January 2005 - Boost-users - lists.preview.boost.org

exception safety of boost graph library
by Ovidiu Deac 28 Jan '05

28 Jan '05

Hello everybody, We are going to use boost graph library within our application and I would like to know if there are any known issues regarding the exception safety. Is the basic guarantee satisfied? Is the strong guarantee satisfied? (see Herb Sutter's definition of basic/strong guarantee in Exceptional C++) "Basic guarantee: Even in the presence of exceptions thrown by T or other exceptions, Stack objects don't leak resources. Note that this also implies that the container will be destructible and usable even if an exception is thrown while performing some container operation. However, if an exception is thrown, the container will be in a consistent, but not necessarily predictable, state. Containers that support the basic guarantee can work safely in some settings. Strong guarantee: If an operation terminates because of an exception, program state will remain unchanged. This always implies commit-or-rollback semantics, including that no references or iterators into the container be invalidated if an operation fails. For example, if a Stack client calls Top and then attempts a Push that fails because of an exception, then the state of the Stack object must be unchanged and the reference returned from the prior call to Top must still be valid." (Herb Sutter, Exceptional C++) Thank you, vidu

2 1

[string_algo] Split and Predicates
by David Brownell 28 Jan '05

28 Jan '05

I am attempting to split a string with the string_algo library, using the code below. It appears that the split algorithm is designed to split on instances of individual characters rather than sequences of characters. Is there something I can do to split on a substring? The code below illustrates the problem, where I am forces to use the is_any_of predicate rather than something that implies equality. It seems like this is a trivial use of the algorithm, so there could easily be something that I am missing. typedef std::vector<boost::iterator_range<char const *> > ResultsType; ResultsType results; static char const STRING[] = "this\r\nis\r\n\r\na test!"; boost::split(results, boost::iterator_range<char const *>(STRING, STRING + sizeof(STRING)), boost::is_any_of("\r\n")); // The problem is here. I would like to be able // to use something like equals, but split requires // a predicate that takes a single argument rather than // a range. for(ResultsType::const_iterator rptr = results.begin(); rptr != results.end(); ++rptr) std::cout << "'" << std::string(rptr->begin(), rptr->end()) << "'\n"; Thanks! David Brownell

2 1

Graphs (boost1.32.0) Trouble with dijkstra visitor and VC 7.1
by Tony Juricic 28 Jan '05

28 Jan '05

Has anybody created custom dijkstra_shortest_paths algorithm visitor for discover_vertex event? I am having all sorts of compilation errors using MSVC 7.1 and defining my visitor using example code. Reshuffling the code (but at this point I hardly know what am I really doing) I got the following to compile: // define visitor for discover_vertex event template <class Vertex, class Tag> struct target_visitor : public default_dijkstra_visitor { target_visitor(Vertex u) : v(u) { } template <class Vertex, class Graph> void discover_vertex(Vertex u, Graph& g) { if ( u==v ) assert(0); } private: Vertex v; }; template <class Vertex, class Tag> target_visitor<Vertex, Tag> target_visit(Vertex u, Tag) { return target_visitor<Vertex, Tag>(u); } ... // call dijkstra_shortest_paths with custom visitor // parameters below are defined as in dijkstra-example.cpp dijkstra_shortest_paths(g, s, predecessor_map(&p[0]).distance_map(&d[0]).visitor(target_visit(D, on_discover_vertex()))); Tony Confused

2 1

error in boost/graph/fruchterman_reingold.hpp
by Jamie Wilkinson 28 Jan '05

28 Jan '05

The doc at http://boost-consulting.com/boost/libs/grpah/doc/fruchterman_reingold.html suggests that there is a 4 argument version of the function, taking only the graph, a position map, width and height; but attempts to compile such a function against the headers from the latest CVS results in a compile error: error: no matching function for call to ` fruchterman_reingold_force_directed_layout(Umbrella::Document&, Umbrella::View::load_from_document()::PositionMap&, double&, double&)' Removing the typenames Param, Tag, and Rest from lines 371-372 of the header allows the compile to continue. Index: boost/graph/fruchterman_reingold.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/graph/fruchterman_reingold.hpp,v retrieving revision 1.5 diff -u -c -r1.5 fruchterman_reingold.hpp cvs diff: conflicting specifications of output style *** boost/graph/fruchterman_reingold.hpp 29 Dec 2004 16:53:05 -0000 1.5 --- boost/graph/fruchterman_reingold.hpp 12 Jan 2005 09:18:16 -0000 *************** *** 368,375 **** params); } ! template<typename Graph, typename PositionMap, typename Dim, typename Param, ! typename Tag, typename Rest> void fruchterman_reingold_force_directed_layout(const Graph& g, PositionMap position, --- 368,374 ---- params); } ! template<typename Graph, typename PositionMap, typename Dim> void fruchterman_reingold_force_directed_layout(const Graph& g, PositionMap position,

