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

[iostreams] Assertion in filtering_stream dtor when using a code_ converter
by EMalenfant＠interstarinc.com 27 Apr '06

27 Apr '06

Hi, The following causes an assertion failure (in iostreams::detail::optional::operator*()) when destroying the filtering_stream: #include <boost/iostreams/filtering_stream.hpp> #include <boost/iostreams/code_converter.hpp> #include <boost/iostreams/device/file_descriptor.hpp> #include <windows.h> namespace io = boost::iostreams; int main (void) { io::wfiltering_stream<io::output> out; HANDLE h = CreateFileW(L"test.txt", GENERIC_WRITE | GENERIC_READ, FILE_SHARE_READ, NULL, OPEN_ALWAYS, 0, NULL ); out.push( io::code_converter<io::file_descriptor_sink>( io::file_descriptor_sink(h, true) ) ); out << L"Hello world!"; return 0; } If I add "out.pop()" before returning, there is no assertion. Looking deeper, it turns out that iostreams::chain's dtor calls close(); and reset();, and both end up calling close() on the code_converter. And code_converter is not "multi-closable". IOW, I get the same assertion failure with: HANDLE h = CreateFile(/*...*/); io::code_converter<io::file_descriptor_sink> cvt(io::file_descriptor_sink(h, true)); cvt.close(); cvt.close(); //<< Asserts here To fix this, I made it possible to call code_converter::close() multiple times by simply adding: if (!(this->is_open())){ return; } right at the beginning of code_converter::close(). Am I right in assuming that it should be possible to close() a Device or Filter multiple times ? Anyone has comments on this proposed fix? ----------------------------- Éric Quidquid latine dictum sit, altum viditur.

1 0

Multi Array problem
by Robert Ramey 27 Apr '06

27 Apr '06

I have a problem what is demonstrated in the attached file. Basically I have the following create multi_array - a1 create a view on the multiarray which has the same number of dimensions a1_view a1 == a1_view fails to compile. The attached example illustrates this. In the example I verify that both a1 and a1_view fullfill the requirements for Multi Array. The documentation states that in order for a type to be a model of the Multi Array concept. The operations a == b must be valid. Actually the documentation could be considered ambiguous on this point. But that is was I get from a plain reading of the document and from a common sense interpretation. If two types meeting the requirements of the Multi Array concept have the same number of dimensions, I expect to be able to compare them as this is natural operation to perform and implement. Robert Ramey begin 666 test_misc.cpp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` ` end

3 3

shared_ptr and RTTI?
by Foster Brereton 27 Apr '06

27 Apr '06

Hi all, I have this request from a colleague. Can anyone help me shed some light on his issue?: > I did not use the boost::shared_ptr in quite some time and I was > surprised to find out that, more recently, the shared_ptr _requires_ > RTTI to be enabled. I did some reading on the websites and groups and > while I admit that the new version is more flexible, powerful, etc, I > would have thought that they also left a switch in for the old behavior > which did not require RTTI. In some projects, RTTI does not add any > benefits and people obviously turn it off. > > Is anybody aware of such a "switch" that will allow somebody to use > boost::shared_ptr without RTTI? Blessings, Foster -- Foster T. Brereton - Computer Scientist Software Technology Lab, Adobe Systems Incorporated fbrereto(a)adobe.com -- http://opensource.adobe.com

2 1

Using boost::lambda::bind with std::transform
by Kobi Cohen-Arazi 27 Apr '06

27 Apr '06

Hi, Would like to do the following with boost::lambda::bind #include <boost/lambda/lambda.hpp> #include <boost/lambda/bind.hpp> #include <algorithm> using namespace boost::lambda; struct Foo{ std::string get(){ ... } }; std::vector<Foo> array; // say it is already populated std::vector<std::string> tr; // the transformed instead using my handcrafted for loop, I would like to transform the std::list<Foo>::iterator endIter=array.end (); for (std::list<Foo>::iterator iter=array.begin(); iter!=endIter ;++iter) { noExtResults.push_back (iter.get().replace (0,5, "")); } I know it should be something like: std::transform (array.begin(), array.end (), std::back_inserter(tr), bind(&Foo::get, _1)); So here I'm stuck. The bind function will know to extract the string using get. But how to I get that string and do something like: bind(&std::string::replace (???)) Thanks in advance for your kind help, Kobi.

2 12

[algorithm - strings] problems with borland c++ 6
by Federico Feller 27 Apr '06

27 Apr '06

Hi, Im having problems using the string algorithm of boost with Borland C++ 6.0. Precisely, I cant compile the sequence.cpp because of the following error: [C++ Error] sequence.hpp(186): E2285 Could not find a match for 'boost::algorithm::detail::replace<InputT,ForwardIteratorT>(StringE,char *,char *,undefined,undefined)' Iam using std::string for string algorithms. Any ideas? Thanks, Federico __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

[checked_delete] pimpl with scoped_ptr<T>
by Herve Alexanian 27 Apr '06

27 Apr '06

Hello, This is my first post regarding boost, I hope I am sending it to the right list. I have a question regarding the use of scoped_ptr versus shared_ptr: I am trying to use scoped_ptr<Impl> to implement the "Pimpl" idiom in order to hide the Impl definition from the user. Since my class is not copyable I thought using a scoped_ptr was better suited than a shared_ptr. So, something like this: class Impl; class MyClass { scoped_ptr<Impl> m_pImpl; ... ... }; When compiling client code, I see a checked_delete error. Reading the blurb about checked_delete this makes sense to me and I am able to work around the error by putting an explicit destructor declaration in MyClass. But, I don't see any error if I use a smart_ptr instead of a scoped_ptr. Why is the requirement about not deleting an incomplete type any different between scoped_ptr and shared_ptr? Thanks, Herve Alexanian Sonics, inc.

1 0

[mpl] no result_type defined
by Frank Birbacher 27 Apr '06

27 Apr '06

Hi! I want to create a container of pointers to objects in a range (first,last(. I tried to use mpl, but have been without success. I understand, that the lambda "&_1" cannot specify its concrete return type, so I added ret<T>(...). But the resulting lambda is rejected by the make_transform_iterator function. I'm using boost 1.33.0 on gcc 4.0.2 //****** CODE BEGIN ****** #include <vector> #include <iterator> #include <boost/iterator/transform_iterator.hpp> #include <boost/lambda/lambda.hpp> template<typename Iterator> void foo(Iterator first, Iterator last) { using boost::make_transform_iterator; using namespace boost::lambda; typedef typename std::iterator_traits<Iterator>::pointer Pointer; std::vector<Pointer> objectPointers( make_transform_iterator(first, ret<Pointer>(&_1)), make_transform_iterator(last, ret<Pointer>(&_1)) ); // ... } int main() { std::vector<int> v; //instantiate template: foo(v.begin(), v.end()); } //***** CODE END ****** // ** ERROR on GCC 4.0.2: ** /mnt/daten/develop/boost/boost/iterator/transform_iterator.hpp:43: Fehler: kein Typ namens »result_type« in »class boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::explicit_return_type_action<int*>, boost::tuples::tuple<boost::lambda::lambda_functor<boost::lambda::lambda_functor_base<boost::lambda::other_action<boost::lambda::addressof_action>, boost::tuples::tuple<boost::lambda::lambda_functor<boost::lambda::placeholder<1> >, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type> > >, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type, boost::tuples::null_type> > >« The error says it cannot find the type "result_type" in "class ....". What is wrong here? Frank

2 2

[BGL] How to build graphviz with MS VC8?
by Rui Carvalho 27 Apr '06

27 Apr '06

Hi, I'm trying to build graphviz so as to be able to use read_graphviz in the boost graph library. My understanding is that the make file in Boost 1_33_1\boost_1_33_1\libs\graph\src is not yet ready for VC -but I may be missing something obvious here. Is there a simple way to compile the graphviz library with VC? Thanks, Rui

2 4

Boost::program_options - Linking Issue
by Adam Hartshorne 27 Apr '06

27 Apr '06

Using the following command lines on a linux box, g++ -c -pipe -Wall -W -O2 -pipe -m64 -fexceptions -D_REENTRANT -DQT_NO_DEBUG -DQT_THREAD_SUPPORT -DQT_SHARED -I/usr/lib64/qt-3.3/mkspecs/default -I. -Idcs/improc2/adam/boost_1_33_1/ -I/usr/lib64/qt-3.3/include -I.moc/ -o .obj/first.o first.cpp g++ -o cmdLineTest .obj/first.o -L/usr/lib64/qt-3.3/lib -L/usr/X11R6/lib64 -L/dcs/improc2/adam/boost_1_33_1/stage/lib/ -lboost_program_options-gcc -lqt-mt -lXext -lX11 -lm -lpthread I get the following error output when trying to compile the first.cpp example provided with the library .obj/first.o(.text+0x57): In function `main': : undefined reference to `boost::program_options::options_description::options_description(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)' .obj/first.o(.gnu.linkonce.t._ZNK5boost15program_options25basic_command_line_parserIcE3runEv+0x5): In function `boost::program_options::basic_command_line_parser<char>::run() const': : undefined reference to `boost::program_options::common_command_line_parser::run() const' .obj/first.o(.gnu.linkonce.t._ZN5boost15program_options25basic_command_line_parserIcEC1EiPPc+0x42): In function `boost::program_options::basic_command_line_parser<char>::basic_command_line_parser(int, char**)': : undefined reference to `boost::program_options::common_command_line_parser::common_command_line_parser(std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > > const&)' And help would be greatly appreciated, Adam

2 1

[1.33.1][mpl] increasing BOOST_MPL_LIMIT_LIST_SIZE doesn't work
by Marc Mutz 27 Apr '06

27 Apr '06

Hi, According to the mpl docs, defining above macro to e.g. 30 should make boost::mpl::list<> take 30 template arguments, and thus enable boost::variant<> to hold 21 different types (default: 20 on my platform (gcc 4.0)). Yet compiling the attached sample program gives me attached error message, which indicates that mpl::list<> still takes only 20 template parameters. Bug? Thanks, Marc -- Marc Mutz -- marc(a)klaralvdalens-datakonsult.se, mutz(a)kde.org Klarälvdalens Datakonsult AB, Platform-independent software solutions

1 0

I/O bug in date_time
by Thomas Lemaire 27 Apr '06

27 Apr '06

Dear list, I encounter a bug while reading a time_duration object, here is a short test case, the results are obtained with boost 1.33.1, gcc4.0.3 using namespace boost::posix_time; time_duration td1(0,0,0); std::stringstream ss1("14:23:11.345678"); ss1 >> td1; std::cout << td1 << std::endl; // 14:23:11.345678 time_duration td2(0,0,0); std::stringstream ss2("18353:23:11.345678"); ss2 >> td2; std::cout << td2 << std::endl; // 18:53:23 time_duration td3(0,0,0); std::stringstream ss3("318353:23:11.345678"); ss3 >> td3; std::cout << td3 << std::endl; // 00:00:00 If I compile with the flag CPPFLAGS += -DUSE_DATE_TIME_PRE_1_33_FACET_IO this is working well. I hope this will help to fix it... -- thomas http://www.laas.fr/~tlemaire

1 0

any and Map
by Joe McCay 27 Apr '06

27 Apr '06

if dt=3.5; is a double, then would dt=3.5F; be a float? and dt=3.5L; would be a long double? I had forgotten about those. People don't talk & wrtie about thosse much. Would something similar happen with int & long? Joe McCay

1 0

bcbboost 0.4 released
by Nicola Musatti 27 Apr '06

27 Apr '06

Hallo, I just released version 0.4 of bcbboost, a set of patches that enable use of the Boost 1.33.1 libraries with Borland's BCB2006. This release is the first to support the bcc32 5.8.2 compiler, which comes with Update2. You can download it from here: http://prdownloads.sourceforge.net/bcbboost/bcbboost-1_33_1-5_8_2-0.4.zip?d… Please post any bug report, suggestion, request for clarification to the project trackers on its SourceForge site: http://sourceforge.net/projects/bcbboost/ Cheers, Nicola Musatti

2 3

debugging link errors
by Dave Compton 27 Apr '06

27 Apr '06

Is there a way to get bjam to display the actual compile/link commands that it uses to build the boost libraries as it builds them? I'm seeing some link failures and it would make it easier to figure out what is going on if I could see the link command. Thanks. - Dave

2 1

any and Map.
by Joe McCay 27 Apr '06

27 Apr '06

I keep getting a bad_any_cast exception when I retrieve data from a map. Basically: class DataType { //protected: public: boost::any value_; DataType(){}; template <class type> DataType(type& value) : value_(value) {}; template <class type> DataType(const type& value) : value_(value) {}; // Provide constant & non constant version of operator = template <class type> type& operator=(type &rhs) { value_=rhs; return rhs; }; template <typename type> const type& operator= (const type& rhs) { value_=rhs; return rhs; }; } map <std::string, DataType> m; DataType dt; dt=3.5; m["Data 1"] = dt; DataType dt = fieldData["Data 1"]; fValue = boost::any_cast<float>(dt.value_); //returns a bad_any_cast What am I doing wrong? Joe McCay

2 1

[program_options] making options mandatory
by Luke Elliott 26 Apr '06

26 Apr '06

Hi Is it possible to make options mandatory in program_options? I see some references to such functionality in various review emails, but I can't find anything in the docs themselves. Somewhat related: I actually want the mandatory options to be positional options. I there's an exception class too_few_positional_options_error but it doesn't appear to be used at all (too_many_positional_options_error is). The solution I have at the moment is to have a separate options_description for mandatory options and then to check for count() != 0 for those options after parsing... Thanks in advance Luke Elliott.

