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

Boost.Python 'Fatal Error'
by Ed Fear 16 Aug '04

16 Aug '04

Hello there, I've just started using Boost.Python, and find it absolutely invaluable. For the moment, my main task is embedding the python interpreter within an application of mine, and I have successfully built a small application that can load a module, execute a Python function and then collect the result. The next logical step was to extend the Python script somewhat into objects. I wrote some extra code in the script which I checked with the command line interpreter and verified to work. Before changing the C++ code to actually /use/ these new objects, I decided to just see what would happen if I kept exactly the same C++ code but with the new Python script (which still contains the function used for the last example). It ran fine - producing the correct output - but then exited abruptly with the following message: "Fatal Python error: PyThreadState_Get: no current thread". The changes to the original script file were the addition of two new classes - OptionsGUI and Button - and, oddly, commenting out Button makes the program work again (I say oddly because it's simply a class with a constructor, nothing more). Evidently I'm using the library incorrectly, and have searched through the documentation for things I might have done wrong but alas, I do find it rather daunting. I also noticed that my code is quite different from examples - is this a correct way to be doing things? And yes, at present, I'm not doing any exception or error handling - I'm a little unclear on it but I'm sure I will be able to do it given time :) For those wishing to look at the source, I have put the two files - EmbeddedPythonFunc.cpp and script2.py - in my webspace to prevent everyone having to download them as attachments. They are available at the following addresses: http://kutsurogu.pwp.blueyonder.co.uk/EmbeddedPythonFunc.cpp http://kutsurogu.pwp.blueyonder.co.uk/script2.py (incidentally, I didn't include stdafx.h because it simply includes iostream, tchar.h and boost/python.hpp). I hope I have sent my help request to the right place! Apologies if I have not. Thank you in advance, Ed Fear

2 1

polymorphic container/boost::any
by goochrules! 16 Aug '04

16 Aug '04

I'm quite sure that this problem has been addressed many times before, so I suspect it as a better name and I suspect that boost::any could be the solution. I want to store related types in the same stl container, for instance, this is the class heirachy I want: class tree { public: void grow() =0; }; class oak : public tree { public: void grow() { std::cout << "oak grows slowly" << std::endl; }; class pine : public tree { public: void grow() { std::cout << "pine grows quickly" << std::endl; }; Then, I try to put them in a std::list: std::list< tree > forest; forest.push_back(oak()); forest.push_back(pine()); ...and I find out that I can't define tree::grow as abstract. So I give it a dummy implentation -- void grow() { std::cout << "tree grows" << std::endl; --, and try to iterate over the objects: for(std::list< tree >::iterator iter = forest.begin(); forest.end() != iter; ++iter) { iter->grow(); } ...and I don't get what I expect: tree grows tree grows I then think about trying to use a std::list< tree& > and discover that it won't compile. I think briefly consider std::list< tree* > but decide this dosen't meet requirements of memory management. So I do some searching for 'polymorphic container' and 'polymorphic iterator' and come up with boost::any, but I've no luck in making it work. Attached is my code. In the end, I'm looking for a way to add related types to a stl container and not have to do anything complicated to call a method on the type's abstract interface and to not have to wory about memory management. Maybe this question is better suited for comp.lang.c++[.moderated]? -- Matthew Peltzer -- goochrules(a)gmail.com

