- Boost-users - lists.boost.org

[BGL] build problems
by Angus Leeming 09 Aug '03

09 Aug '03

Dear all, I'm a bjam newbie trying to build the BGL in boost 1.30.0 on a DEC Alpha running linux. However, the build process is failing to generate the graphviz parser and lexer: $ cd libs/graph/build $ bjam "-sTOOLS=gcc" ...found 437 targets... ...using 2 temp targets... ...updating 3 targets... ...using <libs!graph!build/liblibbgl-viz.a/gcc/debug/inlining-on/runtime-link-dynamic>graphviz_graph_lex.o... ...using <libs!graph!build/liblibbgl-viz.a/gcc/debug/inlining-on/runtime-link-dynamic>graphviz_digraph_lex.o... gcc-C++-action ../../../libs/graph/build/bin/liblibbgl-viz.a/gcc/debug/inlining-on/runtime-link-dynamic/graphviz_graph_parser.o graphviz_parser.y:44:2: #error Need to define the GRAPHVIZ_GRAPH macro to either GraphvizGraph or GraphvizDigraph. graphviz_parser.y:61: error: `GRAPHVIZ_GRAPH' was not declared in this scope graphviz_parser.y:61: error: template argument 1 is invalid graphviz_parser.y:61: error: syntax error before `;' token graphviz_parser.y:62: error: `GRAPHVIZ_GRAPH' was not declared in this scope graphviz_parser.y:62: error: template argument 1 is invalid graphviz_parser.y:62: error: syntax error before `;' token graphviz_parser.y:63: error: syntax error before `;' token [snip] This all seems very reasonable, but I do not know how to resolve the problem. I see that Jamfile.v2 has explicit rules to build graphviz_graph_lex.o, graphviz_digraph_lex.o, setting GRAPHVIZ_GRAPH appropriately. However, my version of bjam (3.1.4) does not understand Jamfile.v2 and I lack the knowledge to do the equivalent in Jamfile. I did try to append some rules to the Jamfile(see below), but unsurprisingly without success. Could anyone provide me with pointers on how to proceed? Kind regards, Angus Lex graphviz_graph_lex.cpp : graphviz_lex.l : <define>GRAPHVIZ_DIRECTED=0 <dependency>graphviz_graph_parser ; Yacc graphviz_graph_parser.cpp : graphviz_parser.y : <define>GRAPHVIZ_DIRECTED=0 ; Lex graphviz_digraph_lex.cpp : graphviz_lex.l : <define>GRAPHVIZ_DIRECTED=1 <dependency>graphviz_digraph_parser ; Yacc graphviz_digraph_parser.cpp : graphviz_parser.y : <define>GRAPHVIZ_DIRECTED=1 ; Object graphviz_graph_lex.o : graphviz_graph_lex.cpp ; Object graphviz_graph_parser.o : graphviz_graph_parser.cpp ; Object graphviz_digraph_lex.o : graphviz_digraph_lex.cpp ; Object graphviz_digraph_parser.o : graphviz_digraph_parser.cpp ;

4 7

what is wrong with boost::bind?
by yxh 08 Aug '03

08 Aug '03

Hi, I am using boost::bind from boost 1.30.0. When I type "boost::bind(", the VS.NET 20003 IDE stopped responding. What is wrong? #include <boost/bind.hpp> int main(void) { boost::bind( // IDE dies here }

5 5

use of REGEX_SEARCH to match lookahed subexpression.
by yanbelinky 08 Aug '03

08 Aug '03

Dear Regex Guru, I am trying to use REGEX_SEARCH algorithm to match lookahed subexpression. Search in the folowing strings (see below) using expression like: /exchange/test/(?!Drafts)/a.EML/b.png fails /exchange/test/Drafts/a.EML/b.gif /exchange/test/Inbox/a.EML/b.png /exchange/test/Inbox/a.EML/b.gif /exchange/test/Drafts/a.EML/b.png At the same moment I can suceesfully search using the expression: /exchange/test/Inbox/a.EML/b.(?!png). >From online documentation I realize that regex supports lookahed search. Please help me understand what I am doing wrong. P.S. It is tested with the program closely modeled from example provided for regex_search. Thank you for your help. Yan Belinky Senior Software Engineer. Whale Communications, LTD.