1 0

[BGL] push_relabel_max_flow
by André Loose 26 Apr '06

26 Apr '06

Hello, I tried to rewrite the example for push relable max flow on http://www.boost.org/libs/graph/doc/push_relabel_max_flow.html, it currently looks like this: #include <boost/config.hpp> #include <iostream> #include <string> #include <utility> #include <boost/graph/push_relabel_max_flow.hpp> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/graph_utility.hpp> #include <boost/property_map.hpp> using namespace boost; typedef adjacency_list_traits<vecS, vecS, directedS> Traits; struct Vertex { std::size_t index; boost::default_color_type color; }; struct Edge { long capacity_0K; long capacity; long residual_capacity; Traits::vertex_descriptor edge_reverse; }; int main(int , char* []) { typedef adjacency_list<vecS, vecS, directedS, Edge, Vertex> Graph; Graph g; Traits::vertex_descriptor s, t; long flow; flow=push_relabel_max_flow(g, s, t, get(&Edge::capacity,g), get(&Edge::residual_capacity,g), get(&Edge::edge_reverse,g), get(&Vertex::index, g)); return 0; } It does not compile (the IDE jumps to line 681 in push_relabel_max_flow.hpp), and I do not see from the code examples that are availiable on the internet what exactly should be done differently and why. Sadly bundled properties are seldom used in availiable examples, although the documentation encourages their use. Thank you!

1 0

bjam crashes when trying to compile boost with Visual Studio 2003
by Nyenyec N 26 Apr '06

26 Apr '06

Hi, I'm completely new to boost. I tried to compile version 1.33.1 under Windows using Visual Studio 2003 but bjam crashed. What am I doing wrong? See below what I did (see also the attached screenshot). Thanks, nyenyec c:\local\boost_1_33_1>c:\local\bin\bjam -version 'c:\local\bin\bjam' is not recognized as an internal or external command, operable program or batch file. c:\local\boost_1_33_1>c:\local\bin\bjam.exe -v Boost.Jam Version 3.1.12. OS=NT. Copyright 1993-2002 Christopher Seiwald and Perforce Software, Inc. Copyright 2001 David Turner. Copyright 2001-2004 David Abrahams. Copyright 2002-2005 Rene Rivera. Copyright 2003-2005 Vladimir Prus. c:\local\boost_1_33_1>"c:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\bi n\vcvars32.bat" c:\local\boost_1_33_1>"C:\Program Files\Microsoft Visual Studio .NET 2003\Common 7\Tools\vsvars32.bat" Setting environment for using Microsoft Visual Studio .NET 2003 tools. (If you have another version of Visual Studio or Visual C++ installed and wish to use its tools from the command line, run vcvars32.bat for that version.) c:\local\boost_1_33_1>bjam "-sTOOLs=vc-7_1" install ***CRASH***

4 6

boost::shmem
by Jan Stetka 26 Apr '06

26 Apr '06

Hi, Is there an example of using a shmem::map? I'm getting: C:\boost_1_33_1\boost\shmem\sync\win32\win32_sync_primitives.hpp(155) : error C2668: 'boost::shmem::GetSystemTimeAsFileTime' : ambiguous call to overloaded function C:\boost_1_33_1\boost\shmem\sync\win32\win32_sync_primitives.hpp(105): could be 'void boost::shmem::GetSystemTimeAsFileTime(_FILETIME *)' C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\PlatformSDK\Include\WinBase.h(3678): or 'void GetSystemTimeAsFileTime(LPFILETIME)' [found using argument-dependent lookup] while trying to match the argument list '(_FILETIME *)' C:\boost_1_33_1\boost\shmem\sync\win32\win32_sync_primitives.hpp(215) : error C2668: 'boost::shmem::GetSystemInfo' : ambiguous call to overloaded function C:\boost_1_33_1\boost\shmem\sync\win32\win32_sync_primitives.hpp(117): could be 'void boost::shmem::GetSystemInfo(_SYSTEM_INFO *)' C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\PlatformSDK\Include\WinBase.h(3706): or 'void GetSystemInfo(LPSYSTEM_INFO)' [found using argument-dependent lookup] while trying to match the argument list '(_SYSTEM_INFO *)' when i try to compile boost::shmem::map<long,PHRunningData,boost::shmem::allocator<std::pair<long,PHRunningData>, boost::shmem::named_shared_object::segment_manager> > *_rp; is this an error in the library? I'm using the code from the vault. Thanks in advance, Jan.

2 7

Boost linking errors
by Richard 26 Apr '06

26 Apr '06

Hi, I'm trying to use Signals/Slots in my xcode project and I'm running into the following linking errors on XCode 2.2 multiple definition of (anonymous namespace)::_1 in section (__DATA,__common) I'm getting this for _1 -> _9 and I can't figure out what is going on, anyone have any ideas? I'm using the latest version of boost I got using Darwin Ports on OS X 10.4 (version 1.33.1 I think) Thanks, --Richard

1 0

adjacency_list directed to undirected conversion
by Sean Kelly 26 Apr '06

26 Apr '06

Hi All, I have a directed adjacency_list graph for which I would like to identify connected components (ignoring the directed-ness). After poking around the BGL docs it seems that my best option is to role my own algo or build a edge_list graph from the original graph and run connected_components. Basically I want to identify disjoint subgraphs starting with a directed graph. Any advice is appreciated. thanks Sean

4 3

BigInteger equivalent
by Noel Yap 26 Apr '06

26 Apr '06

Is there anything in boost (or elsewhere) similar to Java's BigInteger? Thanks, Noel

4 4

Re: [Boost-users] BigInteger equivalent
by Drumheller, Michael 26 Apr '06

26 Apr '06

>>> Message: 4 >>> Date: Tue, 25 Apr 2006 19:26:16 -0700 (PDT) >>> From: Cromwell Enage <sponage(a)yahoo.com> >>> Subject: Re: [Boost-users] BigInteger equivalent >>> To: boost-users(a)lists.boost.org >>> Message-ID: <20060426022616.60087.qmail(a)web53912.mail.yahoo.com> >>> Content-Type: text/plain; charset=iso-8859-1 >>> >>> --- Noel Yap wrote: >>> > Is there anything in boost (or elsewhere) similar to Java's >>> > BigInteger? >>> See the GNU Multi-Precision library (GMP) <<http://www.swox.com/gmp/>> I have not used it, so I cannot really speak to it. I am merely aware of its existence and thought I'd point it out. FWIW google("gnu", "gmp") --> 1,050,000 hits. Michael Drumheller

1 0

Re: [Boost-users] Boost-users Digest, Vol 886, Issue 6
by Wu Yinghui, Freddie 26 Apr '06

26 Apr '06

Hi, boost-users-request(a)lists.boost.org wrote: > Thing is, the only iterator interface to multi_array returns a reference > to the nested containers. Thus, to iterate over all sub-elements, I'm > forced to have 7-11 nested for-loops to iterate over the entire thing. > > - Greg Link You might wanna take a look at creating a 1-dimensional view of your multi-dimensional container. It is described in multi_array documentation that you can create subranges/views of a multi_array (hint: reshape() function template), which may be helpful. Cheers, Freddie -- Wu Yinghui, Freddie Research & Development Software Engineer Volume Interactions Pte Ltd 1 Kim Seng Promenade, #12-01 Great World City East Tower Singapore 237994 Tel: +65 62226962 (Ext 216) Fax: +65 62226215 Email: yhwu(a)volumeinteractions.com URL: http://www.volumeinteractions.com Important: This message is intended for the recipient(s) addressed above. It contains privileged and confidential information. If you are not the intended recipient, please notify the sender immediately by replying to this message and then delete it from your system. You must not read, copy, use, or disseminate this communication in any form. Thank you.

1 0

[regex] converting a wildcard expr into a regex
by Christoph 25 Apr '06

25 Apr '06

Hi, I would like to implement matching of SQL-LIKE clauses by Boost regex. I need to convert strings like "HEL_LO%" into something regex("HEL.LO.*"). As the LIKE-string might contain characters like '.', '+* etc these need to be escaped: "HE+LO%" -> regex("HE\+LO.*"). Another example are shell globs where * and ? take the place of % and _. I have not found an obvious and easy way (besides doing the replacing and escaping by hand) of how to do that using Boost.Regex. Is there any support for this problem present in Boost? Thank you and best regards Christoph

2 1

serialization: deserialization issue
by Olivier Destrebecq 25 Apr '06

25 Apr '06

I have a base class and some derived class. Each class declare a serialize method. each derived class begin its serialization method by: inArchive & BOOST_SERIALIZATION_BASE_OBJECT_NVP (FetchableMetadataObject); to declare the inheritance relationship then they serialize their own members: inArchive & BOOST_SERIALIZATION_NVP(xxxx); I want the base class to declare a serialize as XML method that will return a string containing the serialized XML representation std::string baseClass::SerializeAsXML() const { return baseClass::saveToXML(this); } saveToXML is a static method on the baseClass std::string baseClass::saveToXML(baseClass const* inObject) { std::ostringstream returnValue; boost::archive::xml_oarchive archive(returnValue); archive << boost::serialization::make_nvp("root", inObject); return returnValue.str(); } The serialization appear to work as when the object are printed for the unit test, i pass the xml serialized data and it contains the data that i expect it to contain. The problems start when i want to reload the data into objects I declared a constructor on the base class and on the derived class that accept the XML string and then rebuild the object using that string (is that a correct design decision?) baseClass::MetadataObject(std::string const &serializedObject) { baseClass::LoadFromXML(serializedObject, this); } derivedClass::MetadataObject(std::string const &serializedObject) { baseClass::LoadFromXML(serializedObject, this); } LoadFromXML is a static method on the base class void baseClass::LoadFromXML(std::string const &inXMLString, baseClass *inObject) { std::istringstream inputStream (inXMLString); // open the archive assert(inputStream.good()); boost::archive::xml_iarchive inputArchive(inputStream); // restore the schedule from the archive inputArchive >> boost::serialization::make_nvp("root", inObject); } The serialize method is not even called when de-serializing the object. So i assume i missed something:-) Any help appreciated, Olivier Destrebecq olivierd(a)real.com

1 0

serialization auto_ptr std::string
by Olivier Destrebecq 25 Apr '06

25 Apr '06

I'm including serialization in my library and am running into problems. One of my class include an std::auto_ptr <std::string> I'm including #include <boost/serialization/string.hpp> to handle the string serialization i downloaded the serialization example for an auto_ptr and modified it to allow for XML serialization (see end of email). But when i compile get the following error saying: error: 'struct std::basic_string<char, std::char_traits<char>, std::allocator<char> >' has no member named 'serialize' If i replace the std::auto_ptr <std::string> with just std::string it works, so i assume something is wrong with my auto_ptr code Feel free to ask me more question or send me beginners advice as i'm just beginning with BOOST. Thanks auto_ptrSerialization.hpp: #include <list> #include <memory> #include <fstream> #include <boost/config.hpp> // std::autoptr inteface wrong in dinkumware #include <boost/detail/workaround.hpp> #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/xml_oarchive.hpp> #include <boost/detail/workaround.hpp> #include <boost/serialization/split_free.hpp> namespace boost { namespace serialization { ///////////////////////////////////////////////////////////// // implement serialization for auto_ptr<T> // note: this must be added to the boost namespace in order to // be called by the library template<class Archive, class T> inline void save( Archive & ar, const std::auto_ptr<T> &t, const unsigned int file_version ){ // only the raw pointer has to be saved // the ref count is rebuilt automatically on load T *objectPtr = t.get(); ar << BOOST_SERIALIZATION_NVP(objectPtr); } template<class Archive, class T> inline void load( Archive & ar, std::auto_ptr<T> &t, const unsigned int file_version ){ T *objectPtr; ar >> BOOST_SERIALIZATION_NVP(objectPtr); // note that the reset automagically maintains the reference count #if BOOST_WORKAROUND(BOOST_DINKUMWARE_STDLIB, == 1) t.release(); t = std::auto_ptr<T>(objectPtr); #else t.reset(objectPtr); #endif } // split non-intrusive serialization function member into separate // non intrusive save/load member functions template<class Archive, class T> inline void serialize( Archive & ar, std::auto_ptr<T> &t, const unsigned int file_version ){ split_free(ar, t, file_version); } } // namespace serialization } // namespace boost Best, Olivier Destrebecq olivierd(a)real.com

2 2

Re: [Boost-users] BGL: filtered graph question
by Grégoire Dooms 25 Apr '06

25 Apr '06

boost-users-request(a)lists.boost.org wrote: > Date: Tue, 25 Apr 2006 08:54:12 +0200 > From: Sebastian Weber <sebastian.weber(a)physik.tu-darmstadt.de> > Subject: [Boost-users] BGL: filtered graph question > > Hello everybody! > > I use the filtered_graph framework to restrict a BFS visit to a small > region on my graph. BUT, I provide the starting vertex to the BFS > routine by a vertice_descriptor of the underlying/unfiltered graph. Of > course, this vertice descriptor matches the predicate. My code compiles > fine, but I started to ask myself if this actually behaves correct? > > As I could not find a hint within the docs, does someone know? > > A quick look at the code tells that it is perfectly fine. The type of vertex_descriptor is the same as for the underlying graph and the vertex iterator is simply a filter_iterator using the provided predicate. Best, -- Grégoire Dooms

3 4

Boost::program_optiions - Defining The Config File As A Cmdline Parameter
by Adam Hartshorne 25 Apr '06

25 Apr '06

Hi All, I want to be able to do the following using the Boost::program_options library. 1) Parse a set of command line options in a standard way (I know how to achieve this). 2) Have an command line option which is "get other command line options from a config file defined as the argument to this option. The example multiple_sources is very close to the sort of thing I want, however in that example the config file is hard coded. I want the user to interactively set the config file at runtime. If I try simply putting the following lines in the set of if statements my program crashes (I know this is still hard coding the cfg file, but it was a first attempt before I made it more general), if ((vm.count("input-file")) { ifstream ifs("multiple_sources.cfg"); store(parse_config_file(ifs, config_file_options), vm); notify(vm); } Any help would be much appreciated, Adam

3 3

Boost::program_options - Problem Compiling Examples On Linux Box
by Adam Teasdale Hartshorne 25 Apr '06

25 Apr '06

Hi All, I am having some trouble compiling the examples for the program_options library on a linux box. I following is the output of my attempt to try and compile the first.cpp example. I obviously understand that error message is due to the compiler been unable to see the appropriate function in the given libraries. I therefore assume I should be including so other libraries, but I don't know which ones are needed. Interestingly I can get the simple_ls.cpp example to compile and function ok. Any help would be much appreciated, Adam g++ -c -pipe -Wall -W -O2 -pipe -m64 -fexceptions -D_REENTRANT -DQT_NO_DEBUG -DQT_THREAD_SUPPORT -DQT_SHARED -I/usr/lib64/qt-3.3/mkspecs/default -I. -Idcs/improc2/adam/boost_1_33_1/ -I/usr/lib64/qt-3.3/include -I.moc/ -o .obj/first.o first.cpp g++ -o cmdLineTest .obj/first.o -L/usr/lib64/qt-3.3/lib -L/usr/X11R6/lib64 -L/dcs/improc2/adam/boost_1_33_1/stage/lib/ -lboost_filesystem-gcc-d-1_33_1 -lboost_program_options-gcc-1_33_1 -lqt-mt -lXext -lX11 -lm -lpthread .obj/first.o(.text+0x57): In function `main': : undefined reference to `boost::program_options::options_description::options_description(std::basic_string<char, std::char_traits<char>, std::allocator<char> > const&)' .obj/first.o(.gnu.linkonce.t._ZNK5boost15program_options25basic_command_line_parserIcE3runEv+0x5): In function `boost::program_options::basic_command_line_parser<char>::run() const': : undefined reference to `boost::program_options::common_command_line_parser::run() const' .obj/first.o(.gnu.linkonce.t._ZN5boost15program_options25basic_command_line_parserIcEC1EiPPc+0x42): In function `boost::program_options::basic_command_line_parser<char>::basic_command_line_parser(int, char**)': : undefined reference to `boost::program_options::common_command_line_parser::common_command_line_parser(std::vector<std::basic_string<char, std::char_traits<char>, std::allocator<char> >, std::allocator<std::basic_string<char, std::char_traits<char>, std::allocator<char> > > > const&)' collect2: ld returned 1 exit status make: *** [cmdLineTest] Error 1

2 2

program_options: iterate variable_map
by Stephan Stiglmayr 25 Apr '06

25 Apr '06

Hello! I'm looking for a way to "iterate" through a variable_map after the command line options were stored, in order to print out all the options, or all with newly values. I know that you can test if a specific options was read with the "count()" method, but this is cumbersome. Is there a possiblity to do that? Regards, Stephan.