4 6

problem with static linking on linux gcc 2.95.3
by Daniel Huber 16 Aug '04

16 Aug '04

Hi, I am wondering if any of you boost gurus can help with this problem. I have compiled and installed boost using gcc 2.95.3 on a linux redhat 7 machine. The install actually fails on some targets. I get this message at the end of the install: ...failed updating 24 targets... ...skipped 48 targets... ...updated 2343 targets... But I am guessing that this is expected behavior for an old compiler. Anyway, I am interested in the filesystem library at this time, and so I tried compiling the simple_ls.cpp example. When I compile using the default linking (which uses shared objects), the program works fine, but when I compile using the -static flag to limit to static linking, I get a lot of undefined reference errors (see below). Unfortunately, I haven't been able to figure out why this would happen. Any suggestions on how to debug this or as to what might be going wrong would be greatly appreciated. Sincerely, Daniel /tmp/ccU0ehTU.o: In function `main': /tmp/ccU0ehTU.o(.text+0xc): undefined reference to `boost::filesystem::initial_path(void)' /tmp/ccU0ehTU.o(.text+0x2c): undefined reference to `boost::filesystem::native(basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &)' /tmp/ccU0ehTU.o(.text+0x3e): undefined reference to `boost::filesystem::path::path(char const *, bool (*)(basic_string<char, string_char_traits<char>, __default_alloc_template<true, 0> > const &))' /tmp/ccU0ehTU.o(.text+0x50): undefined reference to `boost::filesystem::system_complete(boost::filesystem::path const &)' /tmp/ccU0ehTU.o(.text+0xe8): undefined reference to `boost::filesystem::exists(boost::filesystem::path const &)' /tmp/ccU0ehTU.o(.text+0x10d): undefined reference to `boost::filesystem::path::native_file_string(void) const' /tmp/ccU0ehTU.o(.text+0x188): undefined reference to `boost::filesystem::is_directory(boost::filesystem::path const &)' /tmp/ccU0ehTU.o(.text+0x1ad): undefined reference to `boost::filesystem::path::native_directory_string(void) const' /tmp/ccU0ehTU.o(.text+0x257): undefined reference to `boost::filesystem::is_directory(boost::filesystem::path const &)' /tmp/ccU0ehTU.o(.text+0x289): undefined reference to `boost::filesystem::path::leaf(void) const' /tmp/ccU0ehTU.o(.text+0x2f5): undefined reference to `boost::filesystem::path::leaf(void) const' /tmp/ccU0ehTU.o(.text+0x427): undefined reference to `boost::filesystem::path::native_file_string(void) const' /tmp/ccU0ehTU.o(.text+0x5ab): undefined reference to `boost::filesystem::path::leaf(void) const' /tmp/ccU0ehTU.o: In function `boost::filesystem::directory_iterator::directory_iterator(boost::filesystem::path const &)': /tmp/ccU0ehTU.o(.boost::filesystem::directory_iterator::gnu.linkonce.t.(boost::filesystem::path const &)+0x26): undefined reference to `boost::filesystem::detail::dir_itr_init(boost::shared_ptr<boost::filesystem::detail::dir_itr_imp> &, boost::filesystem::path const &)' /tmp/ccU0ehTU.o: In function `boost::filesystem::directory_iterator::dereference(void) const': /tmp/ccU0ehTU.o(.boost::filesystem::directory_iterator::gnu.linkonce.t.dereference(void) const+0x12): undefined reference to `boost::filesystem::detail::dir_itr_dereference(boost::shared_ptr<boost::filesystem::detail::dir_itr_imp> const &)' /tmp/ccU0ehTU.o: In function `boost::filesystem::directory_iterator::increment(void)': /tmp/ccU0ehTU.o(.boost::filesystem::directory_iterator::gnu.linkonce.t.increment(void)+0x12): undefined reference to `boost::filesystem::detail::dir_itr_increment(boost::shared_ptr<boost::filesystem::detail::dir_itr_imp> &)' collect2: ld returned 1 exit status

1 0

Re:Re: Regex help needed, please help
by Kosta Adarchenko 16 Aug '04

16 Aug '04

Message: 8 Date: Sun, 15 Aug 2004 10:31:50 +0100 From: "John Maddock" <john(a)johnmaddock.co.uk> Subject: Re: [Boost-users] Regex help needed, please help. To: <boost-users(a)lists.boost.org> Message-ID: <00a601c482ab$78c9dc10$d8ec0352@fuji> Content-Type: text/plain; charset="iso-8859-1" > John, I have stated the exact regex, it's \d\d\d\d\d\d , a simple regex matching 6 numbers, but the same happens even with one \d regex. I don't see a way to refactor the \d expression. OK, in that case I don't know what the problem is - are you able to generate a test case (mail me privately if the files are large)? And again, from where is the exception being thrown (in other words is it regex that's at fault)? John. ---------------------------------- The exception is being thrown from the search function, I generated a test case and sent it to your email at john(a)johnmaddock.co.uk, files are fairly large compressed 246kb, opened the take about 25MB (Love to see zip on text files). Thanks ahead Kosta. --------------------------------- Do you Yahoo!? New and Improved Yahoo! Mail - 100MB free storage!

1 0

Regex help needed, please help.
by Kosta Adarchenko 15 Aug '04

15 Aug '04

boost-users-request(a)lists.boost.org wrote: Send Boost-users mailing list submissions to boost-users(a)lists.boost.org To subscribe or unsubscribe via the World Wide Web, visit http://www.lists.boost.org/mailman/listinfo.cgi/boost-users or, via email, send a message with subject or body 'help' to boost-users-request(a)lists.boost.org You can reach the person managing the list at boost-users-owner(a)lists.boost.org When replying, please edit your Subject line so it is more specific than "Re: Contents of Boost-users digest..." Today's Topics: 1. Re: Regex problems, please help (John Maddock) ---------------------------------------------------------------------- Message: 1 Date: Fri, 13 Aug 2004 11:39:54 +0100 From: "John Maddock" Subject: Re: [Boost-users] Regex problems, please help To: Message-ID: <01d401c48125$ee4e79f0$3c520352@fuji> Content-Type: text/plain; charset="iso-8859-1" > After some seven thousands hits the match algo throughs the memory exhausted exception. > So I cought the exception and moved the iterator by one, printed the location it points to in my searched file to view the problematic section for the regex find algorithm. > I have noticed that the problem happens not with all large files, but with quite a bit of them. > After I have received the problematic section in my file, I ran the program on it and it did not raise any exceptions. > Please , help me I have a Graduate project to complete within 3 weeks from now, and this is one of the major bugs I have not resolved. You don't say *exactly* what the regex is that's causing the problem - it makes a difference - problems like this are usually caused by an expression that behaves badly with specific input (because regexes are NP-Hard to match in the general case). Boost.Regex will throw an exception rather than go on thrashing forever trying to find a match, but to determine whether this is the case here, you need to trap the exception at the point it's being thrown, and see if the regex lib is indeed the culprit, if it is, then refactoring the expression that you're using so that it's less ambiguous is usually the answer. John. ------------------------------ _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://www.lists.boost.org/mailman/listinfo.cgi/boost-users End of Boost-users Digest, Vol 279, Issue 2 ******************************************* John, I have stated the exact regex, it's \d\d\d\d\d\d , a simple regex matching 6 numbers, but the same happens even with one \d regex. I don't see a way to refactor the \d expression. --------------------------------- Do you Yahoo!? New and Improved Yahoo! Mail - Send 10MB messages!

