. > > What are the requirements for such a traits structure? I already told you: > > I am using Boost 1.30, in which the traits are documented in > > <boost/libs/regex/traits_class_ref.htm>. I don't know whether that > > documentation still applies to 1.31. Ben.

1 0

singleton_pool compile error
by ukropina＠sfu.ca 19 Mar '04

19 Mar '04

Hi I'm trying to compile the sample code for singleton pool from http://www.boost.org/libs/pool/doc/interfaces.html just to learn how to use singleton pool. Here is the code. struct MyPoolTag { }; typedef boost::singleton_pool<MyPoolTag, sizeof(int)> my_pool; void func() { for (int i = 0; i < 10000; ++i) { int * const t = my_pool::malloc(); } my_pool::purge_memory(); } but I keep geting this error: Compiling... main.cpp c:\program files\microsoft visual studio\vc98\include\xmemory(59) : error C2248: 'object_creator' : cannot access private struct declared in class 'boost::details::pool::singleton_default<struct boost::singleton_pool<struct MyPoolTag,4,struct boost: :default_user_allocator_new_delete,class boost::details::pool::win32_mutex,32>::pool_type>' e:\lib\boost\boost\pool\detail\singleton.hpp(63) : see declaration of 'object_creator' c:\program files\microsoft visual studio\vc98\include\xmemory(59) : while compiling class-template static data member 'struct boost::details::pool::singleton_default<struct boost::singleton_pool<struct MyPoolTag,4,struct boost::default_user_ allocator_new_delete,class boost::details::pool::win32_mutex,32>::pool_type>::object_creator ol_type>::create_object' Error executing xicl6.exe. I'm using MS Visual C++ 6.0 SP5 on windows XP. What am I doing wrong ? Does anyone have a sample code for singleton pool that works on windows ? Thanks Nenad boost::details::pool::singleton_default<boost::singleton_pool<,4,boost::defa ult_user_allocator_new_delete,boost::details::pool::win32_mutex,32>::po

2 1

regular expression searching for Glib::ustring??
by Paul Elliott 19 Mar '04

19 Mar '04

I am using Gtkmm. I want to do boost regular expression searching on Glib::ustring. http://www.gtkmm.org/gtkmm2/docs/reference/html/classGlib_1_1ustring.html This class represents characters in UTF-8, so each character in the buffer is represented by a varriable number of bytes. But it does have a bidirectional iterator. How would you set up boost regex to search if both the regular expression and the string to be searched is a ustring? Does one need to override any of the types defiened in regex_traits? Thank you. BTW. Quote from:http://www.boost.org/libs/regex/doc/regex_traits.html "Under construction. The current boost.regex traits class design will be migrated to that specified in the regular expression standardization proposal." This is not very useful to someone trying to use the boost_regex library now! -- Paul Elliott 1(512)837-1096 pelliott(a)io.com PMB 181, 11900 Metric Blvd Suite J http://www.io.com/~pelliott/pme/ Austin TX 78758-3117

2 1

RE: [Boost-users] regular expression searching for Glib::ustring??
by Ben Hutchings 18 Mar '04

