- Boost-users - lists.boost.org

[serialization] transparently treating a class like a built-in type
by Kenneth Riddile 28 Jul '10

28 Jul '10

I have some xml archives that store a lot of floating point values. I would like to read these into a program that is designed to easily switch between using floats and a custom fixed-point type via a typedef. The way that would be easiest would be to add a serialize() method to the fixed-point class definition. However, the xml archive requires that I use an nvp wrapper. Since the data was originally serialized out as plain floats, the data looks something like this: <x>0.95260203</x> <y>0.87517798</y> <z>0.54323343</z> whereas, to serialize this into my fixed-point type, it would need another level of indirection due to the additional required nvp wrapper like so: <x><value>0.95260203</value></x> <y><value>0.87517798</value></y> <z><value>0.54323343</value></z> Is there a way to accomplish what I'm trying to do? The fixed-point type is convertible from (but not to) a float if that helps.

3 4

assigning a value to a matrix shared_ptr
by Max S. Kaznady 28 Jul '10

28 Jul '10

Hello, I am trying to assign a value to the matrix, which is pointed to by the shared pointer, but I am not sure which operator to call, please have a look at the following code, I commented the problem: #include <boost/numeric/ublas/matrix.hpp> #include <boost/numeric/ublas/io.hpp> int main () { using namespace boost::numeric::ublas; matrix<double> m (3, 3); boost::shared_ptr<matrix<double>> m2(new matrix<double>(3,3)); for (unsigned i = 0; i < m.size1 (); ++ i) for (unsigned j = 0; j < m.size2 (); ++ j) { m (i, j) = 3 * i + j; /// now do the same operation to the shared pointer's matrix m2->operator*(i, j) = 3 * i + j; // this results in a runtime error } std::cout << m << std::endl; // And how do we print m2 ? } Thanks, Max

1 2

Using boost::thread to keep all cores busy
by Internet Retard 27 Jul '10

27 Jul '10

I have a new 2 core CPU and I've been experimenting with boost::threads by attempting to keep both cores busy independent of each other. Say I have a queue of 4 tasks to do. Each task will be handled by one thread. Some tasks will take longer to complete than others so I can't use thread_group with 2 threads at a time and join_all() as one core may finish before the other and set idle until the second thread completes. I use boost::thread::hardware_concurrency() to determine how many CPU cores I have available. My problem is *knowing* when a thread/CPU core finishes a task and is available for another task. Here is some sample code. I commented what I think would be an approach. Any suggestions or hints? std::queue<std::string> string_queue() { std::queue<std::string> sq; sq.push("AAA"); sq.push("BBB"); sq.push("CCC"); sq.push("DDD"); return sq; } void worker(std::string& S) { std::cout << S << std::endl; } int main() { std::cout << "Number of Cores: " << boost::thread::hardware_concurrency() << std::endl; std::queue<std::string> strings = string_queue(); while ( !strings.empty() ) { // if ( a CPU core is available ) // { boost::thread t(boost::bind(&worker, strings.front())); t.join(); strings.pop(); // } // else // { // wait awhile and check if a CPU core is available again. // } } return 0; } _________________________________________________________________ Hotmail: Trusted email with Microsoft’s powerful SPAM protection. https://signup.live.com/signup.aspx?id=60969

4 10

subcommands with program_options (like svn command)
by Adrien Guillon 27 Jul '10

27 Jul '10

Hello all, I am writing a moderately complex command-line application, and I have been looking at program_options for simplifying my life. The best way to describe what I want to do, is in term of the svn command (which I assume most people are familiar with). In particular, what I want to do is expose a number of subcommands for controlling the application, each of which has their own particular options. For example, if I were to write svn using program_options, then the first command I have to accept could be one of: add, checkout, del, log, etc. Each of these commands has a separate list of arguments that they accept themselves. I'm a bit stumped on how exactly to do this with program_options, because the first argument (which I can do using positional arguments) decides what other arguments are valid. Any thoughts or suggestions? Thanks! AJ

1 0

enumeration type in boost
by archie14 27 Jul '10

27 Jul '10

Is there any support for proper usage of enumerations in any of the boost libraries? I am looking for something similar to Qt's QFlags class but can't find anything similar in boost.