2 1

[boost-users]_BGL:_copy_component_bugs
by Erik Jenelius 13 Aug '04

13 Aug '04

Hi. In boost/graph/copy.hpp, besides the copy_graph function, I noticed a nice, undocumented function called copy_component which uses breadth-first search to copy a component from a graph to a new graph. Unfortunately, it doesn't work. I think there are at least two bugs in it: 1. The vertices are copied at the examine_vertex event point and the edges are copied at the examine_edge event point. However, the target of the edge has not yet been examined at that point, which leads to strange results. If the edges are copied at the black_target event point instead, the algorithm seems to work correctly. 2. The std::vector orig2copy is templated on the vertex descriptor type of the input graph, while it should be the output graph. The function copy_graph is correct here. Also, both copy_graph and copy_component by default assume the vertices have a vertex_all property and the edges an edge_all property, which are copied to the output graph. Wouldn't it be more intuitive to not assume and copy any properties by default? - Erik Jenelius

1 0

Regex problems, please help
by Kosta Adarchenko 13 Aug '04

13 Aug '04

Here is the problem: I want to find a simple regular _expression something like \\d\\d\\d\\d\\d\\d in a large file (15 mb) using VC++ 7.1 . I use Spirit's File Iterator as the input iterator for the search algorithm. After some seven thousands hits the match algo throughs the memory exhausted exception. So I cought the exception and moved the iterator by one, printed the location it points to in my searched file to view the problematic section for the regex find algorithm. After the "problematic section" of some 300 chars ends the algorithm keeps working fine. here is the code. //------------------------------------------------------------------------------------------- void TheFunction() { //print the problematic section here ofstream of; of.open("d:\\err1.txt"); boost::spirit::file_iterator< > begin("d:\\log1.txt"); boost::spirit::file_iterator<> beginning_saved(begin); //save the beginning in order to compare later boost::spirit::file_iterator< > ending; ending= begin.make_end(); boost::match_results<boost::spirit::file_iterator<char > > what; boost::match_flag_type flags = boost::match_default; long matchescounter=0; std::exception e; int i; bool found; found=false; try { found=boost::regex_search(begin, ending, what, _expression, flags); if (found) { begin =what[0].second; cout<<what[0]; matchescounter=1; } } catch (std::exception &e) { cout<<e.what(); } while(found) { try { found=boost::regex_search(begin, ending, what, _expression, flags); begin =what[0].second; if (found) { cout<<matchescounter<<"----------"<<what[0]<<"-------------------"<<what[0].second-beginning_saved<<"\n"; matchescounter++; } } catch (std::exception &e) { cout<<*begin; of<<*begin; begin+=1; } } cout<<"---------------the end------------\n"; cout<<matchescounter; of.close(); } //--------------------------------------------------------------------------------------- I have noticed that the problem happens not with all large files, but with quite a bit of them. After I have received the problematic section in my file, I ran the program on it and it did not raise any exceptions. Please , help me I have a Graduate project to complete within 3 weeks from now, and this is one of the major bugs I have not resolved. Yours truely, Kosta. Desparatly waiting for your reply. --------------------------------- Do you Yahoo!? Y! Messenger - Communicate in real time. Download now.