18 Mar '04

Paul Elliott <pelliott(a)io.com> wrote: > I am using Gtkmm. I want to do boost regular expression searching on > Glib::ustring. > http://www.gtkmm.org/gtkmm2/docs/reference/html/classGlib_1_1ustring.htm l > > This class represents characters in UTF-8, so each character in the > buffer is represented by a varriable number of bytes. But it does > have a bidirectional iterator. > > How would you set up boost regex to search if both the regular > expression and the string to be searched is a ustring? > > Does one need to override any of the types defiened in regex_traits? The character type should be gunichar. > Thank you. > > BTW. > Quote from:http://www.boost.org/libs/regex/doc/regex_traits.html > > "Under construction. > The current boost.regex traits class design will be migrated to > that specified in the regular expression standardization proposal." > > This is not very useful to someone trying to use the boost_regex > library now! I am using Boost 1.30, in which the traits are documented in <boost/libs/regex/traits_class_ref.htm>. I don't know whether that documentation still applies to 1.31. You don't necessarily need custom traits to use custom iterator types. However, you do need to define custom traits to make character classes, case folding and so on work for all Unicode characters. To create a regex from a ustring I would write something like this: boost::reg_expression<gunichar, unicode_regex_traits> re; re.assign(s.begin(), s.end(), flags); When searching in a ustring I would write code like: my_predicate pred; boost::regex_grep(pred, s.begin(), s.end(), re, flags); boost::match_results<Glib::ustring::const_iterator, std::allocator<gunichar> > results; boost::regex_search<Glib::ustring::const_iterator, std::allocator<gunichar>, gunichar, unicode_regex_traits, std::allocator<gunichar> >( s.begin(), s.end(), results, re, flags); Glib::ustring result; boost::regex_merge( std::back_inserter(result), source.begin(), source.end(), re, std::basic_string<gunichar>(repl.begin(), repl.end())), flags); (The above is untested as we're not actually using Gtkmm but it's based on what we're doing with our own UTF-8 and UTF-16 strings.)

2 1

RE: [Boost-users] regular expression searching for Glib::ustring??
by Yarwood, Martin (UK) 18 Mar '04

18 Mar '04

> "Under construction. > The current boost.regex traits class design will be migrated to > that specified in the regular expression standardization proposal." > > This is not very useful to someone trying to use the boost_regex > library now! > I'm sure they'll give you a full refund ******************************************************************** This email and any attachments are confidential to the intended recipient and may also be privileged. If you are not the intended recipient please delete it from your system and notify the sender. You should not copy it or use it for any purpose nor disclose or distribute its contents to any other person. ********************************************************************

1 0

adjacency_matrix BGL problem
by Gustavo Sánchez 18 Mar '04

18 Mar '04

Hi everybody, Is it posible to resize a graph with this representation?. The idea is start with an empty matrix and add vertex based on the new vertices and edges we read. That is, we read a new edge and if their vertex aren't in the graph we add them and then the edge. Is this posible in an adjacency_matrix representation??? Thanks a lot. --------------------------------- Antivirus Filtros antispam 6 MB gratis ¿Todavía no tienes un correo inteligente?

2 1

problem with vertex properties in BGL
by Gustavo Sánchez 17 Mar '04

17 Mar '04

I have to build a graph from data in a file, some information is strings which are related to vertexes. I want to prove if those vertexes already exist in the graph before the insert operation. Next is the source: typedef adjacency_matrix<undirectedS,VertexProperty,EdgeProperty> UGraph; // an undirected graph UGraph g(0); // start from scratch property_map<UGraph, edge_name_t>::type ename = get(edge_name, g); // not sure if this is right (the graph have no elements at this point) property_map<UGraph, vertex_name_t>::type vname = get(vertex_name, g); typedef property_map<UGraph, vertex_name_t>::type vname_type; typedef boost::property_traits<vname_type>::key_type vktype; typedef boost::property_traits<vname_type>::value_type vvtype; vktype indice_nombre; vvtype valor_nombre; for(i=0;i<datgen.NumHilos;i++){ valor_nombre = hiloNormal[i].viaOrigen; /// here we get the string indice_nombre = vname[valor_nombre]; /// here I try to get the index /// I think this is not posible because there isn't [] operator in this way valor_nombre = vname[1]; /// if I do this the compiler is right but that isn't useful for me because the information I have is the string not the index.I can use vertex iterators and search for the string, but if posible i prefer not to do that. thanks a lot. --------------------------------- Antivirus Filtros antispam 6 MB gratis ¿Todavía no tienes un correo inteligente?