2 1

RE: [Boost-Users] shared_ptr memory leak (newb)
by Mark Storer 07 Aug '03

07 Aug '03

*string allocates it's own copy of the string, so your pmem pointer is leaked. STL's string templates manage their own memory, you never need to use new or delete to [de]allocate for them. --Mark Storer Software Engineer Cardiff Software #include <disclaimer> typdef std::disclaimer<Cardiff> Discard; > -----Original Message----- > From: Aaron Campbell [mailto:yg-boost-users@m.gmane.org] > Sent: Thursday, August 07, 2003 1:22 PM > To: boost-users(a)yahoogroups.com > Subject: [Boost-Users] shared_ptr memory leak (newb) > > > why does this leak memory from the first call to "new"? > how can I prevent the leak? > > wchar_t* pmem = new wchar_t[10]; > boost::shared_ptr<std::wstring> cp ( new std::wstring(pmem) ); > > thanks for the help, > .Aaron > > > > > > ------------------------ Yahoo! Groups Sponsor > ---------------------~--> > Buy Ink Cartridges or Refill Kits for Your HP, Epson, Canon or Lexmark > Printer at Myinks.com. Free s/h on orders $50 or more to the > US & Canada. http://www.c1tracking.com/l.asp?cid=5511 > http://us.click.yahoo.com/l.m7sD/LIdGAA/qnsNAA/EbFolB/TM > -------------------------------------------------------------- > -------~-> > > Info: <http://www.boost.org> > Wiki: <http://www.crystalclearsoftware.com/cgi-bin/boost_wiki/wiki.pl> > Unsubscribe: <mailto:boost-users-unsubscribe@yahoogroups.com> > > > Your use of Yahoo! Groups is subject to http://docs.yahoo.com/info/terms/

1 0

shared_ptr memory leak (newb)
by Aaron Campbell 07 Aug '03

07 Aug '03

why does this leak memory from the first call to "new"? how can I prevent the leak? wchar_t* pmem = new wchar_t[10]; boost::shared_ptr<std::wstring> cp ( new std::wstring(pmem) ); thanks for the help, .Aaron

1 0

Problems with if_then_else_return in boost::lambda library
by Ioffe, Robert 07 Aug '03

07 Aug '03

Hi, If I have a vector of strings vector<string> v; v.push_back("One"); v.push_back("Two"); v.push_back("Three"); v.push_back("Four"); v.push_back("Five"); and define arithmetic action plus like this: namespace boost { namespace lambda { struct plain_return_type_2<arithmetic_action<plus_action>, std::string, std::string> { typedef std::string type; }; } } then the following code works just fine: cout << accumulate(v.begin()+1, v.end(), v.front(), _1 + string(", ") + _2) << endl; and the following also works wonderfully: int i = 1; cout << accumulate(v.begin()+1, v.end(), v.front(), _1 + bind(constructor<string>(), var(i)++, ' ') + _2) << endl; Now, if I want to use if_then_else_return as shown below, it fails miserably, producing a huge error message: cout << accumulate(v.begin(), v.end(), string(), if_then_else_return(_1 == string(), _1, _1 + bind(constructor<string>(), var(i)++, ' ') + _2)) << endl; Now for the questions: What is the proper way to use if_then_else_return? Why do I need to define a separate arithmetic action for strings, even though they are a common library type? Is there a way to say the most intuitive: cout << accumulate(v.begin()+1, v.end(), v.front(), _1 + ", " + _2) << endl; Is there way to do more intuitive: cout << accumulate(v.begin()+1, v.end(), v.front(), _1 + string(i++, ' ') + _2) << endl; Thank you for your help. Robert Ioffe [Non-text portions of this message have been removed]

1 0

[function] How to extract a tuple<A0,A1> of arguments from a boost::function<>?
by Fernando Cacciola 07 Aug '03