2 1

RE: [Boost-users] polymorphic container/boost::any
by Klaus Nowikow 13 Aug '04

13 Aug '04

> dp12-ipvi(a)spamex.com wrote: > > Use std::list<boost::shared_ptr> std::list<boost::shared_ptr<tree> > to be exact ;-) > > I hope I spelled all that correctly. > > ++PLS -- Klaus > > -----Original Message----- > From: goochrules! [mailto:goochrules@gmail.com] > Sent: Thursday, August 12, 2004 2:02 PM > To: boost-users(a)lists.boost.org > Subject: [Boost-users] polymorphic container/boost::any > > * Replies will be sent through Spamex to boost-users(a)lists.boost.org > * For additional info click -> http://www.spamex.com/i/?v=4259897 > I'm quite sure that this problem has been addressed many > times before, so I > suspect it as a better name and I suspect that boost::any could be the > solution. > I want to store related types in the same stl container, for > instance, this > is the class heirachy I want: > class tree { public: void grow() =0; }; > class oak : public tree { public: void grow() { std::cout << > "oak grows > slowly" << std::endl; }; class pine : public tree { public: > void grow() { > std::cout << "pine grows quickly" << std::endl; }; Then, I > try to put them > in a std::list: > std::list< tree > forest; > forest.push_back(oak()); > forest.push_back(pine()); > ...and I find out that I can't define tree::grow as abstract. > So I give it > a dummy implentation -- void grow() { std::cout << "tree grows" << > std::endl; --, and try to iterate over the objects: > for(std::list< tree >::iterator iter = forest.begin(); > forest.end() != iter; > ++iter) { > iter->grow(); > } > ...and I don't get what I expect: > tree grows > tree grows > I then think about trying to use a std::list< tree& > and > discover that it > won't compile. I think briefly consider std::list< tree* > > but decide this > dosen't meet requirements of memory management. > So I do some searching for 'polymorphic container' and 'polymorphic > iterator' and come up with boost::any, but I've no luck in > making it work. > Attached is my code. > In the end, I'm looking for a way to add related types to a > stl container > and not have to do anything complicated to call a method on the type's > abstract interface and to not have to wory about memory > management. Maybe > this question is better suited for comp.lang.c++[.moderated]? > -- Matthew Peltzer > -- goochrules(a)gmail.com > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

1 0

compiling boost using Intel C++ 8.0 under Windows
by Lars Schouw 13 Aug '04

13 Aug '04

I am havng problems building boost using the intel C++ 8.0 compiler. It looks as if bjam can't fine the icl.exe compiler executable itself. This file is installed under this directory C:\Program Files\Intel\CPP\Compiler80\Ia32\Bin Any ideas what I do wrong? I try this: run the iclvars.bat batch file. Then I use bjam like this: C:\dev\boost\boost_1_31_0_icc>bjam "-sTOOLS=intel-win32" "-sINTEL_VERSION=80" "-sINTEL_BASE_MSVC_TOOLSET=vc7.1" But when compiling I get these errors from each command to the compiler: /Zm800 -nologo -GX -c -DNDEBUG -DNDEBUG -DBOOST_DATE_TIME_POSIX_TIME_STD_CONFIG -DBOOST_ALL_DYN_LINK=1 /Ogity /O2 /Gs /Ob2 /GX /GR /MD /Op /Zc:forScope /Zc:wchar_t /Qwn5 /Qwd985 -Qoption,c,--arg_dep_lookup /Qansi_alias -D_NATIVE_WCHAR_T_DEFINED /Qvc7.1 -I"bin\boost\libs\date_time\build" -I"C:\dev\boost\boost_1_31_0_icc" -I"C:\Program Files\Microsoft Visual Studio .NET 2003\VC7\include" -Fo"bin\boost\libs\date_time\build\boost_date_time.dll\intel-win32\release\greg_weekday.obj" -Tp"C:\dev\boost\boost_1_31_0_icc\libs\date_time\build\../src/gregorian/greg_weekday.cpp" ...failed vc-C++ bin\boost\libs\date_time\build\boost_date_time.dll\intel-win32\release\greg_weekday.obj... vc-C++ bin\boost\libs\date_time\build\boost_date_time.dll\intel-win32\release\date_generators.obj '/Zm800' is not recognized as an internal or external command,operable program or batch file. Lars --------------------------------- Do you Yahoo!? Yahoo! Mail Address AutoComplete - You start. We finish.

1 0

RE: [Boost-users] polymorphic container/boost::any
by dp12-ipvi＠spamex.com 12 Aug '04

12 Aug '04

Use std::list<boost::shared_ptr> I hope I spelled all that correctly. ++PLS -----Original Message----- From: goochrules! [mailto:goochrules@gmail.com] Sent: Thursday, August 12, 2004 2:02 PM To: boost-users(a)lists.boost.org Subject: [Boost-users] polymorphic container/boost::any * Replies will be sent through Spamex to boost-users(a)lists.boost.org * For additional info click -> http://www.spamex.com/i/?v=4259897 I'm quite sure that this problem has been addressed many times before, so I suspect it as a better name and I suspect that boost::any could be the solution. I want to store related types in the same stl container, for instance, this is the class heirachy I want: class tree { public: void grow() =0; }; class oak : public tree { public: void grow() { std::cout << "oak grows slowly" << std::endl; }; class pine : public tree { public: void grow() { std::cout << "pine grows quickly" << std::endl; }; Then, I try to put them in a std::list: std::list< tree > forest; forest.push_back(oak()); forest.push_back(pine()); ...and I find out that I can't define tree::grow as abstract. So I give it a dummy implentation -- void grow() { std::cout << "tree grows" << std::endl; --, and try to iterate over the objects: for(std::list< tree >::iterator iter = forest.begin(); forest.end() != iter; ++iter) { iter->grow(); } ...and I don't get what I expect: tree grows tree grows I then think about trying to use a std::list< tree& > and discover that it won't compile. I think briefly consider std::list< tree* > but decide this dosen't meet requirements of memory management. So I do some searching for 'polymorphic container' and 'polymorphic iterator' and come up with boost::any, but I've no luck in making it work. Attached is my code. In the end, I'm looking for a way to add related types to a stl container and not have to do anything complicated to call a method on the type's abstract interface and to not have to wory about memory management. Maybe this question is better suited for comp.lang.c++[.moderated]? -- Matthew Peltzer -- goochrules(a)gmail.com

