November 2003 - Boost-users - lists.preview.boost.org

Re: filesystem::directory_iterator::operator++() never returns <--never mind
by John.Wismar＠alldata.com 20 Nov '03

20 Nov '03

OK, I found it. Upon yet another search of the archive, I found it: I was using the single-threaded version of the library in a multi-threaded app. Sorry for the bandwidth! --------------------- John Wismar john.wismar(a)alldata.com In a previous state of enlightnment, I wrote: I'm using the fielsystem library's directory_iterator class to collect information about the files in a directory. I'm using VC7.0, and Boost 1.30.2. My logic looks something like this: ---------------------------------------------------------------------- class DirInfo { public: DirInfo(fs::path aPath) : myPath(aPath){} void scan(); void addFile(fs::path); private: fs::path myPath; } void DirInfo::scan () { fs::directory_iterator itEnd; for (fs::directory_iterator it(myPath); it != itEnd; ++it) { TRACE("File: %s\n", it->string().c_str()); if (!fs::is_directory(*it)) { addFile(*it); } } } DirInfo di(fs::path("C:\\temp\\", fs::native); di.scan(); ---------------------------------------------------------- This successfully goes through all of the files (and subdirectories) in the given directory, but after the last item, the next call to fs::directory_iterator::operator ++() never returns. If anyone has any suggestions or ideas, I would love to hear them!! Thanks! The call stack at the point where execution seems to lock up looks like this: ntdll.dll!NtDelayExecution() + 0xc kernel32.dll!_Sleep@4() + 0xb MyApp.exe.exe!boost::detail::lightweight_mutex::scoped_lock::scoped_lock(boost::detail::lightweight_mutex & m={...}) Line 96 + 0xa C++ MyApp.exe.exe!boost::detail::sp_counted_base::release() Line 132 + 0xf C++ MyApp.exe.exe!boost::detail::shared_count::~shared_count() Line 380 C++ MyApp.exe.exe!boost::shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>:: ~shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>() + 0x2e C++ MyApp.exe.exe!boost::shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>::reset () Line 217 C++ MyApp.exe.exe!boost::filesystem::directory_iterator::m_inc() Line 256 C++ > MyApp.exe.exe!boost::filesystem::directory_iterator::operator++() Line 88 + 0x2b C++ MyApp.exe.exe!DirInfo::scan() Line 63 + 0x1a C++ --------------------- John Wismar john.wismar(a)alldata.com

1 0

filesystem::directory_iterator::operator++() never returns
by John.Wismar＠alldata.com 19 Nov '03

19 Nov '03

I'm using the fielsystem library's directory_iterator class to collect information about the files in a directory. I'm using VC7.0, and Boost 1.30.2. My logic looks something like this: ---------------------------------------------------------------------- class DirInfo { public: DirInfo(fs::path aPath) : myPath(aPath){} void scan(); void addFile(fs::path); private: fs::path myPath; } void DirInfo::scan () { fs::directory_iterator itEnd; for (fs::directory_iterator it(myPath); it != itEnd; ++it) { TRACE("File: %s\n", it->string().c_str()); if (!fs::is_directory(*it)) { addFile(*it); } } } DirInfo di(fs::path("C:\\temp\\", fs::native); di.scan(); ---------------------------------------------------------- This successfully goes through all of the files (and subdirectories) in the given directory, but after the last item, the next call to fs::directory_iterator::operator ++() never returns. If anyone has any suggestions or ideas, I would love to hear them!! Thanks! The call stack at the point where execution seems to lock up looks like this: ntdll.dll!NtDelayExecution() + 0xc kernel32.dll!_Sleep@4() + 0xb MyApp.exe.exe!boost::detail::lightweight_mutex::scoped_lock::scoped_lock(boost::detail::lightweight_mutex & m={...}) Line 96 + 0xa C++ MyApp.exe.exe!boost::detail::sp_counted_base::release() Line 132 + 0xf C++ MyApp.exe.exe!boost::detail::shared_count::~shared_count() Line 380 C++ MyApp.exe.exe!boost::shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>:: ~shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>() + 0x2e C++ MyApp.exe.exe!boost::shared_ptr<boost::filesystem::directory_iterator::dir_itr_imp>::reset () Line 217 C++ MyApp.exe.exe!boost::filesystem::directory_iterator::m_inc() Line 256 C++ > MyApp.exe.exe!boost::filesystem::directory_iterator::operator++() Line 88 + 0x2b C++ MyApp.exe.exe!DirInfo::scan() Line 63 + 0x1a C++ --------------------- John Wismar john.wismar(a)alldata.com

1 0

regex static lib for x86 emulator
by Reuti＠sdc.canon.co.in 19 Nov '03

19 Nov '03

I can build static library for ARM processor. But when building for X86 emulator I am getting following error. LINK : fatal error LNK1104: cannot open file 'boost_regex_vc6_sss.lib' Please help. Thanks in advance. Reuti

1 0

gregorian: explicit date::date(date_int_type) ?
by Michael Haubenwallner 19 Nov '03