4 8

Re: [Boost-users] Problem with iterator, map, shared_ptr and temp lated class
by James Hughes 25 Apr '06

25 Apr '06

Spot on! Thanks for that. reading my templates book has given me a good idea of why, so hopefully I wont make the same mistake again! James > -----Original Message----- > From: Joaquín Mª López Muñoz [mailto:joaquin@tid.es] > Sent: 25 April 2006 15:17 > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Problem with iterator, map, shared_ptr and > templated class > > > > James Hughes ha escrito: > > > Hello all, > > > > I am a bit stumped by a compile fault.. > > > > This is a fragment of the code > > [...] > > > for (handlerMap::iterator pos = LogMap.begin(); pos > != LogMap.end(); > > I think this has to be > > for (typename handlerMap::iterator pos = > LogMap.begin(); pos != > LogMap.end(); > > HTH, > > Joaquín M López Muñoz > Telefónica, Investigación y Desarrollo > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

Why Linux May Never Be a True Desktop OS
by Andreas Sæbjørnsen 25 Apr '06

25 Apr '06

http://www.technewsworld.com/rsstory/50091.html Why Linux May Never Be a True Desktop OS E-Mail Article<http://www.technewsworld.com/perl/mailit/?id=50091> Print Version <http://www.technewsworld.com/rsstory/50091.html#> [image: Why Linux May Never Be a True Desktop OS] By Rob Enderle TechNewsWorld 04/24/06 5:00 AM PT With Linux, the customer often expects to get the product for free and wants the retail price of Windows deducted from his/her purchase price. There are no funds passed back to the vendor and, because Linux is different, customers tend to place more service calls -- at $85 a call. As a result, the vendor generally ends up losing money. *The Formula for Total Customer Experience* Nothing is more powerful than seeing real users interact with your e-Commerce application, Web site, or software. Use Morae, a single, collaborative software solution, to identify barriers to conversion, while watching real users navigate your site. Free White Paper!<http://www.technewsworld.com/rsstory/50091.html> This week I'm speaking at the Desktop Linux Summit in San Diego on why the PC OEMs don't, and probably never will, fully support Linux on the desktop. This is somewhat of a deja vu for me since a decade ago my team made a similar presentation at IBM <http://www.ibm.com/> (NYSE: IBM) [image: Latest News about IBM]<http://www.technewsworld.com/perl/search.pl?query=IBM&scope=network>on OS/2 and why it wouldn't beat Windows in its time. For the purpose of this piece I'm using the word "Linux" as an open source OS catch-all so I don't have to call out every distribution or variant. Now, as it was then, the core of the argument wasn't the technology, the strength of Microsoft <http://www.microsoft.com/> (Nasdaq: MSFT) [image: Latest News about Microsoft]<http://www.technewsworld.com/perl/search.pl?query=Microsoft&scope=network>, or anything, really, outside IBM's control. The reason was that IBM would simply not do what was necessary to take the market away from Microsoft. When I look at Microsoft these days I recall something former Hewlett-Packard <http://www.hp.com/> (NYSE: HPQ) [image: Latest News about Hewlett-Packard]<http://www.technewsworld.com/perl/search.pl?query=Hewlett-Packard&scope=net…>Chairman Lou Platt once said: "My biggest problem is finding people who will tell me the truth." I've never seen Microsoft as weak, from an execution standpoint, before, but currently the firm appears weak. Taking advantage of that weakness remains elusive for both Linux players and Apple<http://www.apple.com/>(Nasdaq: AAPL) [image: Latest News about Apple]<http://www.technewsworld.com/perl/search.pl?query=Apple&scope=network>, however. Let's take a look at the reasons why it's been difficult for them to move dramatically into the business or consumer desktop computer spaces. Classifieds Find IT Consultants<http://www.ectnews.com/adsys/link/?creative=3338&ENN_target=http://links.in…> Post Your Project for Free. Get Bids from Thousands of Pre-Screened Consultants. ------------------------------ Laptop DC power jacks parts & service<http://www.ectnews.com/adsys/link/?creative=3338&ENN_target=http://links.in…> Your laptop doesn't power up? You don't have to replace the whole motherboard, you can just replace the part. We stock DC power s... ------------------------------ Computer, PDA & iPod Accessories<http://www.ectnews.com/adsys/link/?creative=3338&ENN_target=http://links.in…> Dell, IBM, Sony, Apple, HP, Compaq, Toshiba, Palm, iPod, and more. Factory-direct Prices, Same-day Shipping, 30-day Money-back Gu... Doing What Is Necessary When it comes to competition, typically you have the folks who are willing to do what is necessary to win and the folks who aren't. In my personal experience, I've seen again and again scenarios where a team would lay out a plan for what they deemed was needed in order to be successful in a market and then they'd watch as the executive in charge cherry picked the things he/she wanted to do. As a result, the project failed, often disastrously. People like to put filters over their eyes and ignore important steps they don't want to perform on the road to achieving a goal. All of us have seen examples of when this derails the plan. With OS/2, the last major challenger to Windows, IBM had a list of things it needed to do in order for the product to be successful. This list laid out staffing levels, marketing expense, structural changes to the company and partnerships, and its foundation was one of the most extensive pieces of research that company ever did. However as these recommendations went up the chain of command they were changed or in some cases, ignored. The end result? OS/2 failed. The requirements for OS/2's success included broad third party hardware<http://www.technewsworld.com/rsstory/50091.html#> [image: Powerful Yet Simple: HP ProLiant ML110 G3. Just $688 with the Intel(r) Pentium(r) 4 Processor.] <http://www.technewsworld.com/rsstory/50091.html> and software [image: operating system]<http://www.technewsworld.com/rsstory/50091.html>support, developer support, and Windows 32-bit support. In order to get into the hardware, however, OS/2 couldn't come from IBM. This was probably the single biggest deal breaker. Compaq, HP and others wouldn't support an OS from a competitor and, much like it was with Sun and Java<http://java.sun.com/> [image: Latest News about Java]<http://www.technewsworld.com/perl/search.pl?query=Java&scope=network>and Netscape and Navigator, the biggest weakness that OS/2 had was IBM not doing what the market required of it to be successful. After reading up <http://ask.slashdot.org/article.pl?sid=06/04/18/122246> on a subject that speaks to a likely OEM desktop requirement -- strong vendor supported drivers -- I deeply doubt whether the folks working on Linux will make the difficult tradeoffs needed to assure success. Think of the iPod, for example. In order to achieve success Apple had to support Windows, as distasteful as that must have been for the firm. The OEM Cost of Linux and Windows Windows is free to the OEMs. In fact, not only is it free, but Microsoft<http://www.technewsworld.com/rsstory/50091.html#>, in effect, pays them to take it. Regardless of the cost, Windows is a logical choice, and a straight pass. Dell <http://www.dell.com/> (Nasdaq: DELL) [image: Latest News about Dell]<http://www.technewsworld.com/perl/search.pl?query=Dell&scope=network>pays about $80 for it and typically charges about $80 for it. There is rarely much of a mark up. If Microsoft were to lower its price that lowered price would be reflected in virtually all desktop hardware immediately. Microsoft provides a number of services which include development support, service support, marketing support, technicians, classes, databases and support materials, and it picks up a lot of the service load as well. In addition, it provides marketing co-op dollars, incentives for early adoption of new products, and ensures a somewhat level playing field (which could be good or bad) for the vendors. Linux, on the other hand, is often viewed as "free." Even when it comes with a fee, that fee is also passed through without any mark up. However, the hardware vendor must now provide all of the things that Microsoft has been providing, too. Particularly painful are services, as customers expect a similar experience with open source that they have with Windows and will often call the vendor to help troubleshoot. The vendor takes all of the related costs; there are typically no dependable (on-site) resources, unless the vendor supplies some of its own. With Linux, the customer often expects to get the product for free and wants the retail price of Windows deducted from his/her purchase price. There are no funds passed back to the vendor and, because Linux is different, customers tend to place more service calls -- at $85 a call. As a result, the vendor generally ends up losing money on average with Linux. This is the primary reason vendors don't want to do desktop Linux, unless there is an extremely lucrative services contract tied to it. Given services revenue often flows to the services division, the desktop hardware divisions still do their best to throw their bodies in front of these efforts. Too Many Versions An old rule of the desktop is that support costs go up by the square of the number of different products you have to support. This may actually understate the problem now that much of this support is done in remote sites like India. For each version of a product the support team has to be trained, support scripts have to be created and updated; patches have to be collected, tested and deployed; and line changes have to be tested against every operating system version shipped. This is part of the reason that IBM hardware was not competitive in the early nineties; the extra cost of supporting OS/2 was theirs and theirs alone while IBM software got the largest portion of OS/2 revenue (that division was profitable, though OS/2 wasn't). There is broad diversity among Linux buyers, and Red Hat<http://www.redhat.com/>(Nasdaq: RHAT) [image: Latest News about Red Hat]<http://www.technewsworld.com/perl/search.pl?query=Red%20Hat&scope=network>, which is the closest to a standard Linux OS, isn't yet dominant enough to claim victory. The OEMs might be able to handle one version of Linux but until there is only one they have to handle they will probably avoid doing it. Excessive Advocacy A few years ago I recall getting a note from an Apple guy who said something to the effect of, "I'm sure glad all of the crazy Apple supporters moved over to Linux." While I seriously doubt all of the crazy Apple zealots moved to Linux, I can say that they were instrumental in getting Apple out of large enterprises. Zealots, regardless of cause, have a tendency to treat very poorly those who don't share their beliefs, and, historically, they have been much more likely to misuse e-mail [image: Integrate BlackBerry with your CRM Solution]<http://www.technewsworld.com/rsstory/50091.html>or take a dispute to a public forum. They tend to be human resource nightmares and anyone wanting to create a powerful argument against either platform has only to take the decision maker to any of a number of fan Web sites to create the impression of significant risk. This can be particularly problematic now with employee blogs starting to spread and more internal disagreements being made public. Management typically does not want to create an environment where company dirty laundry is aired or where employees organize to fight executive orders. Yet zealots appear to do this as a matter of course: their beliefs come first, even though this is technology, not religion. With all this in mind, IT executives want to be able to decide to abandon a Linux or Apple trial with faith that the employees would follow those orders and not revolt. They also want to feel in control with regard to how the technology is used and not cede that to an external or internal employee advocacy group without conscious thought. In conclusion, there are a few key requirements for alternative platforms to be successful, success being measured by significant market share, in the general market. A Reliable Road Map to Linux Desktop Success The list of clear requirements includes: - Clear OEM resources to match or exceed those currently provided by Microsoft. This would include R&D support and co-marketing dollars. - Embrace existing desktop requirements (roadmaps, application support, proprietary drivers, consistent patch releases schedules and documentation, clear escalation lines for support). - Seek out reasonable advocates who will take direction from their executives and not from other advocates. - Make it profitable for the OEM. Recall that for Microsoft it took about 5 years, US$10 billion and Y2K to move the majority of the installed PC base to another kernel -- and they had to kill the old one off to do it. In the end, I look at this short list and still doubt it is reasonably achievable. I also wonder if Linux were actually modified to be able to comply with this list whether it would even be Linux anymore. That deserves a little thought. Generally speaking, you have to give something up in order to get something else. However there are times when what you sacrifice is actually more important than what you gain.

1 1

Using bind/lambda to extract graph bundled properties from overloaded operator[]
by Sean Kelly 25 Apr '06

25 Apr '06

Hi All, I'm attempting to use bind to extract bundled properties from a adjacency_list graph via the syntax below. ToyGraph g; ToyGraph::vertex_descriptor v; bind(&ToyGraph::operator[], ref(g), _1)(v); this fails to compile (I assume because opertor[] is overloaded) bind(&ToyGraph::swap, ref(g), _1)(g); works fine -- any suggestions on how to make the former work are appreciated. thanks Sean Windows XP Professional SP1, Boost 1_32_0

1 0

Problem with iterator, map, shared_ptr and templated class
by James Hughes 25 Apr '06

25 Apr '06

Hello all, I am a bit stumped by a compile fault.. This is a fragment of the code template<typename T> void LCLogSource<T>::SetLoggers(std::map<T, boost::shared_ptr<LCLogHandler<T> > > &LogMap) { typedef std::map<T, boost::shared_ptr<LCLogHandler<T> > > handlerMap ; for (handlerMap::iterator pos = LogMap.begin(); pos != LogMap.end(); ++pos) // **** compile error here { yada yada I get a compile fault on the for line, 'expected ; before pos'. Now this fault has only started happening since I decided to template this class. It appears that the combination of the template, and the complexity of the map declaration is causing things to go awry, and I don't know why. I have tried a few alternatives but always get a problem of some description. Is there some reason I cannot use iterators on a stl::map where the map key defn. depends on the template parameter? In case this is nothing to do with the shared_ptr in the map, does anyone have a list where templates programming is discussed so I can post there? James

2 1

General Boost documentation
by John Krajewski 25 Apr '06

25 Apr '06

I'm new to generic programming and finding the various concepts used in Boost quite confusing (named parameters, property maps, etc). Is there general documentation for the programming methods used in Boost? Perhaps a good book to buy?