1 0

Random Library
by Andy Tompkins 12 Aug '04

12 Aug '04

Hi, I have a question about the random library. I'm using vc7.1 and I'm getting a warning when I include the random library. The warning I get when I, #include <boost\random.hpp> is: d:\includes\boost_1_31_0\boost\random\linear_congruential.hpp(233) : warning C4244: 'return' : conversion from 'boost::uint64_t' to 'boost::int32_t', possible loss of data When I look at the file, rand48::operator() returns an int32_t, but lcf is defined with uint64_t. Thanks in advance for any help. Andy.

1 0

RE: [Boost-users] Regex problem
by Evgeny Benediktov 12 Aug '04

12 Aug '04

It was the case. Thank you. Evgeny Benediktov -----Original Message----- From: John Maddock [mailto:john@johnmaddock.co.uk] Sent: Thursday, August 12, 2004 1:04 PM To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Regex problem >When extensive usage of regex_match and regex_search functions is done numerous times, then after several thousands iterations >Regex library starts throwing "Regular expression too big" exception on the same regular expressions it worked perfectly with at >several thousands previous iterations. >Looks like Regex fails to do some vital clean up. >If anyone saw this problem before, or has any suggestion about it, this may be very helpful. It's most likely that those exceptions are coming from the matching algorithms (regex_match/regex_search)? If so then it's because the combination of the particular regular expression, and the particular text being matched, were becoming "pathological", so an exception is thrown to prevent the match taking forever to obtain (regexes are NP-Hard to match in the worst case). Often these issues can be resolved by refactoring the regular expression, but you really need to track down which specific expression, against what specific text is causing the problem - try setting your debugger to break when an exception is thrown, and then work back up the stack to see what caused the issue. John Maddock. _______________________________________________ Boost-users mailing list Boost-users(a)lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users ***************************** Finjan Software This e-mail and any attached files are confidential and may be legally privileged. The unauthorized use, disclosure or copying of this email or any information contained within it is strictly prohibited. This also confirms that Finjan Software's Vital Securityfor E-Mail has scanned this message for the presence of known viruses and potentially malicious code. Finjan Software - Prevention is the Best Cure! *****************************

1 0

uBLAS - How to use a range of a matrix row?
by Pascal Bauer 12 Aug '04

12 Aug '04

Hi, I have some (strange?) problems with the ublas: I'm writing a SPH-Code for fluids and brittle solids. I use a uBLAS matrix to store all the particle-Data. Now, since the SPH code itself includes a lot of physics, i wrote a "particle" class, that give me access to the values stored in the matrix in an intuitive way. As template argument, i'll give a matrix, so i do no have to take care about matrix internals. It looks somehow thatlike: template <typename T> class Particle { static T& SPHData; //Matrix where the values are stored typename T::size_type GID; //Particle ID (...) matrix_row<T> LData; // Proxy to the particle data vector_range<matrix_row<T> > Pos; // Position vector, for // calculations... (...) typename T::value_type& mass; // Reference to physical values... Particle(typename T::size_type ParticleID); // Constructor, needs // Particle ID } template <typename T> Particle<T>::Particle(typename T::size_type ParticleID) : mass(SPHData(ParticleID, MASS)) // Set the references... { LData.reset(SPHData, ParticleID); // OK, works fine... Pos.reset(LData, range(0, 2)); // Give strange error message: // (at the end of this mail...) } I tried different methods to get the Pos to the right place in my matrix, but if i use the "=" operator, i've got some strange run-time assertion errors. My thing is, that i'd like to have a range of a matrix_row or column, that i can use later for vector-operations. How can I do that? - Pascal Bauer PS: Compile Error for my example code. */home/pbauer/oosph/src/particlecontainer.h:103: instantiated from `sph::Particle<T>::Particle(typename T::size_type) [with T = ParticleTest()::Matrix]' */home/pbauer/oosph/src/test.cpp:36: instantiated from here */usr/include/boost/numeric/ublas/vector_proxy.hpp:126: error: no matching function for call to ` boost::numeric::ublas::matrix_row<boost::numeric::ublas::matrix<double, boost::numeric::ublas::row_major, boost::numeric::ublas::unbounded_array<double> > >::reset( boost::numeric::ublas::matrix_row<boost::numeric::ublas::matrix<double, boost::numeric::ublas::row_major, boost::numeric::ublas::unbounded_array<double> > >&)' */usr/include/boost/numeric/ublas/matrix_proxy.hpp:112: error: candidates are: void boost::numeric::ublas::matrix_row<E>::reset(M&) [with M = boost::numeric::ublas::matrix<double, boost::numeric::ublas::row_major, boost::numeric::ublas::unbounded_array<double> >] */usr/include/boost/numeric/ublas/matrix_proxy.hpp:117: error: void boost::numeric::ublas::matrix_row<E>::reset(M&, typename M::size_type) [with M = boost::numeric::ublas::matrix<double, boost::numeric::ublas::row_major, boost::numeric::ublas::unbounded_array<double> >] -- \ | Pascal Bauer | | Space Research & Planetary Sciences | Research Division | | University of Bern | | Phone: +41316314416 | | e-mail: mailto:pascal.bauer@phim.unibe.ch /