4 4

Bind Overload Workaround?
by Duane Murphy 28 Jan '05

28 Jan '05

How do I work around this problem with bind? #include <boost/bind.hpp> namespace { class functions { public: int func( int ); int func( int, int ); }; void test() { functions a_func; boost::bind( functions::func, a_func, 1, _1 ); } } Bind can't determine which function to call because it is overloaded and it doesn't know the context in which it will be called. With CodeWarrior 8 on Mac OS X I get "illegal use of template argument dependent type 'T::result_type' It references an "unspecified" template deep in the bowels of bind. As I recall there was some way of helping bind to pick one of the overloaded functions using a typedef or something similar. (Maybe that was function?). Thanks for any tips, ...Duane

2 2

boost serialization
by Daniel Goller 28 Jan '05

28 Jan '05

are there any problems with boost serialization including linker errors or does this indicate a broken install i have heard of problems with boost serialization but do not know which they might be before running off in the wrong direction i figured i would gather information here to solve this: http://bugs.gentoo.org/show_bug.cgi?id=78043 problem any help on this will be appreciated Daniel

4 6

Regex lib and thread safety
by Leon Mergen 27 Jan '05

27 Jan '05

Hello, I am sorry if this is a stupid question, or simply inappropriatie for this list. I am currently camping with this bug for way too long, so I decided to ask this mailinglist. I am currently in the process of finding out where a segfault is, and how to fix it. Now, I am using the boost regex library to parse my regexes, and for some reason when multiple threads use this regex library, it seems to mess up and generate segfaults (it works perfectly with only one thread). Since it also doesn't always occur on the same place, and doesn't even occur /all/ the times, I have a strong feeling this has something to do with thread safety. On the Boost website is was able to find out that the boost regex library should be thread safe when BOOST_HAS_THREADS is defined; I've tested this, and this worked. ( http://www.boost.org/libs/regex/doc/thread_safety.html ) However, I also was able to find a mailinglist message that provides some instructions on how to make certain boost regex functions thread-safe : ( http://lists.boost.org/MailArchives/boost/msg59110.php ) Now, I am confused; is the boost regex library thread-safe or not ? The solution provided in the mailinglist message ( move the regex_replace () function inside its own scope ) can't be applied here... :( I've posted the stacktrace when the problem occurs and the source of the code 'that matters' below. It also occurs in another function, I've posted the other function too. I hope anyone has any experience with this, or is able to help. Thanks in advance! Regards, Leon Mergen One stacktrace: --- #0 std::string::compare(char const*) const (this=0x0, __s=0x40338b81 "C") at /root/gcc-3.3.2/i686-pc-linux-gnu/libstdc++-v3/include/bits/basic_string.h:257 #1 0x4004b101 in boost::c_regex_traits<char>::update() () from /usr/local/lib/libboost_regex-gcc-1_31.so.1.31.0 #2 0x40084cd9 in boost::reg_expression<char, boost::regex_traits<char>, std::allocator<char> >::set_expression(char const*, char const*, unsigned) () from /usr/local/lib/libboost_regex-gcc-1_31.so.1.31.0 #3 0x080648aa in unsigned boost::reg_expression<char, boost::regex_traits<char>, std::allocator<char> >::set_expression<std::char_traits<char>, std::allocator<char> >>(std::basic_string<char, std::char_traits<char>, >>std::allocator<char> > const&, unsigned) (this=0x413d29cc, >>p=@0x413d2adc, f=34055) at