07 Aug '03

Short question: As the Subject says, given a boost::function{1..N}<>, how can I extract a tuple of all its arguments? Long question: Anyway, what I'm really trying to do is this: Given some classes with different methods: struct X { void f(int n ) ; } ; struct Y { void g(char c, double d ) ; } ; I want to easly create _polymorphic_ closures binding member functions of X /Y with instances of X/Y and function arguments. The goal is the have a deferred 'action' list for easy backtracing of an algorithm. That is, instead of directly calling: X x ; Y y ; x.f(1); y.g('a',1.2); I want to do something that would look like: X x; Y y; ActionList actions; actions.push_back( make_closure(&X::f, x,1) ); actions.push_back( make_closure(&Y::g, y,'a',1.2) ); std::for_each(actions.begin(),actions.end(),execute_action()); TIA Fernando Cacciola

2 2

filesystem compilation on tru64 unix
by allysonlurena 07 Aug '03

07 Aug '03

Hi all, I have been trying to build the filesystem library for one of our alpha servers running tru64 unix. I have managed to compile bjam using the tru64cxx toolset. Then, I went into the filesystem build directory and ran the following, but it has been hanging for over 30 minutes now on the last command you see below. Can anyone help? It could just be our machines, but I am at a loss. Does anyone have any ideas? Any help would be appreciated! Thanks Allyson $ cxx -V Compaq C++ V6.5-014 for Compaq Tru64 UNIX V5.1A (Rev. 1885) Compiler Driver V6.5-014 (cxx) cxx Driver rum:/.../boost_1_30_0/libs/filesystem/build $ bjam "-sTOOLS=tru64cxx65" ...found 432 targets... ...using 1 temp target... ...updating 9 targets... tru64cxx65-C++-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/exception.o tru64cxx65-C++-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/operations_posix_windows.o cxx: Warning: ../src/operations_posix_windows.cpp, line 236: #767-D conversion from pointer to smaller integer assert( m_imp.get() ); // fails if dereference end iterator ------^ cxx: Warning: ../src/operations_posix_windows.cpp, line 242: #767-D conversion from pointer to smaller integer assert( m_imp.get() ); // fails on increment end iterator ------^ tru64cxx65-C++-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/path_posix_windows.o tru64cxx65-C++-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/convenience.o tru64cxx65-Archive-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/libboost_filesystem.a ar: Warning: creating ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/libboost_filesystem.a Ranlib ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/debug/libboost_filesystem.a ...using <libs!filesystem!build/libboost_filesystem.a/tru64cxx65/release>exception.o... tru64cxx65-C++-action ../../../libs/filesystem/build/bin/libboost_filesystem.a/tru64cxx65/release/operations_posix_windows.o So, it seems to hang at the last command above. Thanks again! Allyson

3 2

cvs checkout probs
by Angus Leeming 07 Aug '03

07 Aug '03

Hi, I have just tried to checkout the boost cvs tree from sourceforge, but get stuck at cvs server: Updating boost/libs/numeric/mtl/test cvs server: [05:07:19] waiting for anoncvs_boost's lock in /cvsroot/boost/boost/libs/numeric/mtl/test I can subsequently check out the remaining stuff with $ cat tmp pool preprocessor property_map python random rational regex signals smart_ptr spirit static_assert test thread timer tokenizer tuple type_traits utility $ for dir in `cat tmp`; do cvs -z3 co boost/libs/$dir; done so it appears that the problem is specific to the boost/libs/numeric/mtl/test directory. Kind regards, Angus

1 0

[boost] filesystem lib & linking (MSVC.NET)
by Jonathan Roewen 07 Aug '03

07 Aug '03

Hi All, I'm trying to use the filesystem lib, but I'm having some problems figuring out how to build it properly. I'm using MSVC.NET on WindowsXP, and I ran the bjam tool like in the docs for the whole set of libraries. It took me a while to figure out where it had put everything, and finally found a file called libboost_filesystem.lib. I copied it to a convenient location, added the library path and lib file to my project settings, and then tried to link my app. MSVC then gave a bunch of linker errors, and suggested to add MSVCRT to the nodefaultlib setting, so I did that, and got less linker errors. All that is left is some thing to do with exception & string classes. Linker errors are all about multiply defined symbols. I suppose I'm doing something wrong somewhere along the line, but not sure what. I can't tell if bjam is actually doing what it's supposed to, and the documentation around the place can be pretty sparse in places, so I hope someone can point out what I can do to get things working. Regards, Jonathan Roewen

4 4