July 2009 - Boost-users - lists.boost.org

Status of Boost.Extension from sandbox?
by Rex Kerr 10 Sep '09

10 Sep '09

I'm evaluating various plugin solutions for a project and am considering boost.extension as one of the alternatives. Looking at http://boost-extension.blogspot.com/ and the mailing list archives it appears that the project hasn't been active for a while. I have tried, unsuccessfully, to find a page detailing the current status of sandbox projects to determine whether or not it has been accepted for inclusion in future boost versions, has significant flaws preventing its inclusion, etc. I'm hesitant to choose a project that may be orphaned. Any guidance in finding such status updates, or outright answers to my questions, would be helpful. Thank you, -Rex

4 6

Re: [Boost-users] [Graph] Problems with adjacency_matrix
by Juan Antonio Farré Basurte 29 Aug '09

29 Aug '09

> > > Hello, > I have a problem with adjacency_matrix template. > It claims to implement VertexAndEdgeListGraphConcept, BidirectionalGraphConcept and MutablePropertyGraphConcept. > But when I try to compile, it complains about using the next functions with an adjacency_matrix: > > num_vertices(graph) > vertices(graph) > num_edges(graph) > edges(graph) > source(edge_descriptor, graph) > target(edge_descriptor, > graph) > > and also with > > get(&V::property, graph), where V is a vetex bundled-properties type get(&E::property, > graph), where E is an edge bundled-properties type > > (these get functions should return a property map for the corresponding bundled property > and they work correctly for adjacency_list, just like the rest). > > Is this a bug or an error in the documentation? > > Thanks a lot, > > Juan > Mmmm... I just tried to include adjacency_matrix.hpp and now I don't get errors about those functions anymore. It's strange to have to inclue this header in a template library that doesn't explicitly deal with any concrete implementations of graph concepts. In fact, I don't need to include adjacency_list.hpp for it to work with adjacency_list. I just need to include graph_concepts.hpp and graph_traits.hpp. Anyway, I'm still getting lots of errors in a code that perfectly compiles and works with adjacency lists. If anybody is interested, I can send the listing of compiler errors. Thanks again, Juan

9 20

pkg-config and FreeBSD
by Roman Shmelev 29 Aug '09

29 Aug '09

Hi! I've downloaded boost 1.38 from the official site and successfully build it. I was able to compile the program that uses regexes. OS is FreeBSD 7.2. gcc is 4.2.1. The question is: how can I do something like this: pkg-config --libs boost-regex I'm new to FreeBSD, so I guess I really miss something basic. Googling for subject gave me a lot but not instructions how to do what I want. Thank you

3 10

Re: [Boost-users] [ANN] Boost 1.39.0
by Tom Kent 28 Aug '09

28 Aug '09

I've built windows libraries (vc7.1, vc8, vc9) off of this, if you want them, you can get them from my website: http://boost.teeks99.com Tom > Subject: [Boost-users] [ANN] Boost 1.39.0 > From: Beman Dawes (bdawes_at_[hidden]) > Date: 2009-05-02 16:57:10 > > Boost 1.39.0 is available for download from SourceForge. > > See > http://sourceforge.net/project/showfiles.php?group_id=7586&package_id=8041&… > > New Libraries: Signals2. > > Updated Libraries: Asio, Flyweight, Foreach, Hash, Interprocess, > Intrusive, Program.Options, Proto, PtrContainer, Range, Unordered, > Xpressive. > > Updated Tools: Boostbook, Quickbook. > > Plus the usual bug fixes to numerous libraries. > > Please report any problems to the Boost users or developers mailing > list. > > Thanks, > > --Beman Dawes, Daniel James, and all the other Boosters who help get > releases out

1 1

Race condition with Boost::Thread
by Rob Yull 27 Aug '09

27 Aug '09

I've been writing a thread pool class as kind of an experiment with boost.threads, boost.bind, and boost.function. The code below is the worker thread, that will eventually pull jobs from a queue, execute them, then look and see if there is another job in the queue. I think I've got an interesting race condition. The problem I'm having is when the class goes out of scope. I'm getting an access violation from the run() member function (which is function being run by m_Thread). Right now, run() is just entering and leaving (no actual queue checking). By the time run() is getting executed, the class is starting to execute the destructor. I thought that having the join() call in the destructor should allow the thread to finish before the class data is destroyed. In Visual Studio 2005's debugger, when it is getting to run(), it looks like the object has already been destroyed (this = 0x7fffffff), so the access violation is coming from the assignment to the m_Running member variable. If I take out the assignment, and run the debug, when it is entering run(), the this pointer is still showing as invalid, but the function completes and ends up back at the join() in the destructor, where the this pointer in the destructor is valid again. So as far as the destructor seems to be conserned, the class hasn't been destroyed yet, but as far as the member class is, it has been destroyed already. If I do pad some instructions/time before the class goes out of scope, run() executes just fine (this pointer is valid). I have also tried moving the creation of m_Thread outside of the constructor to a separate member function call, but the result is the same. I know debugging multithreaded apps can be difficult. Is there right out something I'm doing wrong here or making some kind of bad assumpion about the execution of run()? #include <boost/config.hpp> #include <boost/thread.hpp> #include <boost/functional.hpp> class WorkerThread { public: typedef boost::function<void ()> JobType; WorkerThread() { m_Thread = boost::thread(boost::bind(&WorkerThread::run, boost::ref(this))); m_Running = true; } ~WorkerThread() { m_Thread.join(); } const bool isRunning() const { return m_Running; } private: volatile bool m_Running; boost::thread m_Thread; void run() { m_Running = false; // Access violation here } private: WorkerThread(const WorkerThread& rhs); WorkerThread& operator = (const WorkerThread& rhs); }; int main() { WorkerThread wt; // new thread should start executing here return 0; } -Rob Yull