4 4

Graphs - push/relabel algorithm
by Leandro Melo 25 Apr '06

25 Apr '06

Hi. One of the restrictions of boost's graph push relabel algorithm is the existence of 'reverse' edges in the graph. I don't know exactly what the procedure should be for graphs already have the 'reverse' edges. For example, if I want to compute a max flow of an undirected network (it's not the rare) - in this case the networks would already have the reverse edges, which are not necessary set with capacity 0, as the algorithm requires. From what I can see in the implementation (a very fast look), the algorithm would not work correct in this cases. Maybe I'm 'thinking' wrong. Can anyone help me understand that? Thanks. -- Leandro Terra C. Melo

1 0

BGL: filtered graph question
by Sebastian Weber 25 Apr '06

25 Apr '06

Hello everybody! I use the filtered_graph framework to restrict a BFS visit to a small region on my graph. BUT, I provide the starting vertex to the BFS routine by a vertice_descriptor of the underlying/unfiltered graph. Of course, this vertice descriptor matches the predicate. My code compiles fine, but I started to ask myself if this actually behaves correct? As I could not find a hint within the docs, does someone know? Greetings, Sebastian

4 3

[BGL] Neighbourhood as a subgraph.
by Seweryn Habdank-Wojewódzki 25 Apr '06

25 Apr '06

Hi! Did any one check how to get neighbourhood of the node from given graph with respect to the properties or connections? Like in example below: V = { 1, 2, 3, 4, 5 }; E = { (1,2), (2,5), (5,4), (2,4), (1,4), (1,3), (3,4) }; So if there is no property, let assume that we are interesting in searching neighbourhood level 1 of point 1 we will have {1,2,3,4} level 2 {1,2,3,4,5}, but neighbourhood level 1 of point 4 is {1,2,3,4,5}. Of course I am interesting in copying proper edges too. If we have numerical properties on edges we could define (as on all problems with paths) that distance between node A and B is a sum of properties between them. And if we set adjacent matrix as 0 1 3 2 0 1 0 0 1 1 3 0 0 2 0 2 1 2 0 2 0 1 0 2 0 Than neighbourhood of level (or in that case "distance") 1 of point 1 is {1,2} level 2 {1,2,4,5}, and for point 4 level 1 {2,4}. The question is. If algorithms "breadth_first_*" will give me all functionality for that or not? Regards. -- |\/\/| Seweryn Habdank-Wojewódzki `\/\/

2 2

proposal for a clearly visible roadmap on boost.org
by miele di vespa 25 Apr '06

25 Apr '06

Hi all, I am a programmer interested in using boost in my application. I was studying the statechart library that, how I learnt from http://www.boost.org/more/formal_review_schedule.html, is part of boost CVS but not yet part of an official boost release. I was wandering in boost.org site looking for sort of a roadmap telling me the status of each new library, for example: statechart;Accepted -- Added to CVS;will be in 1.33.4;august 2006 I think that this kind of roadmap will be very useful for end-user evaluating the adoption of boost library and it should be in a very noticeable place in the boost.org web site. Thank you for your attention. Best regards, Alessandro Gentilini

2 2

Macro definitions in the wave
by Anatoli Sakhnik 25 Apr '06

25 Apr '06

Hello, boost users! This question is adressed rather to wave developers. Why doesn't the wave expose possibility to query or iterate over the map of macro definitions. This would be very useful for manipulation on the context. Say, one could serialize macro definitions to the disk, then there would it be possible to continue preprocessign the file without walking over unchanged included headers. Currently, I myself collect macro definitions through the ContextPolicy, but this doubles memory consumption. Thank you! -- Anatoli Sakhnik.

1 0

Re: [Boost-users] General Boost documentation
by James Hughes 25 Apr '06

25 Apr '06

> James Hughes wrote: > > >Slightly off topic, but what I would find a great help with > regards to > >learning Boost is a better description on what each library > can do, and more > >importantly, what problems it solves. > > > >For example, perhaps someone has a problem they need to > solve that a Boost > >lib will help with, but just reading through the top level of the > >documentation may not give them enough of a hint that this > is the case. > > > > > This issue has recently been the subject of discussion on the > developer > list. I can't remember exactly what the outcome was, though. > If you have > any concrete suggestions on how to make the library overview without > making it unreadable long, we'd love to hear them. Even a simple list of common problems solved (or if inapplicable, the area of computer theory covered by the component) at the top level of the docs would help I think. Some (most) of the short descriptions are already pretty good (I think that now, but when I first looked they were rather bewildering), but for the more complicated libs, maybe more information would be useful. For example, Property map is described thus "Concepts defining interfaces which map key objects to value objects". At first glance this seems to be pretty similar to stl::map (to someone new perhaps to both Boost and STL), so perhaps needs little more explanation at the top level as to what it's for. > > >My pet one is simple loading and saving of configuration > information in XML. > >Which library should/can I use for that? Serialisation? Or > is there a better > >option? Or is there a boost lib in the wings perhaps aimed at that > >particular task? > > > > > The property_tree library currently under review can do exactly that. > The Serialization library can load and save XML, but only in its own > particular format, i.e. you cannot make it load a schema you have > externally defined. > > >Another is Boost::graph - I see lots of questions about it > on the list, but > >what real world problems does it solve? > > > > > That's what studying computer science is about: learning how to solve > problems. Having learned graph theory, the simple statement that > Boost.Graph helps with graph problems is worth a lot to me. > There is no > way to explain in a short statement what kinds of problems > the library > solves. Many kinds problems can be mapped to graphs. > Already did that - although it was twenty years ago! Problem with Computer Science degrees I always thought (and that's what I did), was the lack of real world application. A lot of theory, and *some* good stuff on application, but not enough. Graph theory is a good example - I understand the concepts (even from twenty years ago), but have never been given enough examples of real world application to be able to apply it myself (perhaps I have never had a need of course!). James > Sebastian Redl > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

[date_time] Need help with a couple simple problems
by Scott 25 Apr '06

25 Apr '06

I've spent a couple hours looking over the docs for the date-time library but still haven't been able to figure out how to do what I need to do. I'm hoping someone on this list can shed some light on how to solve my problems. I'm creating a logging system for our software and each log is timestamped. I want to store in a file the boost::posix_time::ptime of each log. I realize ptime supports boost::serialization, but we had to pull the boost::serialization code from our server because of really strange errors (DLL exceptions, DLLs getting unloaded while serializing, etc). What's the best way to serialize/deserialize a ptime? The file is in a binary format, so the most compact format is preferred. I'm storing the ptime for each log in UTC, however I would like to convert it to local time when the log is displayed. I saw all kinds of conversions to local time, but all of them had the time zone hard coded. Is there a simple way to just convert UTC to the time zone of the computer? Thanks for any and all help, Scott

2 1

Boost::Test without main?
by Johannes 24 Apr '06

24 Apr '06

In one of our applications we develop modules that are loaded in to another application (AutoCAD). Since we don't have access to the main function and the modules don't work outside of the application I was wondering if there is a way to make Boost::Test work in an ordinary function? I have not been able to find any such information. /Johannes

2 1

fixed point arithmetic
by Raider 24 Apr '06

24 Apr '06

Is fixed point arithmetic available it Boost libraries? Or, may be, suggested for new versions of Boost? Raider

1 0

repost: undefined symbols
by Olivier Destrebecq 24 Apr '06

24 Apr '06

I'm including the serialization library in the library I'm building. It compiles without errors but produces a lot of warning (~1600): serialization/base_object.hpp:136:57: warning: "__MWERKS__" is not defined serialization/base_object.hpp:147:60: warning: "__BORLANDC__" is not defined archive/basic_xml_iarchive.hpp:41:41: warning: "BOOST_MSVC" is not defined archive/basic_text_oprimitive.hpp:29:51: warning: "BOOST_DINKUMWARE_STDLIB" is not defined I'm building on a mac (10.4.6, intel), and the header file in which I'm including the serialization stuff looks like this: #include <string> #include <sstream> #include <boost/archive/tmpdir.hpp> #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/xml_oarchive.hpp> Any help appreciated Thanks, Olivier Destrebecq olivierd(a)real.com

1 0

BGL compared to LEDA
by xiao-hui.wu＠exxonmobil.com 24 Apr '06

24 Apr '06

Has anyone compared the performance of Boost Graph Library with LEDA? The following is a link I received that contains some comparisons. I would like to know your experience since I am trying to decide which one is more appropriate for my problems. http://www.quappa.com/leda-outperforms-boost.htm I like the generic interface of BGL, although LEDA has a C# counterpart called GOAL, which uses C# interfaces. I am not sure if the performance of GOAL matches LEDA, but it seems to provide some customization capability. Thanks in advance for your feedbacks and suggestions. Xiao-Hui Wu

2 1

Lambda expressions and VC++6
by Sliwa, Przemyslaw (London) 24 Apr '06

24 Apr '06

Hi, I just wanted to know if Lambda expressions are supported by VC++6.0 SP5 compiler. Thanks for help, Pshemek Kind regards Pshemek Przemyslaw Sliwa Global Foreign Exchange FX Risk Engine Development Merrill Lynch ( +44 (0)20 7995 4182 ) +44 (0)79 6325 2319 * Email: przemyslaw_sliwa(a)ml.com -------------------------------------------------------- If you are not an intended recipient of this e-mail, please notify the sender, delete it and do not read, act upon, print, disclose, copy, retain or redistribute it. Click here for important additional terms relating to this e-mail. http://www.ml.com/email_terms/ --------------------------------------------------------

2 1

Boost regexp (Perl) throws initialization error for [^\0x21-\0xfe]
by Brad Barber 24 Apr '06

24 Apr '06

If my code contains the following regexp, const boost::tregex non_printables(__T("[^\0x21-\0x7E]+")); The program throws the following dialog box on initialization. The stack trace follows. Please forward to someone who cares :-) The following regexp, works as expected const boost::tregex non_printables(__T("[^!-~]+")); --Brad ntdll.dll!_KiFastSystemCallRet@0() ntdll.dll!_ZwWaitForMultipleObjects@20() + 0xc bytes kernel32.dll!_UnhandledExceptionFilter@4() + 0x84b bytes mscoree.dll!_abort() + 0xeb bytes mscoree.dll!terminate() + 0x2a bytes mscoree.dll!__CxxUnhandledExceptionFilter() + 0x39 bytes kernel32.dll!_UnhandledExceptionFilter@4() + 0x149 bytes > msvcr80d.dll!_XcptFilter(unsigned long xcptnum=3765269347, _EXCEPTION_POINTERS * pxcptinfoptrs=0x0007da0c) Line 237 + 0xa bytes C msvcr80d.dll!__CppXcptFilter(unsigned long xcptnum=3765269347, _EXCEPTION_POINTERS * pxcptinfoptrs=0x0007da0c) Line 147 + 0xd bytes C mathdlld.dll!__DllMainCRTStartup(void * hDllHandle=0x01010000, unsigned long dwReason=1, void * lpreserved=0x00000000) Line 517 + 0x17 bytes C msvcr80d.dll!@_EH4_CallFilterFunc@8() + 0x12 bytes Asm mathdlld.dll!_except_handler4(_EXCEPTION_RECORD * ExceptionRecord=0x0007db20, _EXCEPTION_REGISTRATION_RECORD * EstablisherFrame=0x0007e15c, _CONTEXT * ContextRecord=0x0007db40, void * DispatcherContext=0x0007daf4) + 0x22 bytes C ntdll.dll!ExecuteHandler2@20() + 0x26 bytes ntdll.dll!ExecuteHandler@20() + 0x24 bytes ntdll.dll!_KiUserExceptionDispatcher@8() + 0xe bytes kernel32.dll!_RaiseException@16() + 0x52 bytes msvcr80d.dll!_CxxThrowException(void * pExceptionObject=0x0007deb0, const _s__ThrowInfo * pThrowInfo=0x00b98b08) Line 166 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::throw_exception<boost::regex_er ror>(const boost::regex_error & e={...}) Line 40 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::regex_error::raise() Line 74 + 0x9 bytes C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::re_detail::basic_regex_parser<w char_t,boost::regex_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::fail(boost::regex_constants::error_type error_code=error_paren, int position=2) Line 161 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::re_detail::basic_regex_parser<w char_t,boost::regex_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::parse(const wchar_t * p1=0x01735a34, const wchar_t * p2=0x01735a38, unsigned int flags=0) Line 137 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::re_detail::basic_regex_implemen tation<wchar_t,boost::regex_traits<wchar_t,boost::w32_regex_traits<wchar _t> > >::assign(const wchar_t * arg_first=0x01735a34, const wchar_t * arg_last=0x01735a38, unsigned int f=0) Line 96 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::basic_regex<wchar_t,boost::rege x_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::do_assign(const wchar_t * p1=0x01735a34, const wchar_t * p2=0x01735a38, unsigned int f=0) Line 524 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::basic_regex<wchar_t,boost::rege x_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::assign(const wchar_t * p1=0x01735a34, const wchar_t * p2=0x01735a38, unsigned int f=0) Line 255 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::basic_regex<wchar_t,boost::rege x_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::assign(const wchar_t * p=0x01735a34, unsigned int f=0) Line 240 C++ boost_regex-vc80-mt-gd-1_33_1.dll!boost::basic_regex<wchar_t,boost::rege x_traits<wchar_t,boost::w32_regex_traits<wchar_t> > >::basic_regex<wchar_t,boost::regex_traits<wchar_t,boost::w32_regex_trai ts<wchar_t> > >(const wchar_t * p=0x01735a34, unsigned int f=0) Line 209 C++ mathdlld.dll!`dynamic initializer for 'non_printables''() Line 65 + 0x15 bytes C++ msvcr80d.dll!_initterm(void (void)* * pfbegin=0x01701520, void (void)* * pfend=0x01702e5c) Line 855 C mathdlld.dll!_CRT_INIT(void * hDllHandle=0x01010000, unsigned long dwReason=1, void * lpreserved=0x00000000) Line 313 + 0xf bytes C mathdlld.dll!__DllMainCRTStartup(void * hDllHandle=0x01010000, unsigned long dwReason=1, void * lpreserved=0x00000000) Line 489 + 0x11 bytes C mathdlld.dll!_DllMainCRTStartup(void * hDllHandle=0x01010000, unsigned long dwReason=1, void * lpreserved=0x00000000) Line 459 + 0x11 bytes C ntdll.dll!_LdrpCallInitRoutine@16() + 0x14 bytes ntdll.dll!_LdrpRunInitializeRoutines@4() + 0x1c7 bytes ntdll.dll!_LdrpLoadDll@24() - 0x1b1 bytes ntdll.dll!_LdrLoadDll@16() + 0x110 bytes kernel32.dll!_LoadLibraryExW@12() + 0xc8 bytes regsvr32.exe!_wWinMain@16() + 0x741 bytes regsvr32.exe!_wWinMainCRTStartup() + 0x198 bytes kernel32.dll!_BaseProcessStart@4() + 0x23 bytes