1 0

RE: [Boost-users] Feature for using C++ as script language
by Persson Jonas 12 Aug '04

12 Aug '04

> -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org]On Behalf Of Jeff Garland > Sent: den 11 augusti 2004 17:19 > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] Feature for using C++ as script language > > > On Wed, 11 Aug 2004 11:33:50 -0300, Maurício wrote > > Hi, > > > > After I discovered Boost, which has regex and filesystem > > libraries, I decided to use C++ for tasks that I usually do with > > Ruby (a Perl-like language), because I know C++ a lot better than > > other languages. However, there's a feature of those languages I > > miss: the ability to call external programs and get the output as a > > string. Does any of Boost libraries (or other free libraries you > > know) have this capability? > > There isn't a boost library, but it would be a nice addition. > Turns out the > 'system' call is available on both windows and unix so you > can do something > like this: > > #include <stdlib.h> > > std::string command = "...your command here..." > std::string tmpfile_name("...your temporary file name here..."); > std::string cmd = command + "> " + tmpfile_name; > system(cmd.c_str()); > std::ifstream output(tmpfile_name.c_str()); > > Of course this has many limits an issues. If you want > something more formal > pstreams might work for you: > > http://pstreams.sourceforge.net/ > > HTH, > > Jeff > Another one is here: http://libexecstream.sourceforge.net/ / Jonas

1 0

Regex problem
by Evgeny Benediktov 12 Aug '04

12 Aug '04

Hello, I want to report a problem with Regex library. When extensive usage of regex_match and regex_search functions is done numerous times, then after several thousands iterations Regex library starts throwing "Regular expression too big" exception on the same regular expressions it worked perfectly with at several thousands previous iterations. Looks like Regex fails to do some vital clean up. If anyone saw this problem before, or has any suggestion about it, this may be very helpful. Thanks, Evgeny ***************************** Finjan Software This e-mail and any attached files are confidential and may be legally privileged. The unauthorized use, disclosure or copying of this email or any information contained within it is strictly prohibited. This also confirms that Finjan Software's Vital Securityfor E-Mail has scanned this message for the presence of known viruses and potentially malicious code. Finjan Software - Prevention is the Best Cure! *****************************

2 1

[BLG] Problem: PropertyMap with three properties
by Ambichl Christoph 11 Aug '04

11 Aug '04

Hello, I am trying to implement an edge-property map with three properties, but it does not work as I expected it. Below is the source code and the compiler error message. Thanks for your help! Christoph *** Source Code *** #include <boost/property_map.hpp> #include <boost/graph/graph_traits.hpp> #include <boost/graph/adjacency_list.hpp> using namespace boost; enum edge_type_t { edge_type }; enum edge_cplex_x_t { edge_cplex_x }; namespace boost { BOOST_INSTALL_PROPERTY(edge, type); BOOST_INSTALL_PROPERTY(edge, cplex_x); } // first try typedef property<edge_type_t, int, property<edge_weight_t, int, property<edge_cplex_x_t, float> > > EdgeProperty; // second try typedef property<edge_type_t, int> EdgeType; typedef property<edge_weight_t, int> EdgeWeight; typedef property<edge_cplex_x_t, int> EdgeX; typedef property< EdgeType, EdgeWeight, EdgeX > EdgeProperty2; typedef adjacency_list<vecS, vecS, directedS, no_property, EdgeProperty> Eag; typedef graph_traits<Eag>::vertex_descriptor vertex_descriptor; typedef graph_traits<Eag>::edge_descriptor edge_descriptor; typedef pair<int,int> Edge; int main() { Eag eag(5); property_map<Eag,edge_type_t>::type e_type = get(edge_type,eag); property_map<Eag,edge_weight_t>::type e_weight = get(edge_weight,eag); property_map<Eag,edge_cplex_x_t>::type e_x = get(edge_cplex_x,eag); add_edge(vertex(0,eag), vertex(1,eag), EdgeProperty(0,3,1.0), eag); return 0; } *** Error Message *** propertymap.cpp: In function `int main()': propertymap.cpp:44: error: no matching function for call to ` boost::property<edge_type_t, int, boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> > >::property( int, int, double)' /usr/local/include/boost/pending/property.hpp:16: error: candidates are: boost::property<edge_type_t, int, boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> > >::property(const boost::property<edge_type_t, int, boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> > >&) /usr/local/include/boost/pending/property.hpp:22: error: boost::property<Tag, T, Base>::property(const T&, const Base&) [with Tag = edge_type_t, T = int, Base = boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> >] /usr/local/include/boost/pending/property.hpp:21: error: boost::property<Tag, T, Base>::property(const T&) [with Tag = edge_type_t, T = int, Base = boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> >] /usr/local/include/boost/pending/property.hpp:20: error: boost::property<Tag, T, Base>::property() [with Tag = edge_type_t, T = int, Base = boost::property<boost::edge_weight_t, int, boost::property<edge_cplex_x_t, float, boost::no_property> >]