4 6

Random library, discrete distribution
by Kevin Martin 18 Aug '09

18 Aug '09

I made a post sometime last year about requiring a distribution that could generate enums with specified probabilities. I find I am regularly requiring a discrete distribution like this - where I specify the probabilities. Am I the only one who requires it, or would it be useful to a wider audience? I'm just wondering whether to submit a feature request for it or not? Thanks, Kevin Martin This is what I'm using at the moment: template<class IntType = int> class discrete_distribution { public: typedef boost::uniform_int<>::input_type input_type; typedef IntType result_type; template<size_t N> discrete_distribution(const boost::array<int, N> &outputs, const boost::array<int, N> &weights) : outputs_(outputs.begin(), outputs.end()), weights_(weights.begin(), weights.end()) { for(int i=1;i!=weights_.size();++i) weights_[i] += weights_[i-1]; } inline void reset() {} template<class Engine> result_type operator()(Engine& eng) { boost::uniform_int<> ui(0, weights_.back()-1); boost::variate_generator<Engine&, boost::uniform_int<> > vg(eng, ui); int idx = std::lower_bound(weights_.begin(), weights_.end(), vg()) - weights_.begin(); return outputs_[idx]; } private: std::vector<int> outputs_; std::vector<int> weights_; };

3 11

[random] Singleton engine
by Diederick C. Niehorster 17 Aug '09

17 Aug '09

Hi All, I have been writing a wrapper around the variate_generator as I always want to use the same number generator, but do want to use different distributions. As I want to use only one instance of the engine instance for all instanced of my random number generator class, I used a singleton pattern. as is my first time using it, I decided to test it generously, to try and see if the same instance of the enigine is used. So not only did i check if the engine pointer for different instances of my random class is the same, but i also checked if the destructor for the engine (does not) get called (i've added a constructor to the boost header file temporarily) and whether the random numbers generated are indeed not the same. It seems my code does not work, the destructor gets called weirdly often, e.g., 4 times every time i initialize my random class. Also, the numbers generated by the two generators are the same, indicating the engine is reinitialized for every instance of my random class. Yet, the pointer to the engine is the same (note that it is also the same if i dont call the destructor in between). I am using MSVC 2008 on windows XP 32 bit. What am I doing wrong? As I am learning, any other feedback is also very welcome. Best, Diederick ---- output of run of code below: check 0 00000000 ctor 1 destructor destructor destructor destructor after ctor 1 003462E0 random: 8148 dtor 1 destructor ctor 2 destructor destructor destructor destructor after ctor 2 003462E0 random: 8148 destructor destructor ---- code // includes #include <boost/random.hpp> #include <time.h> #include <iostream> using namespace std; // declaration template<class Dist> class CRandom { public: /** use for Dist: CRandomI Ci(min,max); uses uniform_int<int> - integral types CRandomR Cr(min,max); uses uniform_real<double> - floating point types for other ditributions or types, use: CRandom<distribution<optional type>> Cr(0--3 params, depending on distribution); for distributions and params, see: http://www.boost.org/doc/libs/1_39_0/libs/random/random-distributions.html */ // forwarding constructors explicit CRandom() : Gen(*EngineSingleton::Eng(), Dist()) { } template<class P> explicit CRandom(const P & p1) : Gen(*EngineSingleton::Eng(), Dist(p1)) { } template<class P> explicit CRandom(const P & p1, const P & p2) : Gen(*EngineSingleton::Eng(), Dist(p1, p2)) { } template<class P> explicit CRandom(const P & p1, const P & p2, const P & p3) : Gen(*EngineSingleton::Eng(), Dist(p1, p2, p3)) { } ~CRandom(void) {}; typedef typename Dist::result_type result_type; result_type operator()() { return Gen(); } template<class T> result_type operator()(T value) { return Gen(value); } private: boost::variate_generator<boost::mt19937,Dist> Gen; }; // end declaration // shorthand typedefs typedef CRandom<boost::uniform_int<int>> CRandomI; // even int is default type , specify it in case it ever changes typedef CRandom<boost::uniform_real<double>> CRandomR; // even double is default type, specify it in case it ever changes /** engine wrapper class following singleton pattern, so every instantiation of CRandom uses the same engine based on: http://www.aristeia.com/Papers/DDJ_Jul_Aug_2004_revised.pdf */ // declaration class EngineSingleton { public: static boost::mt19937* Eng(); static boost::mt19937* GetEngPoint() {return pEng;}; // for test only protected: // disallow (copy) construction and assignment operator EngineSingleton(); EngineSingleton(const EngineSingleton&); EngineSingleton& operator=(const EngineSingleton&); private: static boost::mt19937* pEng; }; // end declaration // implementation boost::mt19937* EngineSingleton::pEng = NULL; boost::mt19937* EngineSingleton::Eng() { if (pEng == NULL) { pEng = new boost::mt19937(/*unsigned int(time(NULL))*/); // initialize with same seed always, for test only } return pEng; } // end implementation //main int main(int argc, char* argv[]) { cout << "check 0" << endl; cout << EngineSingleton::GetEngPoint() << endl; cout << "ctor 1" << endl; CRandomI Cr(1,10000); cout << "after ctor 1" << endl; cout << EngineSingleton::GetEngPoint() << endl; int test = Cr(); cout << "random: " << test << endl; cout << "dtor 1" << endl; Cr.~CRandom(); cout << "ctor 2" << endl; CRandomI Cr2(1,10000); cout << "after ctor 2" << endl; cout << EngineSingleton::GetEngPoint() << endl; int test2 = Cr2(); cout << "random: " << test2 << endl; // exit return 1; }