3 2

Re: [Boost-users] General Boost documentation
by James Hughes 24 Apr '06

24 Apr '06

Slightly off topic, but what I would find a great help with regards to learning Boost is a better description on what each library can do, and more importantly, what problems it solves. For example, perhaps someone has a problem they need to solve that a Boost lib will help with, but just reading through the top level of the documentation may not give them enough of a hint that this is the case. My pet one is simple loading and saving of configuration information in XML. Which library should/can I use for that? Serialisation? Or is there a better option? Or is there a boost lib in the wings perhaps aimed at that particular task? Another is Boost::graph - I see lots of questions about it on the list, but what real world problems does it solve? James ps. If nothing else of course, shared_ptr is ESSENTIAL!! > -----Original Message----- > From: Lynn Allan [mailto:l_d_allan@adelphia.net] > Sent: 24 April 2006 14:25 > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] General Boost documentation > > > John Krajewski wrote: > > I'm new to generic programming and finding the various concepts used > > in Boost quite confusing (named parameters, property maps, etc). Is > > there general documentation for the programming methods used in > > Boost? Perhaps a good book to buy? > > <alert comment="boost newbie"> > > Welcome to the club of confused, disoriented newbies <g>. > > Several suggestions: > > * Start with a VERY basic library and try to absorb "The Boost > Mentality". > > * Stair-step up to a somewhat more difficult library ... repeat as > necessary > > * At this point, the Karlsson book, "Beyond the C++ standard library" > might be helpful. For my purposes, the book had helpful regex > examples > (but regex is already relatively well documented, at least > compared to > several of the other Boost libraries I've wrestled with. Xpressive is > very well documented, IMHO) > > * Adapt a mindset that learning Boost is critical to your career, so > giving up is not an option. (and when your neurons align "just so", > Boost is kinda fun <g>) > > * Also a humble attitude ... you are "rubbing shoulders" with some > very smart people who are VERY busy, but passionate about Boost > prevailing in the marketplace. > > * Keep notes to share with others to perhaps shorten their learning > curve. > > I've found participants on this list to be VERY helpful and patient > with newbies, but they won't write your code for you. > > </alert> > > > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

3 2

Re: [Boost-users] General Boost documentation
by James Hughes 24 Apr '06

24 Apr '06

> > From: "James Hughes" <JHughes(a)Linx.co.uk> > > My pet one is simple loading and saving of configuration > information in > > XML. > > Which library should/can I use for that? Serialisation? Or > is there a > > better > > option? Or is there a boost lib in the wings perhaps aimed at that > > particular task? > > I'm assuming you are not following the current raging debate > on the boost > developers list, then. Your criterion is 'simple' meaning > simple to use, I > take it.... > Not following the developer list, but have seen the library in the 'up and coming' sectioon, hence the slightly loaded question!! > Currently being reviewed is just such a library, property_tree. > > It solves this particular constrained problem relatively > simply but isn't to > everyone's taste in design and implementation. > > The current debate has an element of 'purist vs pragmatist' > programming. > I'm purely pragmatic myself... > A 'simple' configuration file library *could* have multiple > layers and > components to decouple many different elements of design. In > becoming more > useful and extensible as separate components with clearly > defined roles, it > may perhaps be seen as less easy to use for this one role. > > That said, good motivating examples in documentation *can* > make a seemingly > complex collection of interacting classes seem just as simple > as a dedicated > 'property tree' as currently proposed. > I think that is am important point - an 'example' paints a thousand words as it were. I would add that a 'pragmatic' example is the best sort! > Sit down, relax and read the many current threads on the > developer list or > just go to the boost vault and download property_tree rev 5, > then run for > cover. > > > Another is Boost::graph - I see lots of questions about it > on the list, > > but > > what real world problems does it solve? > > On a more serious note, I completely agree (in general). A few more > motivating examples of problems solved would benefit both > existing and > potential new users. Quite. James

1 0

Re: [Boost-users] [program_options] multitoken
by Sohail Somani 24 Apr '06

24 Apr '06

> -----Original Message----- > Edward Diener > Why not have the parameter as: > > "--first=11,12" > > instead of trying to create more than one parameter with the same > parameter name. > > Personally because of the type of solution I suggested, which is > probably the way most uses of such a multi-value value > parameter would > look like, with possibly a different separator than a comma, I do not > think a parameter library should have to cater to the same parameter > name more than once on the command line. What about parameters like -I for gcc? Utilities can add -I's to the command line. It would be an added headache to synchronize all these tools to eventually come up with a -I arg1,arg2,...

6 6

Re: [Boost-users] General Boost documentation
by paul.floyd 24 Apr '06

24 Apr '06

> Slightly off topic, but what I would find a great help with regards to > learning Boost is a better description on what each library can do, and more > importantly, what problems it solves. The Karlsson book mentioned in the post you replied to contains a short summary of all the Boost libraries, and then goes on to cover 12 of them to greater depth. There's a book each dedicated to MPL and Boost::Graph. A+ Paul Accédez au courrier électronique de La Poste : www.laposte.net ; 3615 LAPOSTENET (0,34 /mn) ; tél : 08 92 68 13 50 (0,34/mn)

1 0

managing "interesting syntax" checks
by Howard Gardner 24 Apr '06

24 Apr '06

/* This code detects the presence of particular syntax in a type. The relevant part is the template class syntax_member_insert: it will ramify into an unbounded set of similar classes, each of which is checking for some particular interesting bit of syntax. I am aware of parts of boost that grapple with this issue very directly: boost::concept_check The concepts to be checked are close kin to the syntax I'm looking for. boost::concept_traits Again, the concepts to be checked are close kin to the syntax I'm looking for. And I'm aware of parts of boost that touch on this issue indirectly: boost::function Contains a really nice mechanism for expressing function signatures. In point of fact, one effort that I made to generalize the implementation of syntax_member_insert yielded a solution with template classes like: template< typename x, typename xRet, typename xDummy = dummy, ... > struct member_function; template< typename x, typename xRet > member_function< x, xRet >; template< typename x, typename xRet, typename xP0 > member_function< x, xRet, xP0 >; These look suspiciously like the boost::functionX< > templates. boost::type_traits The is_* classes are used in one of the ways that I would want to use supports_syntax. boost::mpl Represents a *lot* of other uses for supports_syntax. It seems to me that it would be good to express this unbounded set of interesting syntax checks so that all of those facilities can use it. Does such a mechanism already exist? Is it feasable to create it using existing boost facilities? Is it feasable to create it at all? */ #include<ostream> using namespace std; template< typename xParm > struct syntax_member_insert { template< typename x, x & (x::*)( xParm ) = &x::insert > struct match; }; template< typename x, typename xSyntax > struct supports_syntax { private: typedef char ret_true; struct ret_false{ char c[2]; }; template< typename > static ret_false func( ... ); template< typename xX > static ret_true func( typename xSyntax::template match< xX > * ); public: static const bool value = ( sizeof( func< x >( 0 ) ) == sizeof( ret_true ) ); }; struct passes { passes & insert( char ); }; struct fails{}; int main() { cout << supports_syntax< passes, syntax_member_insert< char > >::value << endl; cout << supports_syntax< fails, syntax_member_insert< char > >::value << endl; }

1 0

[program_options]How to save options to storage
by Ken 24 Apr '06

24 Apr '06

HI, BOOST I am sorry if this is a stupid question. I did not found similar question yet. Sometimes we need to read config from a file and change some values then need to save it. So how can I serialize an option_description to stream? Thanks a lot. Regards Ken

3 4

i18n (and program_options)
by Christoph Duelli 24 Apr '06

24 Apr '06

Hi, I am using spirit and program_options so far. When Boost.Program_options fails to parse some command line options it throws an exception (which is fine). So far, I write (if e is such an exception) e.what() to stderr. An obvious example of the the output I get is 'unknown option ...' I would like to present my users a text in their own language (i.e. german). Does Boost or Boost.Program_options support i18n? If yes, how can I set the language to use? (Or is the intended use to catch the exception, analyze it and produce my own translation?) Best regards, Christoph

2 1

Re: [Boost-users] Fwd: Program Options: string iterator not dereferencable
by Vladimir Prus 24 Apr '06

24 Apr '06

On Monday 24 April 2006 13:13, Olaf van der Spek wrote: [I CC boost-users mailing list, where this discussion originates] > > > There also seem to be 'typos' in some comments. Should I report those > > > to you too? > > > > Yes, please. > > > > > And what about: > > > Is it possible to allow long options without -- prefix? > > > The code doesn't seem to support it, but I think it's handy in cases > > > where positional options aren't used. > > > > Did you try 'allow_long_disguise' style option? > > No, but isn't that with a "-" prefix? I'd like no prefix at all. Hmm, never though this will be requested. Anyway, it's possible with a custom style parser, as the attached example shows. For me, both: extra3 test=10 and extra3 test 10 work. HTH, Volodya

2 1

[program_options] additional parser
by Alexander Kondratyuk 24 Apr '06

24 Apr '06

Have a problem with using additional parser feature: I want to allow user to use double dashes for short options, so user will use "-h" or "--h". But using additional parser the code throws an error about syntax. It seems there is a bug in additional parser wrapper - in case then my option has no parameters defined and declared in description it adds an empty value to the vector of values when additional paser returns pair of strings. Perhaps there is another way to get what I want, please let me know how to do it? Here is an example that illustrates the problem: #define BOOST_WHATEVER_DYN_LINK #define BOOST_LIB_DIAGNOSTIC #include <boost/program_options/variables_map.hpp> #include <boost/program_options/options_description.hpp> #include <boost/program_options/parsers.hpp> #include <boost/program_options/detail/utf8_codecvt_facet.hpp> #include <boost/token_iterator.hpp> #include <boost/shared_ptr.hpp> using namespace boost::program_options; #include <string> #include <vector> using namespace std; pair<string, string> parserDashDashShort(const string& s) { string str = s.substr( 0, s.find('=') ); if( str.length()==3 && str[0]=='-' && str[1]=='-' && str[2]<='z' && str[2]>='a') { string val = s.substr(str.length()); return make_pair( str.substr(1), val ); } return make_pair(string(), string()); } int _tmain(int , _TCHAR* argv[]) { int style = command_line_style::allow_short| command_line_style::allow_dash_for_short | command_line_style::short_allow_adjacent; options_description desc; desc.add_options() ("help,h", "help message") ; char* args[]={"program.exe", "--h"}; int argc = sizeof( args ) / sizeof( args[0] ); parse_command_line<char>( argc, (char**)args, desc, style, parserDashDashShort ); return 0; }

2 1

BGL:"topological_sort" on the output of "strong_components"
by John Krajewski 24 Apr '06

24 Apr '06

For my app (a physics simulator), I need to find the order to process sets of nodes in such a way that no node is processed before any node that it has a directed edge to, and all strongly connected nodes are processed in a set together (this is for determining the order to test stacked objects). So basically, I want to detect all strongly connected components, then create a directed graph from that and perform a topological sort. Then, when I process that sorted list of strongly connected components, I'll need to get the list of original nodes that are in each component. My first inclination to solve this is to just create a second graph from the output of the strongly_connected_components algorithm, and then topological_sort that. But perhaps there is some clever way to avoid that extra expense? This will be my first use of BGL/boost so dumbing-down is encouraged :) Thanks, -John gmail.com account: johnnyk1

3 2

BGL: *** glibc detected *** free(): invalid next size
by Irek Szczesniak 23 Apr '06

23 Apr '06

Vertexes in my graph have the distance and predecessor properties. These properties at each vertex are vectors: they store distances and predecessor of shortest paths to every other node in the graph. Below is my program. I run the program and as input I give the file also listed below, and get this run-time error: *** glibc detected *** free(): invalid next size (fast): 0x085dd110 I would appreciate your advice on what I'm doing wrong. Best, Irek ********************************************************************** #include <string> #include <boost/graph/adjacency_list.hpp> #include <boost/graph/dijkstra_shortest_paths.hpp> #include <boost/graph/iteration_macros.hpp> #include <boost/graph/graphviz.hpp> using namespace std; using namespace boost; typedef adjacency_list_traits<vecS, vecS, undirectedS>::vertex_descriptor Vertex; typedef boost::adjacency_list <vecS, vecS, undirectedS, property<vertex_name_t, string, property<vertex_distance_t, vector<int>, property<vertex_predecessor_t, vector<Vertex> > > >, property<edge_weight_t, int, property<edge_weight2_t, int> > > Graph; int main () { Graph g; dynamic_properties dp; dp.property("node_id", get(vertex_name, g)); dp.property("distance", get(edge_weight, g)); dp.property("lambdas", get(edge_weight2, g)); read_graphviz(cin, g, dp); BGL_FORALL_VERTICES(v, g, Graph) { get(vertex_distance, g, v).resize(num_vertices(g)); get(vertex_predecessor, g, v).resize(num_vertices(g)); dijkstra_shortest_paths (g, v, predecessor_map(&get(vertex_predecessor, g, v)[0]). distance_map(&get(vertex_distance, g, v)[0])); } } ********************************************************************** Graph { a; b; c; d; e; f; a -- b [distance = "100", lambdas = "10"] b -- c [distance = "200", lambdas = "10"] c -- d [distance = "80", lambdas = "10"] b -- e [distance = "40", lambdas = "10"] c -- f [distance = "100", lambdas = "10"] a -- e [distance = "210", lambdas = "10"] e -- f [distance = "340", lambdas = "10"] f -- d [distance = "50", lambdas = "10"] }

1 1

[BGL]adjacency_list_traits does not define edges_size_type
by Louis Lavery 23 Apr '06

23 Apr '06

Hello, The class adjacency_list_traits<EdgeList,VertexList,Directed> provides an alternate method for accessing the types of edge_descriptor and vertex_descriptor, with out this we'd run into a problem with mutually recursive types. I need to know the order in which edges are added to the vertices of my graph. To do that I intend to stamp each edge with an (increasing) integral number as it is added to the graph. It seems to me the stamp's type should be edges_size_type, but that's not available via adjacency_list_traits<EdgeList,VertexList,Directed>. Should it be? Can it be? Thanks, Louis.