basic_regex.hpp:110 #4 0x080647bf in reg_expression<std::char_traits<char>, std::allocator<char> > (this=0x413d29cc, p=@0x413d2adc, f=33031, a=@0x413d29bc) at basic_regex.hpp:114 #5 0x0806474c in basic_regex<std::char_traits<char>, std::allocator<char> > (this=0x413d29cc, p=@0x413d2adc, f=33031, a=@0x413d29bc) at basic_regex.hpp:361 #6 0x080633ee in RegexEngine::parseReplaceRegex(std::string*, std::string*, std::string*) (this=0x80d1160, regex=Cannot look up value of a typedef) at RegexEngine.cc:8 #7 0x0808070a in ModuleDigitsToText::applyModule(std::string*, std::string) (this=0x80d3190, data=0x80d16e0, language= {static npos = 4294967295, _M_dataplus ={<allocator<char>> = {<No data fields>}, _M_p = 0x80d244c "en"}, static _S_empty_rep_storage = {0, 0, 77, 0}}) at ModuleDigitsToText.cc:565 #8 0x08080270 in ModuleDigitsToText::applyModule(InputData*) (this=0x80d3190, inputData=0x80d16d8) at ModuleDigitsToText.cc:527 #9 0x08062200 in ModuleParser::parse(InputData*) (this=0x80d1150, inputData=0x80d16d8) at ModuleParser.cc:105 #10 0x0805e1db in Normaliser::normalise() (this=0x80d1698) at Normaliser.cc:107 #11 0x080831f4 in Client::operator()() (this=0x80ca808) at Client.cc:44 #12 0x080542bf in boost::detail::function::void_function_obj_invoker0<Client, void>::invoke(boost::detail::function::any_pointer) (function_obj_ptr= {obj_ptr = 0x80ca808, const_obj_ptr = 0x80ca808, func_ptr = 0x80ca808, data = "\b"}) at function_template.hpp:128 #13 0x400a69f4 in boost::thread_group::join_all() () from /usr/local/lib/libboost_thread-gcc-mt-1_31.so.1.31.0 #14 0x401cff60 in pthread_start_thread () from /lib/i686/libpthread.so.0 #15 0x401d00fe in pthread_start_thread_event () from /lib/i686/libpthread.so.0 #16 0x402f7327 in clone () from /lib/i686/libc.so.6 --- And are the two functions where the segfault occurs. Line 8 of RegexEngine.cc is the "boost::regex expression(*regex);" line of the parseReplaceRegex() function. --- std::string RegexEngine::parseReplaceRegex (const std::string * regex, const std::string * replace, const std::string * subject) { // Create an expression out of the regex boost::regex expression(*regex); // Return the replaced regex { return boost::regex_replace(*subject, expression, *replace, boost::match_default | boost::format_sed); } } bool RegexEngine::parseMatchRegex (std::string * regex, std::string * subject, RegexResultSet * what) { #ifdef DEBUG std::cout << "+ RegexEngine::parseMatchRegex[long] ( " << *regex << ", " << *subject << " )" << std::endl; #endif // DEBUG // Prepare result set boost::match_results <std::string::const_iterator> results; // Create string iterators std::string::const_iterator start = subject->begin(); std::string::const_iterator end = subject->end(); // Initialise boolean to know wether a match was reached bool success = false; // Create an expression out of the regex boost::regex expression(*regex); std::vector < std::string > regexResults; // Iterate over the results while (boost::regex_search(start, end, results, expression, boost::match_default)) { regexResults.clear(); for (boost::match_results<std::string::const_iterator>::iterator i = results.begin(); i != results.end(); ++i) { regexResults.push_back(*i); } // Push back a result what->push_back(regexResults); // Set new string iterator start = results[0].second; // Mark that a match was reached success = true; } #ifdef DEBUG std::cout << "- RegexEngine::parseMatchRegex ()" << std::endl; #endif // DEBUG // Return wether a match was reached return success; } ---