2 1

Feature for using C++ as script language
by Maurício 11 Aug '04

11 Aug '04

Hi, After I discovered Boost, which has regex and filesystem libraries, I decided to use C++ for tasks that I usually do with Ruby (a Perl-like language), because I know C++ a lot better than other languages. However, there's a feature of those languages I miss: the ability to call external programs and get the output as a string. Does any of Boost libraries (or other free libraries you know) have this capability? Thanks, Maurício

2 1

(newbie) design question (is there an MPL solution?)
by Timothy Perrigo 11 Aug '04

11 Aug '04

I'm just now treading into the (rather intimidating) waters of template metaprogramming, so I'm not even sure if it will fit what I'm trying to accomplish. I need to create a kind of registry to keep track of instances of variables of different types (actually, several of these registries, representing different system states); variables of a type would be identified by an unique (integer) id. In other words, the variables within a particular registry represent the state of the system at a given point in time, so although each registry would have variables of the same types and ids, the values of the variables in different registries may be different. If the variables were all of the same type, a simple map would work. What I would like is a way of asking this registry for a variable of a specified type and id, and get back the appropriate instance (of the appropriate type). I.e., something like the following would be ideal: MyType mt = registry.get<MyType>(my_id); Is there anything in the MPL (or elsewhere in Boost) that might help me accomplish what I'm looking to do? If not, does anyone have any design suggestions to accomplish this in another way? Any help would be appreciated! Thanks, Tim

4 7

Newbie questions and possible bug reports
by Daniel Huber 11 Aug '04

11 Aug '04

Hi, I am a new user to Boost, and I am trying to compile and install Boost for the first time. In the process, I have run into some problems that the Boost maintainers may be interested in. First, I was wondering if there is a specification as to which version of gcc the Boost libraries are known to compile with. In particular, I have a lot of code that only compiles under 2.95.3, and I get a lot of error messages when compiling. Here's an example: /usr/bin/objcopy --set-section-flags .debug_str=contents,debug "bin/boost/libs/test/build/libboost_test_exec_monitor.so/gcc/debug/shared-linkable-true/test_tools.o" ...failed gcc-C++-action bin/boost/libs/test/build/libboost_test_exec_monitor.so/gcc/debug/shared-linkable-true/test_tools.o... If it turns out that Boost is supposed to work with this version of gcc, I can provide more detailed results. I think that due to the failures, the libraries and include files were not installed by the bjam command. Second, I noticed that the zip version of the latest release 1_31_0 contains files that are in windows format (ie, the files contain carriage returns). I'm not sure whether this is intended, since most people who download the zip extension will use it on windows. However, if you try to install from a unix variant, you get very strange errors. Specifically, when running "sh build.sh" in the tools/build/jam_src dir, the program dies with: : command not found : command not found : command not found 'uild.sh: line 13: syntax error near unexpected token `() 'uild.sh: line 13: `echo_run () I don't know whether there are files with this problem outside the tools/build/jam_src directory. Finally, the link to the "boost public CVS repository" on the bug report instructions page (http://www.boost.org/more/bugs.htm) is broken. Sincerely, Daniel Huber

3 2

Thanks - problem solved
by Daniel Huber 10 Aug '04

10 Aug '04

FYI, this did the trick. Thanks for your help. -Daniel boost-users-request(a)lists.boost.org wrote: >Here is what I did to install: >> >> First go to tools/build/jam_src and build jam: >> sh build.sh >> >> Then cd back to the boost root dir, and install: >> tools/build/jam_src/bin.linuxx86/bjam --sTOOLS=gcc >> --prefix=~/local/linux install Two things: a) It should be -sTOOLS=gcc but since the default is gcc on Linux anyway this doesn't really affect the results. b) "~/local/linux" is a shell construct which will only gets expanded by the shell if it's by itself as an argument. You heed to use $HOME/local/linux instead.