3 9

memory problems with BGL
by borech＠tepkom.ru 23 Apr '06

23 Apr '06

Hello, I've been in fight with memory leaks in my application for some days. And now I found that most memory is losing while using adjacency_list<> (boost 1.33.1 + VC++6.0). Basically we've lost about 100Kb after each graph life cycle (created - changed - removing). Moreover, I guess, I found the similar problem on the very simple test: #include "..\boost_1_33_1\boost\graph\adjacency_list.hpp" using namespace boost; typedef adjacency_list<> Graph; void main() { int num_vertex = 10000; Graph* graph = new Graph(); for(int j=0; j<num_vertex; ++j) { add_vertex(graph); } delete graph; } Here I've lost some memory too (about 20 Kb). Maybe I'm doing something wrong? Thanks

1 0

Compatibility problems BDS40 vs boost
by Christian Askeland 23 Apr '06

23 Apr '06

Hi all, I am running boost 1.33.1 with BDS40, the Borland compiler. I upgraded to BDS Update pack 2 yesterday, and the two following files failed to compile: boost/iterator_adaptors.hpp boost/function.hpp Have anyone else seen this? The failures are inside a lot of compiler-ifdefs, so I guess that this might require some rewrite of boost. Regards, Christian

2 2

Is there a way to avoid manually adding each lib dir?
by Andrew Schweitzer 23 Apr '06

23 Apr '06

boost for dummies question here: When I use boost with VC 7.1, if I link to a lib, I have to include the header, get the compiler error, copy the enormous name of the lib onto the clipboard, then search in c:\boost_1_33_1\bin for that lib. Then I select the properties of that file, copy the enormous directory name, then open the visual studio project, and add the enormous directory name to the list of directories in which to look for libs. Repeat for each lib linked to. I tried adding e:\boost_1_33_1\bin\boost\libs, e:\boost_1_33_1\bin\boost, e:\boost_1_33_1\bin e:\boost_1_33_1 as lib directories. That didn't help. Is there a less process with fewer manual steps? Thanks, Andy

4 5

[regex] Library did not build
by Ed Keith 23 Apr '06

23 Apr '06

A while back I installed boost 1.33.1 I followed the instructions to build the libraries with mingw and have been using it successfully. Today, for the first time since installing, I tried to used the regex library. I was surprised to discover that the regex library did not build. date_time, file system, iostreams, program_options, python, serialization, signals, test_exec_monitor, unit_test_framework, thread, & wave all built, but no regex. Has anyone else had this problem? It seems to have built for cygwin (but I have not tested it.) -EdK Ed Keith http://www.freecherrypy.org/edk __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

3 2

Boost.Program_options and optional arguments
by Dale McCoy 23 Apr '06

23 Apr '06

I'm trying to convert from getopt (GNU's getopt_long, to be more specific) to boost::program_options, but have run into trouble converting "D::" and {"data",optional_argument,NULL,'D'} (If you're not familiar with getopt, that means that --data is equivalent to -D, and both take an optional argument, so "-D", "--data", "-Darg", and "--data=arg" are all valid.) I see that the first argument to add_option should be "data,D", but then things get fuzzy. I saw nothing in the documentation concerning optional arguments, either directly, or using additional parsers. What is the appropriate method for implementing such options? I'm using currently using Boost 1.33.1. Dale

1 0

Compatibility problems BDS40 vs boost
by Christian Askeland 22 Apr '06

22 Apr '06

Hi all, I am running boost 1.33.1 with BDS 4.0, the Borland compiler. I upgraded to BDS Update pack 2 (released 2006/04/17) yesterday, and the two following lines/files failed to compile: #include <boost/iterator_adaptors.hpp> #include <boost/function.hpp> The code compiled without problems with Update 1. Have anyone else seen this? I'm no expert on compiler quirks, but I know that there are a few workarounds for the BCC compiler in boost. Maybe something has been changed. Regards, ---------------------------------- Christian Askeland ----------------------------------

2 2

[Random] compiler issue with Visual Studio
by David Walthall 22 Apr '06

22 Apr '06

Hello, I am using Visual Studio 7.1.3008. While using the Random library, I ran into an issue with the seed function of mt19937. From what I can gather, the compiler becomes confused about one of the overloads and tries to call the seed function: template<class Generator> void seed(Generator & gen) I couldn't this issue on the list of known issues for this compiler. Is this known. For now, it appears that the work-around is to cast it as an unsigned long. Below is a small test file that demonstrates this issue. Thank you, David Walthall ////// boost_random_error.cpp ////// #include <boost/random/mersenne_twister.hpp> int main() { boost::mt19937 generator(42u); unsigned int uint = 42u; unsigned long int ulong = 42u; generator.seed(static_cast<unsigned int>(42u)); // ok generator.seed(42u); // ok generator.seed(static_cast<unsigned long>(uint)); // ok generator.seed(static_cast<unsigned int>(ulong)); // ok generator.seed(ulong); // ok // comment out the next lines to get it to compile generator.seed(static_cast<unsigned int>(uint)); // error generator.seed(uint); // error (also an error on g++ 3.4.4) return 0; } ////// end of boost_random_error.cpp //////

2 1

[WAVE] bug: generic lexing error
by Andreas Sæbjørnsen 21 Apr '06

21 Apr '06

Using the wave driver reference implementation on the code found at: http://folk.uio.no/andsebjo/bugInducingCode.C I get the following error: bugInducingCode.C(10157): error: generic lexing error: '\000' in input stream This code compiles with g++ and if any line is cut within the file it does not fail with Wave. The code contains only C++ syntax so it is basically a test of the cpplexer. Is this a know problem? This problem occurs in both my two months old CVS version of boost and the 1.3.1 build. Regards, Andreas Saebjoernsen

2 1

boost::function and free template function returning a dependent class template type
by Howard Gardner 21 Apr '06

21 Apr '06

/* This doesn't work. Is it supposed to? If "broken" returns a "ret< int >" or an "x", then it does seem to work right. In case it's a known compiler issue, I'm using: Comeau C/C++ 4.3.3 (Aug 10 2003 15:39:53) for _MS_WINDOWS_x86_Beta8 */ #include "boost/function.hpp" using namespace boost; #include<ostream> using namespace std; template< typename x > struct ret{}; template< typename x > ret< x > broken( x f ) { cout << f << endl; return ret< x >(); } int main() { // function1< ret< int >, int > func; same problem function< ret< int > ( int ) > func = broken< int >; func( 42 ); }

2 3

BGL: std::map<Edge, int> needs the < operator
by Irek Szczesniak 21 Apr '06

21 Apr '06

I'm trying to have a std::map where the key is an edge of a graph, but I get a compilation error. My graph is: typedef boost::adjacency_list <vecS, vecS, undirectedS, property<vertex_name_t, std::string>, property<edge_weight_t, int, property<edge_weight2_t, int> > > Graph; And an edge: typedef graph_traits<Graph>::edge_descriptor Edge; Then I have a map: std::map<Edge, double> m; But when I try this: m[e] = 0 where e is of type Edge, then I get: /usr/lib/gcc/i386-redhat-linux/3.4.4/../../../../include/c++/3.4.4/bits/stl_function.h:227: error: no match for 'operator<' in '__x < __y' Can I use map with the Edge as the key type? If so, what should I do? Thanks for reading. Best, Irek

4 4

[boost-build]error occured when building boost 1.33.1
by lanser 21 Apr '06

21 Apr '06

sorry, my gmail get something wrong, repeat: os: win2k + sp4 compiler: vc6 + sp6 boost 1.33.1 build with: E:\tmp\BOOST_~1>bjam -sTOOLS=msvc "-sMSVC_ROOT=f:\microsoft visual studio\vc98" --with-python-root=f:\python --prefix=f: \Boost --builddir=f:\Boost\build install > d:\boost-install.log i get some errors listed in boost-install.log, please tell me where's wrong, thanks a lot lanser ps: i compress "boost-install.log" with WinRAR 3.0

2 1

[boost-users][build]where is *.lib or *.dll
by lanser 21 Apr '06

21 Apr '06

hi, i build boost under win2k os using bjam -sTOOLS=msvc --prefix=f:\boost --libdir=f:\boost\lib --includedir=f:\boost\include --builddir=f:\boost\build --with-python-root --with-pydebug install after building, i can only see f:\boost\include , where is f:\boost\lib ?? sincerely --szs

2 1

a question about asynchronous I/O
by 李慧霸 21 Apr '06

21 Apr '06

I'm trying to start network programming with boost::asio library, and I come across a question: if I have issued multiple async_read to the same one socket, will the OS execute these IO command out of order? My platform is WinXPSP2+VC2003. lihuiba 2006-04-21

2 1

Re: [Boost-users] BGL: std::map<Edge, int> needs the < operator
by Ovanes Markarian 21 Apr '06

21 Apr '06

Another way would be to implement a global less than operator overload. You can also implement this operator within a class, so that std::less_than is able to call it. Here are some examples... //gloabl operator overlaod sample class Edge { friend bool operator < (const Edge&, const Edge&); public: //your implementation private: //data members }; inline bool operator < (const Edge& e1, const Edge& e2) { return (e1.prop1 < e2.prop1) && (e1.prop2 < e2.prop2) && ...; } //implementig the comaprison operator in the class // if you don't want this operator to be accessible from other classes, // declare the std::less_than<Edge> as friend and make this operator protected class Edge { friend class std::less_than<Edge>; protected: inline bool operator <(const Edge& e) { return (prop1 < e.prop1) && (prop2 < e.prop2) && ...; } //your class implementation }; I do not say that this implementation is better then the one suggested by Dmitry, I would only like to show up all possiblities. Dmitri's implementation is more flexible in regards of compile time type calculations or if you don't have the source code of Edge. My implementation does not involve additional overhead required to create a function object for comparison and delegate comparison through it. In this case the object is only 1 byte big, but if you add additional fields to it memory usage will grow with every comparison. With Kind Regards, Ovanes Markarian > On Fri, April 21, 2006 10:41, Dmitry Bufistov said: >> I'm using the following for a couple of monthes. Seems to work. >> ////////////////////////////////////////// >> struct edge_less_than >> : public std::binary_function<edge_descriptor_t, edge_descriptor_t, >> bool> >> { >> bool operator()(const edge_descriptor_t& __x, >> const edge_descriptor_t& __y) const >> { >> //return __x.second.get_property() < __y.second.get_property(); >> //return &__x < &__y; >> return __x.get_property() < __y.get_property(); >> } >> }; >> >> You can do something better depending of what you need. Possible this >> will help >> http://lists.boost.org/boost-users/2004/08/7579.php >> Regards, >> --dima >> Thomas Costa wrote: >>> write your own total ordering/comparison function object for Edge >>> objects and declare the map type using this function. >>> >>> On Apr 20, 2006, at 7:55 PM, Irek Szczesniak wrote: >>> >>> >> >> >> > > > With Kind Regards, Ovanes Markarian

3 3

Re: [Boost-users] [Graph] bundled properties and connected_components
by Detlef Mages 21 Apr '06

21 Apr '06

Hi, What you are missing is a color map. One solution is to use bundled properties: #include <boost/config.hpp> #include <vector> #include <map> #include <iostream> // for std::cout #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/connected_components.hpp> using namespace boost; struct Node { int index; //****color property****************************** boost::default_color_type m_algo_color; //****color property****************************** }; struct Boundary { double something; }; typedef boost::adjacency_list< boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS> Graph; int main(int argc, char*argv[]) { Graph g(5); boost::graph_traits<Graph>::vertex_iterator vi, viend; std::vector< int > c( num_vertices(g) ); int num; // manually intialize the vertex index, // because we have listS as vertex list type int n = 0; for (tie(vi, viend) = vertices(g); vi != viend; ++vi, ++n) { g[*vi].index = n; } //****modified call****************************** num = connected_components(g, make_iterator_property_map(c.begin(), get(&Node::index, g)), boost::color_map( get(&Node::m_algo_color, g) )); //****modified call****************************** return 0; } Detlef

1 0

[mpl] Please help with mpl::equal
by rincewind 21 Apr '06

21 Apr '06

Please tell me why the following fragment prints 0 - I thought it should print 1: typedef mpl::vector_c<int, 1, 2, 3> v1; typedef mpl::vector<mpl::int_<1>, mpl::int_<2>, mpl::int_<3> > v2; cout << mpl::equal<v1, v2>::value << "\n";

