February 2009 - Boost-users - lists.preview.boost.org

[boost::filesystem] How to determine access mode?
by Dmitry V. Krivenok 19 Feb '09

19 Feb '09

Hello! Is there a way to determine access mode for arbitrary file using boost::filesystem? I searched through the documentation and source code, but didn't find such functions in boost::filesystem. I'm looking for simple functions like: bool is_readable(const std::string& path) bool is_writable(const std::string& path) bool is_executable(const std::string& path) I also want to determine OS-dependent access rights flags (such as setuid and setgid in Unix). Is it possible with boost::filesystem? I don't want to use low-level stat() function and check for S_ISUID and S_ISGID. Thank you beforehand! -- Sincerely yours, Dmitry V. Krivenok Orange System Co., Ltd. Saint-Petersburg, Russia work phone: +7 812 332-32-40 cellular phone: +7 921 576-70-91 e-mail: krivenok(a)orangesystem.ru web: http://www.orangesystem.ru skype: krivenok_dmitry jabber: krivenok_dmitry(a)jabber.ru icq: 242-526-443

1 0

Re: [Boost-users] extraction invariants from intensive-loop based on template parameters
by Hicham Mouline 19 Feb '09

19 Feb '09

> ----- Original Message ----- > From: "Matthieu Brucher" <matthieu.brucher(a)gmail.com> > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] extraction invariants from intensive-loop based on template parameters > Date: Thu, 19 Feb 2009 14:17:20 +0100 > > > Hi, > > The compiler is smart enough to prune an if condition if it knows that > the value is always false. > So if you add an if(!Lower::IsConstant) before and after your loop, > you don't have to rewrite it. > > Matthieu > > 2009/2/19 Hicham Mouline <hicham(a)mouline.org>: > > Hello, > > > > I have an algorithm like: > > > > template <typename Upper, typename Lower> > > class X { > > > > static void f() > > { > > .... > > for (size_t n=0; n<=N; ++n) > > { > > // process for j=0 at every n if Lower::IsConstant is false > > for (size_t j=1; j<=J-1; ++j) > > { > > } > > // process for j=J at every n if Upper::IsConstant is false > > } > > .... > > } > > > > > > So I extracted the double for loop in a separate wrapper class and > > I verbosely wrote 4 specializations with template args begin bool : > > > > <true, true> > > <false, true> > > <true, false> > > <false, false> > > > > for e.g. for <true, true> > > > > // process for j=0 once > > for (size_t n=0; n<=N; ++n) > > { > > for (size_t j=1; j<=J-1; ++j) > > { > > } > > } > > // process for j=J once > > > > > > Any advice? > > Is there a better way than writing explicitly the tmpl specializations? > > Boost.PP maybe? > > > > rds, > > _______________________________________________ I've just measured with VS2005, g++4.0 and intel11. At least these 3 compilers are smart enough. Thanks,

1 0

[filesystem] how to get symlink info?
by jeti 19 Feb '09

19 Feb '09

hi all im not sure if i just dont find the solution in the docs/list or its not supported in boost::filesystem which means using system functions given is a symlink like this '/tmp/subdir/thisisasymlink' which points to '/some/where/some/file/or/directory' what i need is: 1. the absolute path the symlink is pointing too, which is '/some/where/some/file/or/directory' 2. get the absolute path for something like '/tmp/../some/where/some/file/or/directory' which resolvs to '/some/where/some/file/or/directory' 3. if the absolute path is a subdirectory of some other directory in this example: if '/some/where/some/file/or/directory' is a subdirectory of '/tmp/subdir' which obviously isnt thx in advance

1 0

extraction invariants from intensive-loop based on template parameters
by Hicham Mouline 19 Feb '09

19 Feb '09

Hello, I have an algorithm like: template <typename Upper, typename Lower> class X { static void f() { .... for (size_t n=0; n<=N; ++n) { // process for j=0 at every n if Lower::IsConstant is false for (size_t j=1; j<=J-1; ++j) { } // process for j=J at every n if Upper::IsConstant is false } .... } So I extracted the double for loop in a separate wrapper class and I verbosely wrote 4 specializations with template args begin bool : <true, true> <false, true> <true, false> <false, false> for e.g. for <true, true> // process for j=0 once for (size_t n=0; n<=N; ++n) { for (size_t j=1; j<=J-1; ++j) { } } // process for j=J once Any advice? Is there a better way than writing explicitly the tmpl specializations? Boost.PP maybe? rds,