4 11

time_period constructor question
by Stuart Siegel 27 Jan '05

27 Jan '05

Can anyone tell me why the line: time_period t4(t2); in the attached cpp file doesn't compile but the line: time_period t3(t1); does? it seems like i should be calling the same copy constructor in both cases but g++ 3.3.4 complains: error: no matching function for call to ` boost::date_time::period<boost::posix_time::ptime, boost::posix_time::time_duration>::period(boost::posix_time::time_period (&)(boost::posix_time::ptime, boost::posix_time::ptime))' /usr/include/boost/date_time/period.hpp:64: error: candidates are: boost::date_time::period<boost::posix_time::ptime, boost::posix_time::time_duration>::period(const boost::date_time::period<boost::posix_time::ptime, boost::posix_time::time_duration>&) /usr/include/boost/date_time/period.hpp:109: error: boost::date_time::period<point_rep, duration_rep>::period(point_rep, duration_rep) [with point_rep = boost::posix_time::ptime, duration_rep = boost::posix_time::time_duration] /usr/include/boost/date_time/period.hpp:99: error: boost::date_time::period<point_rep, duration_rep>::period(point_rep, point_rep) [with point_rep = boost::posix_time::ptime, duration_rep = boost::posix_time::time_duration] //--------------------------------cut here------------------- #include "boost/date_time/posix_time/posix_time_types.hpp" #include "boost/date_time/gregorian/gregorian_types.hpp" int main() { using namespace boost::posix_time; using namespace boost::gregorian; //---------------------------------------------- date dmin( min_date_time) ; ptime tmin( dmin ); date dmax( max_date_time) ; ptime tmax( dmax ); time_period t1( tmin, tmax ); //--------------------------------------------- time_period t2( ptime(min_date_time), ptime(max_date_time) ); //why does this compile (on g++ 3.3.4) time_period t3(t1); //but this doesn't? i.e., why isn't a time_period a time_period a time_period? time_period t4(t2); return 0; } __________________________________ Do you Yahoo!? All your favorites on one personal page � Try My Yahoo! http://my.yahoo.com

4 5

analogy to mem_fun_ref to member variables
by Chris Goller 27 Jan '05

27 Jan '05

I have a struct that contains two values. struct box { double width; double length; } vector<box> boxes; I want to accumlate all the widths of the boxes. The only solution I've found is to create a member function get_width that will return the width for use in mem_fun_ref eg: accumulate(make_transform_iterator(boxes.begin(), mem_fun_ref(&boxes::get_width)), make_transform_iterator(boxes.end(), mem_fun_ref(&boxes::get_width)), 0.0, plus<double>())); Is there a boost way of not using mem_fun_ref, but rather returns the public member variable. Something that could be named mem_var_ref that would look like this: accumulate(make_transform_iterator(boxes.begin(), mem_var_ref(&boxes::width)), make_transform_iterator(boxes.end(), mem_fun_ref(&boxes::width)), 0.0, plus<double>())); This way I wouldn't need to add extra get logic to a structure that doesn't really need it othewise. Thanks, Chris

3 2

boost/graph 1.32.0 : compilation errors on a SUN machine
by laurent de Vito 26 Jan '05

26 Jan '05

Hello, I cannot compile the boost/graph library 1.32.0 on a SUN machine using Sun C++ 5.5. I get the following error message : boost_1_32_0/boost/graph/depth_first_search.hpp", line 377: Error: Default arguments cannot be added in later declarations of the template function in the same scope. which does make sense since the compiler cannot decide which routine depth_first_visit to use if the second routine (line 371) has an extra template but the the extra template has a default value. So in boost/graph/depth_first_search.hpp, I simply deleted the default argument : template <class IncidenceGraph, class DFSVisitor, class ColorMap, class TerminatorFunc> void depth_first_visit (const IncidenceGraph& g, typename graph_traits<IncidenceGraph>::vertex_descriptor u, DFSVisitor vis, ColorMap color, TerminatorFunc func = TerminatorFunc()) becomes template <class IncidenceGraph, class DFSVisitor, class ColorMap, class TerminatorFunc> void depth_first_visit (const IncidenceGraph& g, typename graph_traits<IncidenceGraph>::vertex_descriptor u, DFSVisitor vis, ColorMap color, TerminatorFunc func) Unfortunately the compilation was afterwards not successful; this time I got "boost_1_32_0/boost/graph/named_function_params.hpp", line 500: Error: Cannot return const boost::bgl_named_params<boost::detail::components_recorder<int*>, boost::graph_visitor_t, boost::no_property> from a function that should return boost::detail::error_property_not_found. "boost_1_32_0/boost/graph/depth_first_search.hpp", line 354: Where: While instantiating "boost::get_param<boost::graph_visitor_t, boost::root_vertex_t, boost::detail::components_recorder<int*>, boost::no_property>(const boost::bgl_named_params<boost::detail::components_recorder<int*>, boost::graph_visitor_t, boost::no_property>&, boost::root_vertex_t)". "boost_1_32_0/boost/graph/depth_first_search.hpp", line 354: Where: Instantiated from non-template code. "boost_1_32_0/boost/graph/depth_first_search.hpp", line 355: Error: Could not find a match for boost::vertices<boost::T>(const boost::adjacency_list<boost::vecS, boost::vecS, boost::undirectedS, boost::no_property, boost::no_property, boost::no_property, boost::listS>). big thanks for any help, Laurent

3 3