2 1

Bug in Boost Graph Library?
by Heiko Bauke 17 Mar '04

17 Mar '04

Dear all, today I was engaged in searching for a bug in my program. After a while I figured out that boost::num_egdes behaves not as I expected. If G is an undirected adjacency matrix graph boost::num_egdes(G) returns the number of edges in the graph G plus the number of edges in G that are no self edges. That means almost all edges are counted twice in undirected adjacency matrix graphs. In my oppinion this behaviour is a bug. At least it makes the BGL-user's live unnecessary difficult. Let's considder the following program: #include <cstdlib> #include <iostream> #include <boost/tuple/tuple.hpp> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/adjacency_matrix.hpp> template <typename Graph> void add_edges(const char * const c) { std::cout << c; Graph G(6); boost::graph_traits<Graph>::vertex_iterator v, vbegin, vend; boost::tie(vbegin, vend)=boost::vertices(G); for (v=vbegin; v!=vend; ++v) { boost::add_edge(*vbegin, *v, G); std::cout << ' ' << boost::num_edges(G); } std::cout << std::endl; } int main() { typedef boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS> adjacency_list; typedef boost::adjacency_matrix<boost::undirectedS> adjacency_matrix; add_edges<adjacency_list>("adjacency list graph :"); add_edges<adjacency_matrix>("adjacency matrix graph :"); return EXIT_SUCCESS; } It prints to stdout: adjacency list graph : 1 2 3 4 5 6 adjacency matrix graph : 1 3 5 7 9 11 So the result of the template function add_edges in the program above depends of the graph container that the function add_edges is applied to. I think this is verry unintuitive and boost::num_edges should be changed, so that alls edges in undirected adjacency matrix graphs are counted only once. Or is there a special reason that boost::num_edges is implemented in this way? with regards, Heiko -- -- Gute Sitten haben für die Gesellschaft mehr Wert als alle -- Berechnungen Newtons. (Friedrich II., der Große, 1712-1786) -- Supercomputing in Magdeburg @ http://tina.nat.uni-magdeburg.de -- Heiko Bauke @ http://www.uni-magdeburg.de/bauke

2 1

RE: [Boost-users] boost::thread problem
by Ben Hutchings 17 Mar '04

17 Mar '04

Pete <petehug(a)paradise.net.nz> wrote: > I'm having a problem with my boost multithreaded application. > My app runs fine as a Windows Service, but as a Unix daemon it > stops responding after some time. I noticed that the fewer > threads I have available, the quicker the application stops > responding: HP-UX(64) after 20mins, RH8(256) after >1hr, Linux > RH9(1024) >5hrs. This led me to assume that perhaps the threads > are not cleaned up properly. That could well be right. Perhaps there is some kind of deadlock at thread exit? > Therefore, I wrote a little test app which ought to demonstrate > the problem. To my surprise, the test application showed an > entirely different problem, but again, only under Unix (it > works as expected under Windows). I'm totally lost what the > reason is and appreciate any help. > > Expected: Actual: > --------------- --------------- > Start running Start running > Start swimming Finish running > Start cycling Start swimming > Finish swimming Finish swimming > Finish cycling Start cycling > Finish running Finish cycling > > In windows I get the actual output but under Linux RH9, I get > the Actual output. Under RH9, the first five lines of the > actual output appear immediately, but the last one takes > approx. 4secs. In other words, only the last threads don't > sleep the number of secs specified. According to the manual page: "sleep() makes the current process sleep until seconds sec- onds have elapsed or a signal arrives which is not ignored." (Note that each POSIX thread is/was implemented as a process in Linux.) If thread creation involves receiving a signal then that would explain the above behaviour. I would expect it under RH8 because that has the old LinuxThreads implementation of pthreads that uses signals heavily, but not in RH9 which has NPTL. Still, you would be better off using boost::thread::sleep() instead.

