- Boost-users - lists.boost.org

Property Tree - adding child nodes to a parent
by Jon Shuler 30 Jan '10

30 Jan '10

I am trying to build a property tree from hierarchy data in a database. I want to be able to add a child at a given node using something similar to: parentNode.add_child("child name", data_type). There is a add_child method but it expects the path to add the child at. I looked at the examples and did not see anything. Is this possible? Jon This e-mail message and any attachment(s) transmitted with it are intended only for the use of the recipient(s) named above.This message may be privileged and/or confidential. If you are not an intended recipient, you may not review, copy or distribute this message. If you have received this communication in error, please notify us immediately by e-mail and delete the original message.

2 1

Re: [Boost-users] Detect specialized template function?
by Andrew Chinkoff 30 Jan '10

30 Jan '10

Hi, Daniel. This is your example: template <class T> void S::f() { if(g is specialized) // (1) callbacks_ready=true; // else wait for a callback object to be registered } At (1) the runtime detection is used, but "S:g()" was specialized at compile time. Templates are specialized and compiled at compile time, but not runtime. We can't write "if(g is specialized)" - it is impossible to realize it at runtime (and it is senseless). So let me to recommend you some books of H. Sutter and A. Alexandrescu. There are a lot of recommendations about templates, their specializations, overloads, static/dynamic polymorphisms and more. 1st book: [http://www.gotw.ca/publications/xc++s.htm] Pay attention to this chapter of the book: "Item 7: Why Not Specialize Function Templates?" 2nd book; "C++ Coding Standards: 101 Rules, Guidelines, and Best Practices" (authors H. Sutter and A. Alexandrescu) Pay attention to 64, 65, 66 and 67 chapters. I hope you will find some useful ideas about templates in these books. There are also Boost libraries targeted to realize static polymorphism: 1) http://www.boost.org/doc/libs/1_40_0?view=category_Generic 2) http://www.boost.org/doc/libs/1_40_0?view=category_Metaprogramming Best, Andrew. > Slightly OT, but I was hoping someone here might know the magic incantation > (which may involve a Boost library). > Given a template class such as > template <class T> > struct S > { > void f(); > void g() { T t(h()); cb(t) } > T & h(); > boost::function1<void, T &> cb; > }; > Is there a way to have S.f() behave differently depending on whether S.g() uses > the default definition or whether it is specialized for this particular type? > This detection needs to occur before S.g() is ever called. > For example, the mere existance of the specialization > template <> void S<int>::g() { ... } > should trigger special logic in S<int>::f(). > Thanks, > Daniel > P.S. The motivation for this is an attempt to retrofit a callback API that > allows specialization of S.g() or the registration of a boost::function that > gets called by the default S.g(). We need to disable callbacks until an actual > function is registered via either method. It is easy to dispatch on > registration of the boost::function; I don't know how to detect the > specialization.

1 0

Boost 1.36.0 with Linux 64bits
by JongKwan Kim 29 Jan '10

29 Jan '10

Hi all, I would like to know if the boost 1.36.0 is working well with Linux 64 bits. If I want to compile with boost 1.36.0 and Linux 64 bits, how should I install the boost 1.36.0? Is it the same to 32 bits machine? Please, if someone knows that, let me know. Thanks,

5 8

fusion sequences, type erasure across translation units
by Hicham Mouline 29 Jan '10

29 Jan '10

Hello, I have a .cpp file which a sort of top level driver. I have say 100 different struct param_i {...}; where 1<=i<=100 Each param_i is defined in a param_i.hpp and param_i.cpp invokes metafunctions to obtain typedef fusion< variant< vector<double>, triplet<double>, pair<double>, double>, variant<...>, ..., variant<...> > seq_i_t; this sequence type is visible only in the param_i.cpp, not the header file. Generation of this sequence is The size of these sequences = number of members of param_i. The toplevel driver and other .cpp need to move instances of these types around to other .cpp. To avoid having the compiler call the metafunctions for these middle translation units, but only at the generating .cpp and at the consuming .cpp, I created a tag struct fusion_seq_tag {}; typedef <typename Arg> struct make_fusion_seq_i { typedef <undefined> type; }; typedef <typename Arg> struct make_fusion_seq_tag { typedef typename make_fusion_seq_i<Arg>::type base_type; struct type : fusion_seq_tag, base_type { type(...) : base_type(...) {} }; }; return smart_ptr<fusion_seq_tag>( new make_fusion_seq_tag<...>::type ); // this was the generating .cpp The function returns smart_ptr<fusion_seq_tag> which can move around in other .cpp without any metaprogramming until reaching the consuming .cpp Is this type erasure? I don't know how I can get the actual type back in the consuming .cpp if I don't know the originator but if I know only the shape of the fusion sequence. How to dynamic cast the type back to its unerased type? If I know the originator .cpp then I would know the exact fusion sequence and therefore I would just dynamic_cast the fusion_seq_tag. Regards,