1 0

Problem compiling boost under gcc
by Daniel Huber 10 Aug '04

10 Aug '04

Hi, This is a followup to my previous message. I must be doing something wrong, but I am unable to compile boost under gcc 3.2.2 on a Red Hat Linux machine (RH9 linux 2.4.24). Note, I also tried gcc 2.95.3 with the same results. If anyone has the time or the desire to help me out, it would be much appreciated. Here is what I did to install: First go to tools/build/jam_src and build jam: sh build.sh Then cd back to the boost root dir, and install: tools/build/jam_src/bin.linuxx86/bjam --sTOOLS=gcc --prefix=~/local/linux install I get a lot of errors relating to the HardLink command. (I am wondering if the problem might be caused by the bjam program.) I checked the things I thought might be causing problems: The target directory does exist, but the files are not getting copied into that directory (eg. by the FileClone command). I have write permissions in the target directory. I am attaching the output from the compile process. Sincerely, Daniel Huber

2 1

bjam options
by Samuel 09 Aug '04

09 Aug '04

I agree with the "nuclear engineer" that the "Getting Started" instructions are a little confusing, but I think it is at least as good as (for example) the instructions for MySQL. There are a few things that are not clear, and I will post separate messages for most of them. One little detail is that the the "Getting Started" instructions use double quotes in the examples of the bjam options, yet as best as I can determine, the double quotes don't work. In other words, the following does not work: bjam "--help" And the following works: bjam --help It is true that there is an example showing the version that works, but I missed seeing it until I began writing this message and the other examples are confusing to the extent that they make it seem as if the double quotes are needed. So in other words, I don't understand why the double quotes are shown in the sample uses of bjam.

2 7

My Boost got built
by Samuel 08 Aug '04

08 Aug '04

Is there a description of what additional files need to be copied? I used the following to build Boost for Visual C++ with Windows XP: Call "S:\Program Files\Microsoft Visual Studio\VC98\Bin\VCVARS32.BAT" Set MSVC_ROOT=S:\Program Files\Microsoft Visual Studio\VC98 "S:\Software\Boost\boost-jam-3.1.10-1-ntx86\bjam" -sTOOLS=msvc --prefix="S:\ Software\Boost" --builddir="T:\boost_1_31_0\BuildDir" install Pause That seems to have worked. Except I did see some messages saying some stuff were not found. That is not relevant to this question, though. I copied the sample in the Variant documentation to get the directories set properly and ensure that that much (Variant) works. So now (I assume) I don't need to use most of the files (in the "T:" partition) that I used to build and "install" Boost from. However I see there are some documentation files in the "install from" files that are not in my installed files, so I assume I should copy the documentation files. I realize that the files are also online but I will use the local copy that is provided. Is there something somewhere specifying what else should be saved from the "install from" files? I intend to make the "T:" partition unavailable normally but I can keep it so that I can get to it when I need to, but it would help to know what I need to copy first. If this is not described anywhere, then I just want to say that it would help. Until then, I assume I can figure it out.

1 0

install | stage
by Samuel 08 Aug '04

08 Aug '04

The definition of the terms "install" and "stage" is obviously clear to those that know what they mean, but it is not clear to me how they relate to Boost. I have worked with software for more than a quarter of a century, and I have even developed a couple of software promotion packagers, so I do have a basic understanding of the terms "install" and "stage". As best as I can guess, an "install" would be a combination of a "stage" and something else. So I assume I can do an install, but I might be wrong. So I think it would help to define those terms a bit more for the "Getting Started" page, for the benefit of all. For me, I have a Windows system and I have downloaded: http://aleron.dl.sourceforge.net/sourceforge/boost/boost_1_31_0.zip I have unzipped that to a temporary location. It is not clear to me if that will work. I am not sure if I can do an "install" and have everything I need put in the location I specify for using Boost. If that is explained somewhere then I apologize for missing it, but I think it is not made very clear.

1 0

Boost Installation Help needed
by Aaronburro＠netscape.net 07 Aug '04

07 Aug '04

Howdy, Sorry if this gets posted twice. i tried the webpage over at news.gmane.org, but it doesn't look like it got posted, so I subscribed(yay). Anyway, I'm trying to install Boost on Windows with the DJGPP compiler for c++. I've basically done nothing but get one big headache trying to follow the instructions at boost.org. So far, I have bjam in my windows\system32 directory. I tried to set it up to use DJGPP, a variant of gcc, but it doesn't seem to like my request. bjam keeps trying to use g++ instead of gpp, even though I gave it the GXX=gpp option. It also can't seem to find the directory of DJGPP, even though I used the GCC_ROOT_DIRECTORY=C:\DGJPP\ option as well. I am thoroughly stumped, and the documents over at boost.org seem to have been written by monkeys, albeit ones with good spelling and grammar. I am fairly computer literate, but those things are like rocket science to me(and I'm a nuclear engineer ;)) I'd use Boost.Build's .bat file, but I can't get bjam to work right, and then God only knows what I'll do from there, so I'd greatly appreciate a more "step-by-step" approach which is geared towards a WindowsXP user. Bryan Burroughs __________________________________________________________________ Switch to Netscape Internet Service. As low as $9.95 a month -- Sign up today at http://isp.netscape.com/register Netscape. Just the Net You Need. New! Netscape Toolbar for Internet Explorer Search from anywhere on the Web and block those annoying pop-ups. Download now at http://channels.netscape.com/ns/search/install.jsp

3 2