1 0

[Intrusive] problem with inheritance
by Stephen Schleimer 27 Jul '10

27 Jul '10

Folk, I have an object which serves as a header to a "page" of memory which I serialize by simply writing out. I deserialize by reading the page and then doing a bit of fixup. I do this for performance (no copies other than the I/O in both directions). The header has an intrusive list pointer. When I reload the page, I must clear the header from the intrusive list. The page has been relocated (it may not be in the same physical location as it was when it was written). Thus, a simple unlink does not work. I was unable to figure out how to re-init the pointer when I had a base_hook. So, I switched to a member hook supposing that I could simply clear the contents of the hook (violating the abstraction) on the reload. Unfortunately, the headers are members of a hierarchy. The member_hook is defined in the base class and it is used in a subclass. I abstracted the problem into the following code: #include <boost/intrusive/list.hpp> class Foo { public: boost::intrusive::list_member_hook<> hook_; }; class Bar : public Foo { public: int x; }; class BigMess { typedef boost::intrusive::member_hook<Bar, boost::intrusive::list_member_hook<>, &Bar::hook_> mho; typedef boost::intrusive::list<Bar, mho> BarList; BarList list; }; This fails to compile using g++ (GCC) 4.4.0 on centos 2.6.18-128.1.16.el5 with: foo.cpp:15: error: could not convert template argument ‘&Foo::hook_’ to ‘boost::intrusive::list_member_hook<> Bar::*’ Clearly, the type of the base class is taken to not be compatible with the subclass. It is not clear why, however. Any thoughts would be appreciated as to how to either fix the inheritance problem or how to clear the base_hook case (which does compile and runs, by the way :-) Thanks -steve

1 0

Intrusive feature idea - accessor_hook
by Hite, Christopher 27 Jul '10

27 Jul '10

Intrusive has base_hooks and member_hooks. I wish it had a accessor hook, where you can define a member function which returns the a hook reference. My use case: I have a class being used between two threads and I want to break it up to avoid cache thrashing. struct Object{ struct UpdatedState{ } *updatedState; struct PrivateState{ } *privateState; struct PublishedState{ } *publishedState; }; So the users's thread updates stuff in UpdatedState and flags it for consumption by the service thread. The service thread has lots of private data associated with the object. It updates the user asychronously updating publishedState and flagging it for consumption by the user. The a nice way to flag things for consumption is to put them in an slist. The problem is that everyone wants an slist of Object instead of just a child like UpdatedState. The problem of course is that putting the hooks in Object would cause the cache trashing we were trying to avoid. Putting the hooks in the child objects requires a way back to Object. As work around a pointer could be added for each child back to Object. If I were writing my own linked lists, the solution is obvoius: I would add a field Object* nextUpdated to UpdatedState. What I propose is adding a accessor to Object list_member_hook<>& getUpdatedHook(); typedef accessor_hook<Object, list_member_hook<>, &Object::getUpdatedHook> MemberHookOption; Another advantage to allowing an accessor to be used is that the user can share memory between unused hooks by putting them in variant or union. In my case I've got 6 sorts 2 for add/change/delete prioritization. I only ever use 2 of these at a time, but I can't avoid the memory cost for having all the hooks. Chris Hite

2 1

Failure to load boost libraries on Linux
by David Aldrich 27 Jul '10

27 Jul '10

Hi Please will some one help me with what is probably a simple Linux error? I have installed Boost 1.40 on Centos 5.3 using an rpm from blackopsoft.com. When I run my executable I get: Failed to load the following library files: [snip] dlerror = libboost_python.so.2: cannot open shared object file: No such file or directory Executing: /sbin/ldconfig -p | grep boost gives: libboost_python-mt.so.1.40.0 (libc6,x86-64) => /usr/lib64/libboost_python-mt.so.1.40.0 libboost_python-mt.so (libc6,x86-64) => /usr/lib64/libboost_python-mt.so libboost_python-mt-d.so.1.40.0 (libc6,x86-64) => /usr/lib64/libboost_python-mt-d.so.1.40.0 libboost_python-mt-d.so (libc6,x86-64) => /usr/lib64/libboost_python-mt-d.so Please can anyone suggest what is wrong? Best regards David

4 5

[bind][function] Do copy constructors invoked inside boost::bind / boost::function get optimized out?
by Steve Lorimer 27 Jul '10