10 22

property_tree CDATA
by Gordon Smith 15 Aug '09

15 Aug '09

property_tree is working well except that CDATA text is missing. The underlying XML library (rapidxml) contains CDATA after reading but it is not transferred to the property_tree structure. Is there a setting I've missed to include CDATA? If not, I've modified the rapidxml reader to transfer CDATA nodes so I already have a workaround. Is this an area that is being worked on, or is CDATA excluded for a reason?

2 1

Re: [Boost-users] [Accumulators] Questions about the library
by dhruva 10 Aug '09

10 Aug '09

Hello, > Date: Thu, 25 Jun 2009 03:41:49 -0700 (PDT) > From: adish > Subject: Re: [Boost-users] [Accumulators] Questions about the library > To: boost-users(a)lists.boost.org > > > > > > ? ?2. If my platform does not support HW floating point ops, and all my > > > ? ?sample values are, e.g. ints, is there a way to specify that the > > > *computation > > > ? ?*and *return value itself* (i.e. result_type, e.g. the mean) will be done > > > ? ?in ints (naturally, at the cost of precision)? > > > > I have used UDT as I needed some extra information from the results (as I was > working on data from a time series). The reverse, using a POD like int must be > possible. Instead of the first argument being float/double, use an int. If that > does not work, you can wrap an int in a UDT, implement some of the operators > like '=', '<', '/' depending on the accumulator you plan to use (I did that as I > needed to send a UDT). > > Using an int still returns a float -- which is quite sensible. You might be able to write a custom extractor which can return an int. I have not explored it. I always thought the type you give as input is the type that gets returned. > By UDT do you mean User Defined Type? > This is an interesting approach. > I didn't realize that the result_type depends on the value type in > that way, I didn't see any UDTs mentioned in the docs. Yes, I mean writing a class/struct. I had to overload '<', '>', '+=' and '/' and I use it for variance, mean, median, min, max and my implementation of percentile. Remember to specialize std::numeric_limits for the UDT you write. Else, the min and/or max will have values that you have not added in the accumulator (garbage values). I had a tough time debugging it and finding the solution. Also, on Windows, you need to derive your specialization from _Num_base. As for implementing a generic rolling type: 1. You need to store your values in a circular buffer. 2. Each time you add a value into the buffer, pass the contents to a pristine accumulator. This may be sub optimal but works. Especially for those calculations where it is not easy to remove the effect of the first data that you want to pop. I had to do it this way because I was dealing with time series. If I want a rolling computation for a 1 hour window, I might add 1 sample and may have to remove more than 1 sample to ensure first-last is 1 hour window in the buffer. -dhruva Love Cricket? Check out live scores, photos, video highlights and more. Click here http://cricket.yahoo.com

3 2

[Boost.Test] Controlling Windows debug trace output
by Alexander Lamaison 08 Aug '09

08 Aug '09

I've recently noticed the Boost.Test outputs anything output using the Windows debug trace functions such as ATLTRACE. Thanks for this! This makes these trace functions so much more useful. However, I was wondering if there was some way to control this output. Ideally, I don't want my automoated test runs to output all these tracelines. Is there a flag to the test runnner that will turn them of? Many thanks. Alex Lamaison

2 4