2 1

compiling 32 bit boost libraries on a 64 bit system? (gcc, Linux)
by Ákos Maróy 19 Feb '09

19 Feb '09

Hi, I'm trying to compile boost into 32 bit libraries on a 64 bit Linux box. with gcc, with would simply mean to add the flag -m32 to gcc. but I can't seem to make bjam do this. I've tried a number of approaches: - export and set the environment variables CFLAGS, CXXFLAGS and LDFLAGS to -m32 - set the following in tools/build/v2/user-config.jam: cflags=-m32 cxxflags=-m32 linkflags=-m32 - set the following in tools/build/v2/user-config.jam: using gcc : 4.1.2 : g++-4.1.2 : -m32 ; (based on the documentation found at http://beta.boost.org/doc/tools/build/doc/html/bbv2/reference/tools.html#bb… ) but to no avail - the -m32 flag is not honored, and the compiled boost libraries are 64 bit :( what am I doing wrong? Akos

2 2

A need for a complex pool type container
by Robert Dailey 19 Feb '09

19 Feb '09

My requirements are a bit complex and involved, but I'll try to be as brief as possible. We have this "scheduler" type system where for each unit of time (think of 1 unit as 1 ms) there is a "bucket" of events that shall be processed. This is relative. Suppose the current time is 103. We will have a record of events to be processed at time 103, 104, 105, ..., 203. When all of the events in the "bucket" for time stamp 103 are processed, we "pop_front" to remove the bucket for 103. Everything is essentially shifted forward. Hopefully this does not require physically shifting everything forward, and simply is just a matter of using a lesser index to access the same element later on. I insert new events into existing buckets anywhere in the container, it does not necessarily have to be from the back. I always remove items from the front since they are processed linearly. One can kind of think of this as a std::deque< std::vector<> >, but the containing deque would need to be given a fixed size of 100 (as in the example above). The tricky part is shifting all of the "buckets" (the buckets in this case would be the inner-vector objects) forward when something from the front is removed. The key here is that I do not want to call pop_front() each time a bucket is processed because I want to avoid constant new/delete operations since insertion and removal will be happening quite frequently. I'm not sure what the best approach to solving this problem is. I'm hoping someone in the community will be able to suggest a way in which boost can easily solve this problem. Thanks to everyone for taking the time to read this.

2 6

[serialization] problem exporting derived classes
by Jens Weggemann 19 Feb '09

19 Feb '09

Hello, I've been trying out boost::serialization and ran into a problem trying to serialize objects of derived classes. I set up this simple class hierarchy: class BaseNode { std::vector<BaseNode*> children_; ... friend class boost::serialization::access; template<class Archive> void serialize(Archive &ar, unsigned version) { ar & children_; } }; class DerivedNode : public BaseNode { ... friend class boost::serialization::access; template<class Archive> void serialize(Archive &ar, unsigned version) { ar & boost::serialization::base_object<BaseNode>(*this); } }; All works fine as long as I register the types with the archives: arch.register_type(static_cast<BaseNode*>(NULL)); arch.register_type(static_cast<DerivedNode*>(NULL)); Now the documentation suggests exporting the classes as an alternative to that. But when I instead put BOOST_CLASS_EXPORT(BaseNode) BOOST_CLASS_EXPORT(DerivedNode) the polymorphism is not handled correctly and e.g. a tree of 3 DerivedNodes is restored as a single BaseNode object. What am I missing here? The full code for this test is at http://noonti.de/tree_serialization.cpp.txt Thanks, Jens