19 Nov '03

Hi, I'm preparing my company's step from C to C++, and when i use boost::date_time::*, there's a great danger to make mistakes like this with boost-1.30.2: #include <time.h> #include <iostream> #include <boost/date_time/posix_time/posix_time.hpp> using namespace boost::posix_time; using namespace boost::gregorian; int main() { time_t now = time(0); try { ptime time(now); // mistake cerr << to_simple_string(time) << endl; } catch (exception const& e) { cerr << "error: " << e.what() << endl; } } The problem is that the time_t in this case is treated as a day-count from which a gregorian::date() is implicitly constructed. When digging through the gregorian::date, i see that the constructors of date_time::date() from date_int_type and date_rep_type protected. So, wouldn't it be better to make the gregorian::date() constructors from date_int_type and date_rep_type explicit, to get a compile-time error with this sample rather than the out-of-range-exception "Year is out of valid range: 1400..10000" ? And another reason to make this constructor explicit might be that date_rep_type is the internal representation of date(), so when constructing date() with a day_count, you have to know the epoch, and then constructing date() should be done explicitly... What do you think about this ? Thanks, Michael Haubenwallner -- Michael Haubenwallner SALOMON Automation GmbH Forschung & Entwicklung A-8114 Friesach bei Graz T +43 3127 200 308 F +43 3127 200 22 mailto:michael.haubenwallner@salomon.at http://www.salomon.at

2 3

Invert Matrix
by JEFFREY CULLINA 18 Nov '03

18 Nov '03

I have devoleped code to derive a simple matrix expression: [K]*[D]=[R] wher D is an unknown vector. I am looking for code to do two things: 1. Invert [K] to get the expression [D]=inv([K])*[R] 2. Solve the equation in 1 It looks like the UBLAS toolkit provides a function _t_sv to solve the equation, but I do not see anything to help me invert the matrix. Is there code available to help me invert a matrix? While I am at it, I may as well ask if _t_sv is the best function to use to solve the system. Thanks, Jeff

1 0

Smart pointers and STL allocators
by Mark A. Gibbs 18 Nov '03

18 Nov '03

I have a couple of questions about combining smart pointers and STL allocators. First of all, am I right in assuming that I can use the shared_* templates right out of the box, as long as I define a custom deleter that uses the allocator's deallocate function? But if that's so, how exactly is it done? Part of my confusion comes from the fact that the STL allocator documentation I've come across isn't exactly clear on the semantics of allocators (I don't have a copy of the standard). For instance, I'm not 100% sure whether or not I have to call construct() after allocate() and destroy() before deallocate(). This is what I have so far: class allocator_helper { public: allocator_helper(allocator_type& a) : a_(a) {} void operator()(void const* p) const { a_.deallocate(static_cast<allocator_type::pointer>(p)); } private: allocator_type& a_; }; foo* bar = static_cast<foo*>(get_allocator().allocate(1)); shared_ptr<foo> p_foo(bar, allocator_helper(get_allocator())); foo* baz = static_cast<foo*>(get_allocator().allocate(10)); shared_array<foo> p_foo(baz, allocator_helper(get_allocator())); Now, for scoped_* I would probably have to rewrite them to use the allocator, right? Or is there a way to use allocators with the existing scoped_* templates? Mark

2 3

regex problems
by Paul Scruby 18 Nov '03

18 Nov '03

The boost::regex class is crashing when I construct it with some regular expressions. For example: "(value|default)[^}]*\{" will crash and "value|default" does not. Is this a bug, or am I doing something wrong? I am running gcc2.96 and boost1.30.2. I have attached an example src including Makefile. Thanks in advance, Paul

3 3

Newbie question, linking problems
by Mrk 17 Nov '03

17 Nov '03

Hello, I'm trying to get regexp to work under vc6, STL 4.5.3 and boost 1.30.2 but sadly I'm running into some problems. Did anyone ever had this problem with the following methods ? error LNK2001: unresolved external symbol "public: static void __cdecl boost::w32_regex_traits<char>::transform_primary( ... error LNK2001: unresolved external symbol "public: static void __cdecl boost::w32_regex_traits<char>::transform( ... Thanks, Mrk

1 0

Re: [Boost-users] [BGL] Problem with color maps for undirected_dfs_visit
by Jeremy Siek 17 Nov '03

17 Nov '03

Hi Tarjei, On Friday, November 14, 2003, at 05:33 AM, Tarjei Knapstad wrote: > On Thu, 2003-11-13 at 17:09, Jeremy Siek wrote: >> Hi Tarjei, >> >> You need to initialize the edge index property map (only the vertex >> index >> map is automatically initialized, and then only for VertexList=vecS). >> > > Hmm, ok, but I still don't quite understand. I thought that > initializing > the std::vector holding the colors was sufficient...? To my > understanding an iterator_property_map is just a property map which > iterates through some other container (that can provide a > RandomAccessIterator) actually holding the properties? Yes, your edge color map is initialized just fine, but you also need to initialize the edge index map. Here's the reason why: your edge color map uses the edge index map to go from an edge descriptor to an integer offset, which is then used to access into the color vector. Currently, your edge index map is initialized to random garbage, so you are going out-of-bounds when accessing the edge color vector. See the example file exterior_properties.cpp, BTW, you don't need to declare an edge_color property for the adjacency_list because you are using an external edge color map. > > Obviously there's some connection I don't quite get here... I also > don't > quite understand the EColorValue uv_color = get(edge_color, *ei); call > in undir_dfv_impl. It asks for the edge_color at the edge_descriptor > *ei, but from where? > > Anyway, I don't really even need an edge_index property here (or do > I?), > it was more or less to try to get this working. Yes, you need it for the reason given above. BTW, if you switch to using an internal edge color property, then you won't need the edge index property, or the edge color vector. Cheers, Jeremy

2 1

GRAPHVIZ_DIRECTED macro definition?
by Soumitra Chatterjee 17 Nov '03

17 Nov '03

Hi, In the file .../libs/graph/src/graphviz_parser.y we have the following: #ifndef GRAPHVIZ_DIRECTED #error Need to define the GRAPHVIZ_DIRECTED macro to either 0 or 1 #endif I couldn't locate the definition for the above macro in my boost installation. Do we need to define it explicitly, or am I missing some file(s)? Thanks, Soumitra

2 1