1 0

boost::thread_group design flaw
by Pete 17 Mar '04

17 Mar '04

I'm pretty convinced that there is a serious design flaw in boost::thread_group or at the very minimum, the documentation should better address the issue. Let's take a look at the thread_group.cpp example provided with boost::thread: #include <boost/thread/thread.hpp> #include <iostream> int count = 0; boost::mutex mutex; void increment_count() { boost::mutex::scoped_lock lock(mutex); std::cout << "count = " << ++count << std::endl; } int main(int argc, char* argv[]) { boost::thread_group threads; for (int i = 0; i < 10; ++i) threads.create_thread(&increment_count); threads.join_all(); } This example works fine until I make the following small change to main() which breaks it: int main(int argc, char* argv[]) { boost::thread_group threads; for (int j = 0; j < 2; ++i) { for (int i = 0; i < 10; ++i) threads.create_thread(&increment_count); threads.join_all(); } } The above code will break the second time threads.join_all() is called. The problem is that all boost::thread objects created in the first iteration survive the join_all() call. Worse still, the 2nd join_all() call sends another join() message to each thread it created but boost::thread::join() does not check the m_joinable flag and simply tries to free the system resources it previously freed again causing a system fault. To correct the problem above, I can change the code as follows: int main(int argc, char* argv[]) { boost::thread* pt[10]; boost::thread_group threads; for (int j = 0; j < 2; ++i) { for (int i = 0; i < 10; ++i) pt[i] = threads.create_thread(&increment_count); threads.join_all(); for (int i = 0; i < 10; ++i) threads.remove_thread(pt[i]); } } While this code no longer causes a GPF, I do not believe that the boost::thread objects created by boost::thread_group::create_thread() actually ever get destroyed. I have set a break point in boost::thread::~thread() but this breakpoint is never reached. If I physically delete each pt[i] after removing it from the thread_group I'm getting an assertion error telling me that the pointer I'm trying to delete is not a valid debug-heap pointer, but this error will most likely disappear if I stop using the debug heap (I haven't tried yet). As a boost::thread_group user, I made the following assumptions: - a boost::thread created through boost::thread_group::create_thread will be removed and cleaned up properly when a) the thread terminates or b) the thread_group is deleted - after returning from a call to join_all() the boost::thread_group has no boost::thread object members - I need not specifically keep track of boost::thread objects created by boost::thread_group and destruction of these objects is encapsulated within boost::thread_group I now find that none of these assumptions is quite correct. Perhaps I good immediate remedy would be to implement a boost::thread_group::clear() method which would throw an error if any of the threads where still in a running state and else would clear its threads collection and destroy the objects? How about thread objects "remembering" their membership in a thread_group so that the thread group can be informed when a thread terminates?

3 3

boost::thread problem
by Pete 17 Mar '04

17 Mar '04

I'm having a problem with my boost multithreaded application. My app runs fine as a Windows Service, but as a Unix daemon it stops responding after some time. I noticed that the fewer threads I have available, the quicker the application stops responding: HP-UX(64) after 20mins, RH8(256) after >1hr, Linux RH9(1024) >5hrs. This led me to assume that perhaps the threads are not cleaned up properly. Therefore, I wrote a little test app which ought to demonstrate the problem. To my surprise, the test application showed an entirely different problem, but again, only under Unix (it works as expected under Windows). I'm totally lost what the reason is and appreciate any help. Expected: Actual: --------------- --------------- Start running Start running Start swimming Finish running Start cycling Start swimming Finish swimming Finish swimming Finish cycling Start cycling Finish running Finish cycling In windows I get the actual output but under Linux RH9, I get the Actual output. Under RH9, the first five lines of the actual output appear immediately, but the last one takes approx. 4secs. In other words, only the last threads don't sleep the number of secs specified. #include <string> #include <list> #include <iostream> #include <unistd.h> #include <boost/thread/thread.hpp> class MyThread { protected: std::string m_strName; int m_iDuration; public: MyThread(const std::string & s, int i) : m_strName(s), m_iDuration(i) {} void operator()() { std::cout << "Start " << m_strName << std::endl; sleep(m_iDuration); std::cout << "Finish " << m_strName << std::endl; } }; int main(int argc, char *argv[]) { std::list<MyThread*> lst; std::list<MyThread*>::iterator itr; MyThread* pT; boost::thread_group tg; lst.push_back(new MyThread("running", 6)); lst.push_back(new MyThread("swimming", 2)); lst.push_back(new MyThread("cycling", 4)); for (itr = lst.begin(); itr != lst.end(); itr++) { pT = *itr; tg.create_thread(boost::ref(*pT)); } tg.join_all(); while (!lst.empty()) { delete lst.front(); lst.pop_front(); } return 1; }