2 1

Graph lib: read_dimacs_max_flow segment fault
by Xian Chen 18 Feb '09

18 Feb '09

Hi All, I compile read_dimacs_max_flow example ( http://www.boost.org/doc/libs/1_38_0/libs/graph/example/read_write_dimacs-e…) with these warnnings. g++ -I/home/xianchen/opt/include/boost-1_38 -Wno-deprecated -c read.cpp -o objects/simu.o /home/xianchen/opt/include/boost-1_38/boost/graph/read_dimacs.hpp: In function 'int boost::read_dimacs_max_flow(Graph&, CapacityMap, ReverseEdgeMap, typename boost::graph_traits<G>::vertex_descriptor&, typename boost::graph_traits<G>::vertex_descriptor&, std::istream&) [with Graph = main()::Graph, CapacityMap = boost::adj_list_edge_property_map<boost::directed_tag, long int, long int&, unsigned int, boost::property<boost::edge_capacity_t, long int, boost::property<boost::edge_reverse_t, boost::detail::edge_desc_impl<boost::directed_tag, unsigned int>, boost::no_property> >, boost::edge_capacity_t>, ReverseEdgeMap = boost::adj_list_edge_property_map<boost::directed_tag, boost::detail::edge_desc_impl<boost::directed_tag, unsigned int>, boost::detail::edge_desc_impl<boost::directed_tag, unsigned int>&, unsigned int, boost::property<boost::edge_capacity_t, long int, boost::property<boost::edge_reverse_t, boost::detail::edge_desc_impl<boost::directed_tag, unsigned int>, boost::no_property> >, boost::edge_reverse_t>]': read.cpp:86: instantiated from here /home/xianchen/opt/include/boost-1_38/boost/graph/read_dimacs.hpp:108: warning: deprecated conversion from string constant to 'char*' /home/xianchen/opt/include/boost-1_38/boost/graph/read_dimacs.hpp:108: warning: deprecated conversion from string constant to 'char*' ...... My GCC is: Using built-in specs. Target: i686-pc-linux-gnu Configured with: ../gcc-4.2.4/configure --prefix=/home/xianchen/opt --enable-shared --enable-threads=posix --enable-languages=c,c++ --enable-checking=release --enable-mpfr=/home/xianchen/opt/ --with-system-zlib Thread model: posix gcc version 4.2.4 I run this program using data "boost/libs/graph/example/max_flow.dat" But get segment fault: *** glibc detected *** free(): invalid pointer: 0x08053ff4 *** Aborted Can any one tell me the why this happen? The program aborts under GCC.4.3.3 either. Which version of GCC is the most boost friendly? Thanks in advance, Xian

1 0

[serialization] export.hpp changes from 1.37 to 1.38
by Chard 18 Feb '09

18 Feb '09

The BOOST_CLASS_EXPORT.. macros have been moved into a detail namespace (1.38) from a (public) anonymous namespace (1.37). I'm guessing that the reasoning for this was in order to avoid many definitions of the class singletons. However, this appears to have enforced that all of the guids (for all archive types) now have to reside in a single compilation unit. The attached code illustrates the limitation; this compiles under 1.37 but results in duplicate symbol definitions in 1.38 (with MSVC7.1). It worked with 1.37 because the anonymous namespace ensured no symbol clashes, but now the symbols are all in the same namespace. The example shows a case where two archive types (xml and binary) might be in use. The obvious solution to this (simple) case would be to just use a single compilation unit, which brings in the guids for both binary and xml. However, turning a blind eye to the 'untidiness', this fix does not help with the project we're working on. In our project, there is not just a simple type being serialized; there are a lot more classes and a bit of mpl thrown in. This actually resulted in breaching the COFF symbol limit (approx. 65k) with MSVC7.1, so we were forced to split the serialization. In fact, we had to split both the save and load parts of the binary serialization. Our current workaround is to use the archive.register_type<...> method, and ignore the export macros, but this does mean ensuring all types are covered (as described in the documentation).

4 9

Boost and Real Time
by Brian Wrenn 18 Feb '09

18 Feb '09

Hi, I'm an experienced C++ programmer, but very new to Boost. I'm looking into making use of it for my next project, which will run on LynxOS, a real time system. Does anyone have any experience, knowledge of pitfalls, or words of wisdom to share about building and running with Boost, or more likely a subset of Boost's libraries, for a real time system? Thanks for any help. -Brian

2 1