2 1

[boost][task] task and fiber do not compile
by Kishalay Kundu 29 Jan '10

29 Jan '10

Hi Oliver, I have boost 1.41.0 installed. I have been trying to build boost-task and boost-fiber (I guess this is the pre-req for building boost-task). However when I do: "./bjam threading=multi --with-task", the compiler errors out with a long list of compiler errors. I have gcc 4.4.1 installed on my Fedore 11 system. Could you please let me know what I am doing wrong? Thank you. regards, Kishalay

1 0

[BGL] Color Maps
by pablo barrio lópez-cortijo 29 Jan '10

29 Jan '10

Hi there! I'm new to Boost (and also to this mailing list). I'm using depth_first_search and it's vital for me to be able to specify the root vertex (4th argument), and that means also specifying the Color Map, which I don't care too much about. It'd be great to have a default color map but I'm a newbie and I didn't found anything in the docs. My initial options were: a) Using sth like a "default_color_map" class. I ignore if there is such thing. b) Creating a property map out of a vector, something like "vector_property_map<default_color_type>". I tried and it didn't work. c) Including a property in my graph called "color" of type "default_color_type", and creating a property map out of it. Didn't work either. Actually, I get an error when trying to include the property with that type (compiles if the property is an int). d) Just copying the color map that the algorithm uses when called without the color map argument. Hoping to find out how the BGL code works was simply unrealistic so I abandoned the idea. Is there any way to easily instantiate a Color Map? I'd post my code, but I have 6 or 7 versions none of which work. Any help will be much appreciated. Thanks ahead!

2 1

Detect specialized template function?
by dherring＠ll.mit.edu 29 Jan '10

29 Jan '10

Slightly OT, but I was hoping someone here might know the magic incantation (which may involve a Boost library). Given a template class such as template <class T> struct S { void f(); void g() { T t(h()); cb(t) } T & h(); boost::function1<void, T &> cb; }; Is there a way to have S.f() behave differently depending on whether S.g() uses the default definition or whether it is specialized for this particular type? This detection needs to occur before S.g() is ever called. For example, the mere existance of the specialization template <> void S<int>::g() { ... } should trigger special logic in S<int>::f(). Thanks, Daniel P.S. The motivation for this is an attempt to retrofit a callback API that allows specialization of S.g() or the registration of a boost::function that gets called by the default S.g(). We need to disable callbacks until an actual function is registered via either method. It is easy to dispatch on registration of the boost::function; I don't know how to detect the specialization.

2 1

[boost::function] Is there a way to assign a generic function, regardless number of parameters?
by Matheus Araújo Aguiar 29 Jan '10

29 Jan '10

Hello list, My doubt is simple. We need to implement some callbacks, but we don't want to declare a boost::function to every different callback, we just want to declare one generic boost::function to receive any function, with any number of parameters (the return type may be one well defined, no problem with that in our case). Is there a way to do this? What we want to do is something like is done in the asynchronous methods of boost::asio when passing the callback to the async operation. TIA, -- Matheus Araújo Aguiar Computer Scientist matheus.pit(a)gmail.com

3 5

[UTF] Provide a custom init_unit_test
by Ahmed Badran 29 Jan '10

29 Jan '10

Hi, I have a single test module (defined via BOOST_TEST_MODULE), then multiple test suites, each defined in its own .cpp file and I am using a single header variant of the unit test framework. Is there an easy way to provide my own init_unit_test and have it plug into the rest of the system, with main and everything else provided by the utf library? I've looked at the headers of utf and it seemed that the generation of main and other initialization routines are tied to also providing an "init_unit_test" which causes a multiple definition problem at link stage. Regards, Ahmed

2 1

Operator overloading of a class template in a namespace
by Surya Kiran Gullapalli 29 Jan '10

29 Jan '10

Hello all, Let me give you a heads up first. This is a general c++ question. not a boost one. But posting it here as we've C++ gurus out there. I've this piece of code http://paste.pocoo.org/show/171324/ The code fails to compile when *St.1 *is uncommented. I read some documents on operator overloading and namespaces and moved the output operator into the same namespace as the class (skg). This works fine. But my question is different. If the output operator is in global namespace and class in skg namespace *St. 2 * (cout statement) works fine, but not the copy statement. Why ? Thanks in advance, and apologies for offtopic. Surya

1 0