2 1

Trouble building libbgl-viz.a
by Steve Juranich 16 Mar '04

16 Mar '04

Here's my system specs (let me know if I'm leaving something out): SuSE Linux 8.2 gcc v. 3.3 barbet (juranich)$ g++ -v Reading specs from /usr/lib/gcc-lib/i486-suse-linux/3.3/specs Configured with: ../configure --enable-threads=posix --prefix=/usr --with-local-prefix=/usr/local --infodir=/usr/share/info --mandir=/usr/share/man --libdir=/usr/lib --enable-languages=c,c++,f77,objc,java,ada --disable-checking --enable-libgcj --with-gxx-include-dir=/usr/include/g++ --with-slibdir=/lib --with-system-zlib --enable-shared --enable-__cxa_atexit i486-suse-linux Thread model: posix gcc version 3.3 20030226 (prerelease) (SuSE Linux) flex v. 2.5.4 bison v. 1.75 Description of problem: When I cd into $BOOST/libs/graph/src and type 'make', I get the following errors: barbet:/local/src/boost_1_31_0/libs/graph/src # make flex -Pbgl_undir_ -ographviz_graph_lex.cpp graphviz_lex.ll bison -p bgl_undir_ -d -v -o graphviz_graph_parser.cpp graphviz_parser.yy graphviz_parser.yy: warning: 1 shift/reduce conflict g++ -ftemplate-depth-50 -DGRAPHVIZ_DIRECTED=0 -I../../.. -g -c graphviz_graph_lex.cpp graphviz_lex.ll: In function `int bgl_undir_lex(yystype*)': graphviz_lex.ll:85: error: request for member `ptr' in `*lvalp', which is of non-class type `int' graphviz_lex.ll:87: error: request for member `ptr' in `*lvalp', which is of non-class type `int' make: *** [graphviz_graph_lex.o] Error 1 So, line 85 in graphviz_lex.ll looks like this: I've looked at the source with a friend who (unlike myself) actually has some experience with lex and yacc, but he couldn't spot the problem either. I know that the caveat at the top of the file says that the makefile probably won't work on my system, so are there any suggestions on what I could do to make it work? Thanks a lot. -- Stephen W. Juranich Science Applications Intl. Corp. (SAIC) (520) 570-7706

2 4

RE: [Boost-users] VC 7.1 & boost.thread
by Ben Hutchings 16 Mar '04

16 Mar '04

raja <raja(a)genesyslab.com> wrote: > Hi, > I am getting the following error. I see that it is trying to link to > the static library, but I still get this error. It's not trying to link to the static library - you are. > Would appreciate if someone can correct my usage. > Linking to lib file: lib"boost_thread"-vc71-mt-sgd-1_31.lib > > c:\Boost\include\boost-1_31\boost\config\auto_link.hpp(269) : > fatal error > C1189: #error : "Mixing a dll boost library with a static runtime is a > really bad idea..." <snip> In your project's properties, under C/C++ > Code Generation, set the Runtime Library to be "Multi-threaded DLL" in the release configuration and "Multi-threaded Debug DLL" in the debug configuration. If you are not using the Visual Studio IDE then use the /MD and /MDd compiler switches respectively.