2 1

[Graph] bundled properties and connected_components
by André Loose 21 Apr '06

21 Apr '06

Hi, I use an undirected graph with bundled properties and want to find out, whether it is connected, or not. Reading the documentation and some examples using property maps, I did not succeed in getting it to work. A minimal example that reproduces the compile errors follows. What am I doing wrong here: #include <boost/config.hpp> #include <vector> #include <iostream> // for std::cout #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/connected_components.hpp> using namespace boost; struct Node { int index; }; struct Boundary { double something; }; typedef boost::adjacency_list< boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS> Graph; int main(int argc, char*argv[]) { Graph g(5); boost::graph_traits<Graph>::vertex_iterator vi, viend; std::vector<int> c(num_vertices(g)); int num; // manually intialize the vertex index, // because we have listS as vertex list type int n = 0; for (tie(vi, viend) = vertices(g); vi != viend; ++vi, ++n) { g[*vi].index = n; } num = connected_components(g, make_iterator_property_map(c.begin(), get(&Node::index, g), c[0])); return 0; } compile log: Compiler: Default compiler Executing g++.exe... g++.exe "Y:\graph\boost\test2.cpp" -o "Y:\graph\boost\test2.exe" -I"C:\Dev-Cpp\lib\gcc\mingw32\3.4.2\include" -I"C:\Dev-Cpp\include\c++\3.4.2\backward" -I"C:\Dev-Cpp\include\c++\3.4.2\mingw32" -I"C:\Dev-Cpp\include\c++\3.4.2" -I"C:\Dev-Cpp\include" -L"C:\Dev-Cpp\lib" C:/Dev-Cpp/include/boost/property_map.hpp: In member function `R boost::iterator_property_map<RandomAccessIterator, IndexMap, T, R>::operator[](typename boost::property_traits<IndexMap>::key_type) const [with RandomAccessIterator = __gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, IndexMap = boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, T = boost::default_color_type, R = boost::default_color_type&]': C:/Dev-Cpp/include/boost/property_map.hpp:319: instantiated from `void boost::put(const boost::put_get_helper<Reference, PropertyMap>&, K, const V&) [with PropertyMap = boost::iterator_property_map<__gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, boost::default_color_type, boost::default_color_type&>, Reference = boost::default_color_type&, K = void*, V = boost::default_color_type]' C:/Dev-Cpp/include/boost/graph/depth_first_search.hpp:197: instantiated from `void boost::depth_first_search(const VertexListGraph&, DFSVisitor, ColorMap, typename boost::graph_traits<G>::vertex_descriptor) [with VertexListGraph = boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, DFSVisitor = boost::detail::components_recorder<boost::iterator_property_map<__gnu_cxx::__normal_iterator<int*, std::vector<int, std::allocator<int> > >, boost::bundle_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, void*, Node, int>, int, int&> >, ColorMap = boost::iterator_property_map<__gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, boost::default_color_type, boost::default_color_type&>]' C:/Dev-Cpp/include/boost/graph/depth_first_search.hpp:246: instantiated from `static void boost::detail::dfs_dispatch<boost::detail::error_property_not_found>::apply(const VertexListGraph&, DFSVisitor, Vertex, const boost::bgl_named_params<P, T, R>&, boost::detail::error_property_not_found) [with VertexListGraph = boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, Vertex = void*, DFSVisitor = boost::detail::components_recorder<boost::iterator_property_map<__gnu_cxx::__normal_iterator<int*, std::vector<int, std::allocator<int> > >, boost::bundle_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, void*, Node, int>, int, int&> >, P = boost::detail::components_recorder<boost::iterator_property_map<__gnu_cxx::__normal_iterator<int*, std::vector<int, std::allocator<int> > >, boost::bundle_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, void*, Node, int>, int, int&> >, T = boost::graph_visitor_t, R = boost::no_property]' C:/Dev-Cpp/include/boost/graph/depth_first_search.hpp:324: instantiated from `void boost::depth_first_search(const VertexListGraph&, const boost::bgl_named_params<P, T, R>&) [with VertexListGraph = boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, P = boost::detail::components_recorder<boost::iterator_property_map<__gnu_cxx::__normal_iterator<int*, std::vector<int, std::allocator<int> > >, boost::bundle_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, void*, Node, int>, int, int&> >, T = boost::graph_visitor_t, R = boost::no_property]' C:/Dev-Cpp/include/boost/graph/connected_components.hpp:93: instantiated from `typename boost::property_traits<IndexMap>::value_type boost::connected_components(const Graph&, ComponentMap) [with Graph = Graph, ComponentMap = boost::iterator_property_map<__gnu_cxx::__normal_iterator<int*, std::vector<int, std::allocator<int> > >, boost::bundle_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, void*, Node, int>, int, int&>]' Y:\graph\boost\test2.cpp:38: instantiated from here C:/Dev-Cpp/include/boost/property_map.hpp:349: error: no match for 'operator+' in '((const boost::iterator_property_map<__gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, boost::default_color_type, boost::default_color_type&>*)this)->boost::iterator_property_map<__gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, boost::default_color_type, boost::default_color_type&>::iter + boost::get [with PropertyMap = boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, Reference = const boost::detail::error_property_not_found&, K = void*](((const boost::put_get_helper<const boost::detail::error_property_not_found&, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t> >&)((const boost::put_get_helper<const boost::detail::error_property_not_found&, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t> >*)(((const boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>*)((const boost::iterator_property_map<__gnu_cxx::__normal_iterator<boost::default_color_type*, std::vector<boost::default_color_type, std::allocator<boost::default_color_type> > >, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::setS, boost::listS, boost::undirectedS, Node, Boundary, boost::setS, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, boost::default_color_type, boost::default_color_type&>*)this)) + 4u))), ((void* const&)((void* const*)(&v))))' C:/Dev-Cpp/include/c++/3.4.2/bits/stl_iterator.h:654: note: candidates are: __gnu_cxx::__normal_iterator<_Iterator, _Container> __gnu_cxx::__normal_iterator<_Iterator, _Container>::operator+(const typename std::iterator_traits<_Iterator>::difference_type&) const [with _Iterator = boost::default_color_type*, _Container = std::vector<boost::default_color_type, std::allocator<boost::default_color_type> >] C:/Dev-Cpp/include/c++/3.4.2/bits/stl_bvector.h:345: note: std::_Bit_const_iterator std::operator+(ptrdiff_t, const std::_Bit_const_iterator&) C:/Dev-Cpp/include/c++/3.4.2/bits/stl_bvector.h:261: note: std::_Bit_iterator std::operator+(ptrdiff_t, const std::_Bit_iterator&) Execution terminated

2 2

Newbie: undefined symbol
by Olivier Destrebecq 20 Apr '06

20 Apr '06

I'm including the serialization library in the library I'm building. It compiles without errors but produces a lot of warning: serialization/base_object.hpp:136:57: warning: "__MWERKS__" is not defined serialization/base_object.hpp:147:60: warning: "__BORLANDC__" is not defined archive/basic_xml_iarchive.hpp:41:41: warning: "BOOST_MSVC" is not defined archive/basic_text_oprimitive.hpp:29:51: warning: "BOOST_DINKUMWARE_STDLIB" is not defined I'm building on a mac (10.4.6, intel), and the header file in which I'm including the serialization stuff looks like this: #include <string> #include <sstream> #include <boost/archive/tmpdir.hpp> #include <boost/archive/xml_iarchive.hpp> #include <boost/archive/xml_oarchive.hpp> Any help appreciated Thanks, Olivier Destrebecq olivierd(a)real.com

1 0

BGL: add_vertex deletes properties
by Fred Labrosse 20 Apr '06

20 Apr '06

All, I have: typedef boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, place, link> mapType; place* newPlace = new place("here-and-there"); mapType::vertex_descriptor newPlaceDesc = add_vertex(*newPlace, theMap); where the destructor of place is: place::~place() { // Just to make sure... std::cout << "Deleting place " << name << std::endl; // ... } When I run that, I do get: Deleting place here-and-there just after adding the vertex. In other words, a property is being deleted, probably a copy of the one I gave. I even tried (to make it explicitly like the documentation says): mapType::vertex_descriptor newPlaceDesc = add_vertex(static_cast<const place&>(*newPlace), theMap); and the same happens. Is that suppose to happen? Cheers, Fred

2 1

[thread] Multiple readers, one writer locking
by Scott 20 Apr '06

20 Apr '06

I need to design a multiple reader, one writer type thread locking on a STL container. I've been using boost::threads for all my threading and locking so far, but I haven't found a way to do what I want. Then I saw this in the release notes: *************** Read/write mutex functionality added New classes, boost::read_write_mutex, boost::try_read_write_mutex, and boost::timed_read_write_mutex were added. Note Since the read/write mutex and related classes are new, both interface and implementation are liable to change in future releases of Boost.Threads. The lock concepts and lock promotion in particular are still under discussion and very likely to change. *************** But I can't find any documentation on it. Can someone help me out here? Thanks, Scott

2 1

newbie question to Boost.Build linking to an external library jam
by wewolski 20 Apr '06

20 Apr '06

Hi, I am trying to compile the example from regex_token_iterator.html using the following entries in the Jamroot file. #Jamroot exe readdata : readdata.cpp : <include>/data/cpp/boost/include/boost-1_33_1/ ; install dist : readdata : <location>. ; and im getting a linker error (collect2: ld returned 1 exit status) However all my attempts to specify the external library location failed. For example I have tried to add lib boost_regex : : <file>boost_regex <search>/data/cpp/boost/lib/ ; exe regexexp : regexexp.cpp .//boost_regex ... and many more weird stuff without success. I found something in about site-config.jam in the manual but the example is not very clear to me. Can anyone please help? Eryk

1 0

Class method cascade
by John Metta 20 Apr '06

20 Apr '06

Does Boost have any magic to help with a class method cascade? Consider: struct INFO { float mass; float volume; ... virtual void Reset() { mass = 0.0f; volume = 0.0f; ... }; } struct WATER_INFO : virtual INFO { float rateInfow; float rateOutflow; ... virtual void Reset() { rateInflow = 0.0f; rateOutflow = 0.0f; super.Reset(); // Wish to cascade a reset function down. }; }; struct SURFACE_WATER_INFO : virtual INFO { ... } // cascading to reset the variables at any level. Disclaimer: I recognize that this is a ... let's say "less than elegant" solution. Unfortunately, I've been given a 20,000 line monolithic implementation (there should be no fewer than 10 in here) of a jungle and am trying to put a bit of duct tape on some things in order to make it run to finish my thesis. Given time, I would re-write the entire thing correctly, but I don't have that luxury currently. Cheers, -j -- John Metta http://jmetta.com "A computer without windows is like a dog without bricks tied to its head."

2 1

[format] 64bit integer
by Christian Henning 20 Apr '06

20 Apr '06

Hi there, boost::format will not swallow "%I64d". Instead, it throws a boost::io::bad_format_string exception. How can I use 64 bit integers? Thanks, Christian

4 8

[program_options] Description formatting failes
by elviin 20 Apr '06

20 Apr '06

Hello Boost Users, OS: Windows XP Compiler: VS 2005, 8.0.50727.42 Boost: 1.33, cvs Attachments: screen shot of assertion failure I have a problem with description formatting. It's not possible to make longer description. How long the description must be depends on other descriptions. So adding a description longer than some size invokes assertion in xstring file in the Visual Studio distribution. I tested this issue also with example "options_description.cpp". I changed description at --help switch from ("help", "produce help message") //that is working for me to the description on http://www.boost.org/doc/html/program_options/overview.html#id2714805 ("help", " A long help msg a long help msg a long help msg a long help msg a long help msg a long help msg a long help msg a long help msg ") //that failes

1 0

BGL autoconf macros
by Fred Labrosse 20 Apr '06

20 Apr '06

All, Anybody knows of autoconf macros that would test for the presence of graph related stuff? I could not find any in boost itself and the ones on the Web are not specific to BGL. Cheers, Fred

2 2

BGL: write_graphviz and bundled properties
by Fred Labrosse 20 Apr '06

20 Apr '06

All, First, I'm very new to boost and BGL, so please forgive me if this is trivial. I've created a graph (adjacency_list) using bundled properties and am now trying to write it as a dot file. I've written my property writer, but when I compile, I get several of: /usr/include/boost/graph/graphviz.hpp: In function `void boost::write_graphviz(std::ostream&, const Graph&, VertexPropertiesWriter, EdgePropertiesWriter, GraphPropertiesWriter, VertexID) [with Graph = boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, VertexPropertiesWriter = map::placeWriter, EdgePropertiesWriter = boost::default_writer, GraphPropertiesWriter = boost::default_writer, VertexID = boost::adj_list_vertex_property_map<boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>]': /usr/include/boost/graph/graphviz.hpp:260: instantiated from `void boost::write_graphviz(std::ostream&, const Graph&, VertexPropertiesWriter, EdgePropertiesWriter, GraphPropertiesWriter) [with Graph = boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, VertexPropertiesWriter = map::placeWriter, EdgePropertiesWriter = boost::default_writer, GraphPropertiesWriter = boost::default_writer]' /usr/include/boost/graph/graphviz.hpp:278: instantiated from `void boost::write_graphviz(std::ostream&, const Graph&, VertexWriter) [with Graph = map::mapType, VertexWriter = map::placeWriter]' topologicalMap.h:282: instantiated from here /usr/include/boost/graph/graphviz.hpp:241: error: no match for 'operator<<' in 'out << boost::get [with PropertyMap = boost::adj_list_vertex_property_map<boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t>, Reference = const boost::detail::error_property_not_found&, K = void*](((const boost::put_get_helper<const boost::detail::error_property_not_found&, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t> >&)((const boost::put_get_helper<const boost::detail::error_property_not_found&, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t> >*) ((boost::put_get_helper<const boost::detail::error_property_not_found&, boost::adj_list_vertex_property_map<boost::adjacency_list<boost::listS, boost::listS, boost::undirectedS, map::place, map::link, boost::no_property, boost::listS>, boost::detail::error_property_not_found, const boost::detail::error_property_not_found&, boost::vertex_index_t> >*) (&vertex_id)))), ((void* const&)((void* const*) (+(&i)->std::_List_iterator<_Tp>::operator* [with _Tp = void*]()))))' /usr/lib/gcc/x86_64-pc-linux-gnu/3.4.5/include/g++-v3/bits/ostream.tcc:63: note: candidates are: std::basic_ostream<_CharT, _Traits>& std::basic_ostream<_CharT, _Traits>::operator<<(std::basic_ostream<_CharT, _Traits>&(*)(std::basic_ostream<_CharT, _Traits>&)) [with _CharT = char, _Traits = std::char_traits<char>] Not being very good at deciphering all that, I'm not sure what is happening. I do understand that it's missing an extraction operator but I'm not sure on what. The actual line in graphviz.hpp is: out << get(vertex_id, *i); Two questions: - Could somebody tell me exactly what that get returns? - From the dot format, this should be the name of the vertex in the dot file. So this seems to relate to the class used as a property. My second question is then: If this is the case, why then do we need the property writer since we apparently also need the operator<< that could do exactly the same thing? (And isn't a ' ' (space) missing between the vertex name and whatever the property writer writes (or should that be included in it, which is not mentioned by the documentation)? That was 3 questions then ;-) Thanks in advance for any replies. Cheers, Fred