27 Jul '10

Hi I'm using boost::bind to convert a function of the form: T::fun(boost::intrusive_ptr<U>) into a form which can be stored by boost::function. Outputting calls to intrusive_ptr_add_ref and intrusive_ptr_release reveals there are a massive number of copies of the intrusive_ptr. For a single call of boost::bind on a I'm wondering what kind of overhead this will add if I'm calling boost::bind on boost::intrusive_ptr a LOT of times? Will the copies be optimized out? There are 6 calls to intrusive_ptr_add_ref when calling boost::bind to bind a member function pointer to an object which is wrapped up in an intrusive_ptr There are 5 calls to intrusive_ptr_add_ref when calling boost::bind to bind an argument which is an intrusive_ptr There are 12 calls to intrusive_ptr_add_ref when using boost::function to store the result of boost::bind when called to bind a member function pointer to an object which is wrapped up in an intrusive_ptr There are 11 calls to intrusive_ptr_add_ref when using boost::function to store the result of boost::bind when called to bind an argument which is an intrusive_ptr Example code and output showing the calls below: #include <iostream> #include <boost/intrusive_ptr.hpp> #include <boost/function.hpp> #include <boost/bind.hpp> using namespace std; struct ref_ctd { int refs; ref_ctd() : refs(0) { } }; template<typename T> void intrusive_ptr_add_ref(T* p) { cout << "++ setting p (" << p << ")'s refs to " << p->refs + 1 << endl; ++p->refs; } template<typename T> void intrusive_ptr_release(T* p) { cout << "-- setting p (" << p << ")'s refs to " << p->refs - 1 << endl; if (!--p->refs) delete p; } struct foo : public ref_ctd { void func(char a) { cout << a << endl; } }; void bar(boost::intrusive_ptr<foo> f) { f->func('b'); } int main() { cout << "--------- creating intrusive ptr ----------" << endl; boost::intrusive_ptr<foo> f(new foo); cout << "--------- binding intrusive_ptr object ----------" << endl; boost::bind(&foo::func, f, 'a'); cout << "--------- now storing result of bind ----------" << endl; boost::function<void()> func1 = boost::bind(&foo::func, f, 'a'); cout << "--------- binding intrusive_ptr object ----------" << endl; boost::bind(bar, f); cout << "--------- now storing result of bind ----------" << endl; boost::function<void()> func2 = boost::bind(bar, f); return 0; } Here are the results: --------- creating intrusive ptr ---------- ++ setting p (0x33b18)'s refs to 1 --------- binding intrusive_ptr object ---------- ++ setting p (0x33b18)'s refs to 2 ++ setting p (0x33b18)'s refs to 3 ++ setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 -- setting p (0x33b18)'s refs to 3 -- setting p (0x33b18)'s refs to 2 -- setting p (0x33b18)'s refs to 1 --------- now storing result of bind ---------- ++ setting p (0x33b18)'s refs to 2 ++ setting p (0x33b18)'s refs to 3 ++ setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 -- setting p (0x33b18)'s refs to 3 ++ setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 ++ setting p (0x33b18)'s refs to 7 ++ setting p (0x33b18)'s refs to 8 ++ setting p (0x33b18)'s refs to 9 -- setting p (0x33b18)'s refs to 8 -- setting p (0x33b18)'s refs to 7 -- setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 -- setting p (0x33b18)'s refs to 3 -- setting p (0x33b18)'s refs to 2 --------- binding intrusive_ptr object ---------- ++ setting p (0x33b18)'s refs to 3 ++ setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 -- setting p (0x33b18)'s refs to 3 -- setting p (0x33b18)'s refs to 2 --------- now storing result of bind ---------- ++ setting p (0x33b18)'s refs to 3 ++ setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 ++ setting p (0x33b18)'s refs to 5 ++ setting p (0x33b18)'s refs to 6 ++ setting p (0x33b18)'s refs to 7 ++ setting p (0x33b18)'s refs to 8 ++ setting p (0x33b18)'s refs to 9 ++ setting p (0x33b18)'s refs to 10 -- setting p (0x33b18)'s refs to 9 -- setting p (0x33b18)'s refs to 8 -- setting p (0x33b18)'s refs to 7 -- setting p (0x33b18)'s refs to 6 -- setting p (0x33b18)'s refs to 5 -- setting p (0x33b18)'s refs to 4 -- setting p (0x33b18)'s refs to 3 -- setting p (0x33b18)'s refs to 2 -- setting p (0x33b18)'s refs to 1 -- setting p (0x33b18)'s refs to 0

4 6

BOOST_STATIC_ASSERT has different behavior on different platforms
by Alexander Vinokur 27 Jul '10

27 Jul '10

Hi, BOOST_STATIC_ASSERT has different behavior on different platforms Boost 1.34 -------- File foo01. cpp -------- #include <boost/static_assert.hpp> struct Foo { int m_value; Foo(); }; Foo::Foo() { BOOST_STATIC_ASSERT(sizeof(this->m_value) == 4); } int main() { Foo f; return 0; } --------------------------------- No compilation errors on UNIX compilers HP: aCC: HP C/aC++ B3910B A.06.25.01 [May 16 2010] Sun: C++ 5.9 SunOS_sparc Patch 124863-01 2007/07/25 AIX: IBM XL C/C++ Enterprise Edition V8.0 for AIX; Version: 08.00.0000.0014 Linux: Intel(R) C++ Intel(R) 64 Compiler, Version 11.0 There are compilation errors on Windows compiler Microsoft (R) 32-bit C/C++ Optimizing Compiler Version 14.00.50727.762 for 80x86 foo01.cpp foo01.cpp(11) : error C2355: 'this' : can only be referenced inside non-static member functions foo01.cpp(11) : error C2227: left of '->m_value' must point to class/struct/union/generic type foo01.cpp(11) : error C2027: use of undefined type 'boost::STATIC_ASSERTION_FAILURE' C:\ADJUST\3rdParties\boost_1_34_0\boost/static_assert.hpp(34) : see declaration of 'boost::STATIC_ASSERTION_FAILURE' Regards, Alex Vinokur This message and the information contained herein is proprietary and confidential and subject to the Amdocs policy statement, you may review at http://www.amdocs.com/email_disclaimer.asp

2 1