1 0

General problem on compiling a boost regex code...
by Benjamin Kaiser [gamigo AG] 16 Mar '04

16 Mar '04

Hello, I'm using the boost library for the first time, and I get a compiler error. :-( I don't know why, because I think the boost library is installed successful (I have tested the SuSE rpm file and the download version from boost.org) I'm using gcc 3.x and boost 1.30 on SuSE Linux 9.0... Header file: ------------------------------------------- #include <string> #include <fstream> #include <syslog.h> #include <boost/regex.hpp> #include <vector> using namespace boost; ------------------------------------------- CPP file: ------------------------------------------- ... regex expression("([0-9]+)(\\-| |$)(.*)"); ... ------------------------------------------- Thats all I do alluded to the boost regex library... if I want to compile the files, I get the following compiling error: ------------------------------------------- #g++ -ansi -pedantic-errors -Wall -o statd ./projekt/statd.cpp compiling C++ using -ansi -pedantic-errors -Wall -o statd /tmp/cc2KYvxs.o(.text+0x1951): In function `Accounting::SaveTraffic()': : undefined reference to `boost::reg_expression<char, boost::regex_traits<char>, std::allocator<char> >::reg_expression[in-charge](char const*, unsigned, std::allocator<char> const&)' /tmp/cc2KYvxs.o(.text+0x1994): In function `Accounting::SaveTraffic()': : undefined reference to `boost::reg_expression<char, boost::regex_traits<char>, std::allocator<char> >::~reg_expression [in-charge]()' collect2: ld returned 1 exit status ------------------------------------------- It would be nice if someone can help me with this problem... I'm new at C++ and don't know how to fix this. :-( If I only include the regex.hpp and do the "using namespace boost" everything is compiling fine. Only if I do the "regex expression(...)" in the CPP File I get the errors... :-( Many thanks for help! If someone will help me via ICQ: 20320644 Mit freundlichen Grüßen / With best regards Benjamin Kaiser Systemprogrammierer gamigo® AG -Deutschland - Butterstr. 13 48431 Rheine Tel. +49 5971-899060 Fax +49 5971-8990611 bkaiser(a)gamigo.de (DH/DSS) PGP-Key auf Anfrage / PGP-Key on request! -------------------------------------------------- Diese E-Mail enthält vertrauliche und / oder rechtlich geschützte Informationen. Wenn Sie nicht der richtige Adressat sind oder diese E-Mail irrtümlich erhalten haben, informieren Sie bitte sofort den Absender und vernichten Sie diese Mail. Das unerlaubte Kopieren sowie die unbefugte Weitergabe dieser Mail ist nicht gestattet. This e-mail may contain confidential and/or privileged information. If you are not the intended recipient (or have received this e-mail in error) please notify the sender immediately and destroy this e-mail. Any unauthorised copying, disclosure or distribution of the material in this e-mail is strictly forbidden.