2 4

lambda/bind trouble
by Sebastian Weber 20 Apr '06

20 Apr '06

Hello everybody! I'm starting to use the boost lambda library. My first steps with it were quite amazing, but now I'm comming to some more advanced problems and I don't see my fault. The code in question is std::for_each(vertices(bg).first, vertices(bg).second, bind(ll::fill_n(), // line 197, supposenly wrong std::back_inserter(vs), bind(out_degree, _1, bg), _1)); which does not compile with the error message: box_cover.hpp: In function `size_t boost::box_cover(Graph&, RandomGenerator&, BoxMap, unsigned int, unsigned int)': box_cover.hpp:197: error: parse error before `::' token "g++-3.3" -Wall -ftemplate-depth-100 -msse2 -mfpmath=sse,387 -O3 -finline-functions -Wno-inline-fPIC -msse2 -mfpmath=sse,387 -DBOOST_DATE_TIME_DYN_LINK=1 -DDATE_TIME_INLINE -DNDEBUG -I"." -I"/home/sebi/projects/diplom/src/boost_1_33_1" -c -o "../../bin/graph/gcc-3.3/release/boxTest.o" "boxTest.cpp" I thought, I could realize this stuff with the boost::bind library as well and tried the construct std::for_each(vertices(bg).first, vertices(bg).second, boost::bind(std::fill_n, std::back_inserter(vs), boost::bind(boost::out_degree, _1, bg), _1) ); which does not compile as well: box_cover.hpp: In function `size_t boost::box_cover(Graph&, RandomGenerator&, BoxMap, unsigned int, unsigned int) [with Graph = main(int, char**)::graph_t, RandomGenerator = base_rand_gen_t, BoxMap = size_t*]': boxTest.cpp:140: instantiated from here box_cover.hpp:202: error: no matching function for call to `bind(<unknown type>, boost::arg<1>&, boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS, boost::no_property, boost::no_property, boost::no_property, boost::listS>&)' "g++-3.3" -Wall -ftemplate-depth-100 -msse2 -mfpmath=sse,387 -O3 -finline-functions -Wno-inline-fPIC -msse2 -mfpmath=sse,387 -DBOOST_DATE_TIME_DYN_LINK=1 -DDATE_TIME_INLINE -DNDEBUG -I"." -I"/home/sebi/projects/diplom/src/boost_1_33_1" -c -o "../../bin/graph/gcc-3.3/release/boxTest.o" "boxTest.cpp" Any help would be greatly appreciated. Thanks a lot! Almost forgot: My Boost version is 1.33.1 and the compiler I use is g++ 3.3.5 from Debian sarge. Greetings, Sebastian

1 0

[weak_ptr] thread safe?
by Alain Cormier 19 Apr '06

19 Apr '06

Hello group! We are experiencing problems with boost::weak_ptr in a multi-threaded environment. It looks that the lock on the ref counting doesn't work well with weak_ptr in release build. configuration : Windows 2000 SP4 VC6 SP5 boost 1.33.1 In summary, it seems that add_ref_lock in sp_counted_base_w32.hpp has a bug in release. In disassemblies, we observe : 01 lea esi,[eax+4] 02 mov eax,dword ptr [esi] 03 test eax,eax 04 je TestWeakPtr+0E4h (00401244) 05 lea ecx,[eax+1] 06 mov edx,esi 07 lock cmpxchg dword ptr [edx],ecx 08 mov ecx,eax 09 cmp ecx,eax 10 je TestWeakPtr+7Dh (004011dd) At line 08, you'll see that we move eax into ecx and after (line 09) we compare ecx and eax that are obviously the same which will destruct prematurely our pointer. In attachement, I send a complete program reproducing the bug with VC6 in release. It is probably a bug in VC6. But is it caused by a bad use? is it simply a "bug" in boost 1.33.1? By the way, we can remove the "bug" by simply "patching" add_ref_lock implementation like this : bool add_ref_lock() // true on success { for( ;; ) { long tmp = static_cast< long const volatile& >( use_count_ ); if( tmp == 0 ) return false; long tmp2 = tmp + 1; if(InterlockedCompareExchange ( &use_count_, tmp2, tmp ) == tmp2 - 1) return true; } } Thank you for any suggestion! -- Alain Cormier alain.work(a)gmail.com

2 4

Re: [Boost-users] [weak_ptr] thread safe?
by Sohail Somani 19 Apr '06

19 Apr '06

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of > Michael Fawcett > Sent: Wednesday, April 19, 2006 10:09 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] [weak_ptr] thread safe? > > On 4/19/06, John Wilkinson <jwilkinson(a)tsystem.com> wrote: > > There is a pragma in VC6 that will activate or deactivate > optimization. > > It can be used to wrap optimization-sensitive blocks of > code. I do not > > remember what it is, though, and I do not have VC6 handy. > I have used > > it, and it did alleviate an optimization-related bug. > > >From the MSDN: > > #pragma optimize( "[optimization-list]", {on | off} ) > > Specifies optimizations to be performed on a function-by-function > basis. The optimize pragma must appear outside a function and takes > effect at the first function defined after the pragma is seen. The on > and off arguments turn options specified in the optimization-list on > or off. I'm curious, why couldn't this be a compiler workaround in weak_ptr?

2 1

[Pool] Using purge_memory() with std::map
by Finley Lee 19 Apr '06

19 Apr '06

Dear list, I'm using boost::fast_pool_allocator to replace the default allocator in a map. This works great, but I'm having a bit of trouble freeing the memory explicitly. I'd like to call boost::singleton_pool::purge_memory(), but I don't know how to instantiate the correct version. Is there a way to determine the size of the chunks the map allocator will require? I have determined by debugging that the size is 28 bytes, but I imagine that this is dependent on the particular implementation of map. Below is a sample program which illustrates the situation. Any guidance would be very appreciated. - Finley //////////////////////////////////////////////////////////////////////// /// #include <assert.h> #include <iso646.h> #include <map> #include <boost\pool\pool_alloc.hpp> class mydata { public: mydata() {} mydata( int _a, short _b, char _c ) : a(_a), b(_b), c(_c) {} int a ; short b ; char c ; }; typedef std::map< int, mydata, std::less<int>, boost::fast_pool_allocator< std::pair<int, mydata> > > mymap ; int _tmain(int argc, _TCHAR* argv[]) { { mymap m ; for ( int i = 0 ; i < 10000 ; i++ ) m[i] = mydata( i, i, i ) ; } bool memory_freed = boost::singleton_pool< boost::fast_pool_allocator_tag, sizeof( std::pair<int,mydata> ) >::purge_memory() ; //bool memory_freed = boost::singleton_pool< boost::fast_pool_allocator_tag, 28 >::purge_memory() ; assert( memory_freed ) ; return 0; }

3 2

scoped locks
by Paul Baxter 19 Apr '06

19 Apr '06

Having looked at Boost threads and shmem 0.9 recently, I'm trying to get to grips with a 'scoped lock' ONLY design rationale having laboured under the posix threads APIs for a while. While I appreciate the reasons, I am a little confused ... All the examples I've seen show scoped locks being constructed local to a function and then destroyed automatically at the end of function scope. I would like my own lock manager object to manage a group of locks and allow me to leave one or more mutexes locked across function calls until the lock manager object goes out of scope and releases any mutexes still flagged as locked. Is this possible? e.g. a vector/array/deque of pointers to scoped_timer_locks each relating to its own mutex and constructed as unlocked. I understand that leaving mutexes locked is not a great idea! The application does think it knows what it is doing. Paul

4 3

boost::hash
by george 19 Apr '06

19 Apr '06

hello! I wanna replace an std::map<std::string,Services*> to hash table.How I will do this? in the examples of hash library there are only examples for custom hashing.... there isn't a trivial example!:-) transactions.insert(std::pair<std::string,Services*>("Aservice",new xxxx)); transactions.insert(std::pair<std::string,Services*>("Vservice",new fffff)); transactions.insert(std::pair<std::string,Services*>("Rservice",new ttt)); ............... how I should replace this with a boost 'hashtable'? thanks for any help.

1 1

Re: [Boost-users] [weak_ptr] thread safe?
by John Wilkinson 19 Apr '06

19 Apr '06

There is a pragma in VC6 that will activate or deactivate optimization. It can be used to wrap optimization-sensitive blocks of code. I do not remember what it is, though, and I do not have VC6 handy. I have used it, and it did alleviate an optimization-related bug. John -----Original Message----- From: boost-users-bounces(a)lists.boost.org [mailto:boost-users-bounces@lists.boost.org] On Behalf Of Peter Dimov Sent: Wednesday, April 19, 2006 10:45 AM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] [weak_ptr] thread safe? Alain Cormier wrote: > So, can I conclude that it is a bad behavior in boost::weak_ptr? No, it is a bug in VC6. > For now, I have simply changed add_ref_lock to : > > long tmp2 = tmp + 1; > if(InterlockedCompareExchange ( &use_count_, tmp2, tmp ) == tmp2 - 1 ) > return true; > > This way, VC6 doesn't optimise the return value of > InterlockedCompareExchange and the disassembly looks ok. I don't think that there is any guarantee that this modification will reliably produce correct code, even if it happens to work in the examples we try. But if we can't think of a better way to work around the issue, this may be our best bet. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users

3 2

using lambda as function argument.
by chun ping wang 19 Apr '06

19 Apr '06

Heres a simple question, that i want to do. Lets say you have a function call void f(double); and then you want to apply f to each argument inside vector<double> a; for_each(a.begin(), a.end(), f(_1)); // the example above isn't possible..is there a simple way to do this?

2 4

[Xpressive] regex-like boost libraries vs re2c [was: Some timings [was: Baffled about results of search]]
by Lynn Allan 19 Apr '06

19 Apr '06

Stuart Dootson wrote: > Lynn - it might be interesting to use re2c (http://re2c.org/) to > generate regex machines and see what their performance is like - > after > all, they do clain 're2c generates warning free code that is equal > to > hand-written code in terms of size, speed and quality.' (I'm > offering > no opinion on the veracity of that statement....), so (allegedly) > they'd give a reasonable hand-code analogue.... I took a look at it, and it looks like it could possibly be quite useful if executable size was a concern .... but it is pretty obscure (at least to me) .... the only example this dummy could make any sense out of at all was to recognize a number .... and that was missing/obscuring the important illustration of how to invoke and use the generated scanner code. Were you able to get it to work to do anything even close to recognizing Days-of-the-Week (full and abbreviations)? Did you find actual compilable examples with "main" invoking the generated scanners with a test-string to crunch? <alert comment="whining"> I really wish developers would provide COMPLETE "baby steps with training wheels" examples to flatten the learning curve. Eric N. did a great job of documenting and providing examples for xpressive .... that could be a model for other projects (not just boost). I've subscribed to their sourceforge list and will swallow my pride and ask the newbie question: "This dummy doesn't expect to be able to figure out your project in 30 minutes ... how do you get it to something basic like recognize:"? ((Sunday|Sun)|(Monday|Mon)....(Saturday|Sat)) </alert>

3 5

[bgl-python] write graphviz with position property
by David E. Konerding 19 Apr '06

19 Apr '06

Hi, From what I can tell it's not possible to correctly wirte a graphviz file that has a position property (where position is a tuple of xy) using bgl-python. The problem, as I see it, is that the wrapper for write_graphviz in bgl-python writes all the dynamic properties to the graph, but provides no ability to provide custom writer methods for those properties. So the resulting graphviz file prints references to the vertex position *object*, but not its values. Does anybody have a better suggestion? The only think I can see is to update the bgl-python wrappers to take function arguments for custom property writers Dave

1 0

program_options question
by Santosh Joseph 19 Apr '06

19 Apr '06

If I have an option that looks like this: options_description desc("options"); desc.add_options()("server,s",value<string>(&server),""); I know this can parse the following options: --server arg -s arg but how can I also parse the following automatically: -server arg /server arg Is there a style flag to set? I looked at the documentation but I'm a little confused by all the style_t flags. or do I have to write a "additional parser"? thanks, -Santosh btw: i'm using 1.33.1 __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

is there any lib used to parse XML file in boost?
by lanser 19 Apr '06

19 Apr '06

hello, i need a simple xml prasing library, i consider there should be some in boost lib but i can't find it or document for this, help!

10 15

Re: [Boost-users] [boost] [statechart] - BOOST_ASSERTION in state_machine::send_eventif used by multiple threads
by Oliver.Kowalke＠infineon.com 19 Apr '06

19 Apr '06

Same happens if I install a exception_translator in the enclosing state: /opt/boost/include/boost-1_33_1/boost/statechart/state_machine.hpp:833: void boost::statechart::state_machine<MostDerived, InitialState, Allocator, ExceptionTranslator>::send_event(const boost::statechart::event_base&) [with MostDerived = iris::proto::svc::sm<ogx::net::ipv4::stream>, InitialState = iris::proto::svc::proto<ogx::net::ipv4::stream>, Allocator = std::allocator<void>, ExceptionTranslator = boost::statechart::exception_translator<boost::statechart::exception_thr own>]: Assertion `get_pointer( pOutermostUnstableState_ ) == 0' failed. >Hello, >I've several threads running. Each thread has its own state_machine. >Sometime (not reproducible) I get following assertion: > >/opt/boost/include/boost-1_33_1/boost/statechart/state_machine.hpp:833: >void boost::statechart::state_machine<MostDerived, InitialState, >Allocator, ExceptionTranslator>::send_event(const >boost::statechart::event_base&) [with MostDerived = >iris::proto::svc::sm<ogx::net::ipv4::stream>, InitialState = >iris::proto::svc::proto<ogx::net::ipv4::stream>, Allocator = >std::allocator<void>, ExceptionTranslator = >boost::statechart::null_exception_translator]: Assertion `get_pointer( >pOutermostUnstableState_ ) == 0' failed.

1 0