1 0

Re: [Boost-users] regex with msvc-stlport
by Juan Casanueva 16 Mar '04

16 Mar '04

Hi John, I removed BOOST_REGEX_DECL from re_custom_error_messages, re_ten_w, and re_zero_w and It now works perfectly thanks! Juan >Date: Mon, 15 Mar 2004 13:49:51 -0000 >From: "John Maddock" <john(a)johnmaddock.co.uk> >Subject: Re: [Boost-users] regex with msvc-stlport1 >To: <boost-users(a)lists.boost.org> >Message-ID: <03ff01c40a94$68740cd0$ce810252@fuji> >Content-Type: text/plain; charset="iso-8859-1" > > > I have installed stlport 4.6 and tryed to compile regex but I get the >same > > error. > > > > it creates the .lib fine, but then give some unresolved externals when > > creating the dll.. > > what are your setting in regex\user.hpp??? > >Nothing don't change it. > > > if anyone has any idea, I would apreciate it > > thanks > >The only thing I notice is that the errors are only coming from the release >dll build (and not the debug version), is that correct? > >If so it explains why I can't reproduce it, as I only have the standard >edition of vc6 available to me (which doesn't create optimised builds). > >I also notice that the symbols that the linker is complaining about, >probably don't need to be exported anyway, so can you go into both >c_regex_traits.cpp and w32_regex_traits.cpp and remove the BOOST_REGEX_DECL >specifier from the following identifiers: > >re_custom_error_messages >re_ten_w >re_zero_w > >and see if that fixes the issue? If it does be sure to let me know so that >I can patch up the source, > >Thanks, > >John. > > > >------------------------------ > >_______________________________________________ >Boost-users mailing list >Boost-users(a)lists.boost.org >http://lists.boost.org/mailman/listinfo.cgi/boost-users > >End of Boost-users Digest, Vol 144, Issue 2 >******************************************* _________________________________________________________________ Add photos to your messages with MSN 8. Get 2 months FREE*. http://join.msn.com/?page=features/featuredemail

2 1

random library fibonacci bug
by Martin Wartens 16 Mar '04

16 Mar '04

There seems to be a bug in the fibonacci random generators. Only the smaller lagged_fibonacci generators work. Code Sample (tested with VC7.1) follows. The debugger says: Run-Time Check Failure #0 - The value of ESP was not properly saved across a function call. This is usually a result of calling a function declared with one calling convention with a function pointer declared with a different calling convention. //lagged_fibonacci4423 and smaller works, //>=9689 fails typedef boost::lagged_fibonacci9689 rng_type; rng_type rng; rng.seed((unsigned)time( NULL )); boost::uniform_int<size_t> uni(0,len-1); //crashes on this line. boost::variate_generator<rng_type, boost::uniform_int<size_t> > die(rng, uni);

1 0

RE: [Boost-users] boost::thread and std::system
by Ben Hutchings 16 Mar '04

16 Mar '04

Jerry Swan <jeremiah.s(a)ntlworld.com> wrote: > I'd like to be able to issue a call to std::system that will time-out > after a specified period. Is it possible to achieve this using the > boost::threads library? You can stop waiting for it to complete, but you can't terminate the external program. > If so could someone kindly sketch how this might be achieved? You need to have the calling thread create a synchronised flag (bool with condition, protected by mutex) and a child thread that calls system() and then sets the flag when that returns. The original thread then performs a timed wait for the flag to be set. Here's my solution. The synchronised flag can be encapsulated as an "event" like this: struct one_time_event { public: one_time_event() : flag(false) { } void trigger() { boost::mutex::scoped_lock lock(mutex); flag = true; cond.notify_all(); } void wait() { boost::mutex::scoped_lock lock(mutex); while (!flag) cond.wait(lock); } bool timed_wait(boost::xtime & time) { boost::mutex::scoped_lock lock(mutex); while (!flag) if (!cond.timed_wait(lock, time)) return false; return true; } private: boost::mutex mutex; boost::condition cond; bool flag; }; (This is a basic event that you use once and then throw away. More complex event classes are possible but harder to use correctly.) Using this class it's really very easy: void call_system(const char * command, one_time_event & event) { std::system(command); event.trigger(); } bool system_with_time_limit(const char * command, boost::xtime & time) { one_time_event event; boost::thread thread( boost::bind(call_system, command, boost::ref(event))); return event.timed_wait(time); } Ben.

2 1

RE: [Boost-users] Re: Using shared_ptr across shared library boun dary
by Ferdinand Prantl 16 Mar '04

16 Mar '04

> From: David Abrahams [mailto:dave@boost-consulting.com] > > "Jim.Hyslop" <Jim.Hyslop(a)Leitch.com> writes: > > It's been my experience > > (with Visual Studio, at least) that allocating memory in > > one DLL and > > deallocating it in another leads to problems. Sometimes it > > doesn't go > > noticed, but when it does, it's really noticeable. > > The library cleverly avoids that, provided the first > shared_ptr to take ownership of the memory does so in the > same DLL where the object was allocated. Actually, I came into the same problem as Jim - I had to ensure the same version of the msvcrt and moreover its debug/release versions to get the code working well. All classes in checked_delete.hpp, smart_count.hpp and shared_ptr.hpp are either templates or completely in headers. The operations with new/delete and allocators are translated/inlined by the compiler into every unit, IMHO. I could not find the trick in the implementation, how to force the deletion of the pointer by the same runtime, if new() was performed in one DLL and delete() in another one. Could you, please, point me to the right direction to the explatation? Thanks, Ferda

4 3

Newbie Q: Boost Graph stuff
by Ori 16 Mar '04

16 Mar '04

Hi everyone, I have a question concerning the Boost Graph library, specifically with strongly connected components. My problem is this: I have a graph that I need to collapse into strongly connected components, then topologically sort the resulting DAG of components. Then I want to traverse the sorted DAG (in order), and for each component (now just a vertex I guess), look up into the original graph to find all the vertices that belong to this component. The application here is a rigid body simulator, and I'm constructing a graph of contact points between bodies. The idea is to find an ordering to apply collision impulses, which logically means from the ground up. Vertices are thus bodies, and edges denote a contact (collision) case. I've constructed the original graph, and using the example online, performed the strong_components algorithm on it. However, I'm unsure of how to correctly create a graph of the components. I presume using the root_map is somehow involved, but I'm kind of at a loss. Here's what I got: // typedefs, most stolen from graph introduction at boost.org typedef std::pair<int, int> Edge; struct pen_index_t { typedef edge_property_tag kind; }; typedef property<pen_index_t, int> PenProperty; typedef adjacency_list<vecS, vecS, directedS, no_property, PenProperty> Graph; typedef adjacency_list<setS, vecS, directedS, no_property, PenProperty> ComponentGraph; typedef graph_traits<Graph>::vertex_descriptor Vertex; ... // actual code Graph contactGraph(NUM_BODIES); ... various junk to fill up the graph ... // now, collapse strongly connected components vector<int> component(num_vertices(contactGraph)); vector<Vertex> root(num_vertices(contactGraph)); bgl_named_params< Vertex*, vertex_root_t, no_property> rootMap = root_map(&root[0]); int num = strong_components(contactGraph, &component[0], rootMap); // create a new graph with each component as a vertex... // ...how? I tried variations on get(...) on the rootMap and such, but was met with the usual nightmarish stl/boost compiler errors when you mess up something involving templates : ) Any help someone out there can give is really appreciated. Thanks! ~ Ori

2 2

BUG: boost::lexical_cast ignores virtual functions.
by John F Meinel Jr 16 Mar '04

16 Mar '04

I found that when I call boost::lexical_cast<std::string>() with an object that has virtual functions, it fails to dereference the virtual functions. It also cannot use an Abstract base type. Attached is example code. I think the problem is simply that the file boost/lexical_cast.hpp, on line 185 has: Target lexical_cast(Source arg) This should be Target lexical_cast(Source& arg) At least, this fixes all of my problems. John =:-> #include <iostream> #include <string> #include <boost/lexical_cast.hpp> struct A { virtual int f() const = 0; }; struct B : A { virtual int f() const { return 1; } }; struct C : B { virtual int f() const { return 2; } }; std::ostream& operator<<(std::ostream& os, const A& a) { return os << a.f(); } int main() { C c; std::string s; std::cout << c << std::endl; //This should print 2 s = boost::lexical_cast<std::string>(c); std::cout << s << std::endl; //This should print 2 s = boost::lexical_cast<std::string, B>(c); std::cout << s << std::endl; //This returns 1, not 2 //This fails because it tries to instantiate an object of type A. Why??? s = boost::lexical_cast<std::string, A>(c); std::cout << s << std::endl; }

1 0