October 2011 - Boost-users - lists.boost.org

Is there Text Diff in Boost Unit Testing?
by Alexander Feldman 19 Oct '11

19 Oct '11

Hi Guys, I am now using Boost Unit Testing as well and I am quite happy. I am writing, among other things, a compiler and I would benefit if in BOOST_CHECK_EQUAL there is some kind of a text diff... Is there such a facility? Thanks, -- Alex

2 1

Proposal mpl::switch_case
by Hite, Christopher 19 Oct '11

19 Oct '11

My use case is a fast protocol decoder. I've described the protocol in mpl and generate the instructions I need to decode/encode/store packets. During decoding I have to be able to take a run time id and map it to the correct compile time generated code. Traditionally such code was written as a switch/case. The obvious solution is to use metaprogramatic recursion, for example mpl::for_each to check the id against each value. You would hope the compiler would inline all the recursion, detect the ifs on an id, and do general switch/case optimization. Unfortunately the code generated by gcc and it's not as good as I hoped. They actually generate a function for every ~6 levels of recursion and just do a bunch of if/equal/else. boost::variant has this same problem. They use BOOST_PP_REPEAT to generate an actual switch/case statement(visitation_impl.hpp). The compiler does a good job with the swich/case. It removes redundant cases and then optimizes using array lookups and binary searches depending on size. Good job! 1) Could we factor this logic out into a utility like mpl::switch_case? 2) What would the best interface look like? 3) What other bits of code could benefit from it? phoenix::switch_? Meta State Machine? Chris

2 1

[serialization] unregistered_class exception that wasn't occurring before
by Igor R 19 Oct '11

19 Oct '11

Hello, I'm serializing the following type: (pseudo-definition) multi_index_container< variant< shared_ptr<template1<arg1> >, shared_ptr<template2<arg2> >, etc... > > It works well when I make and load archives with the current version of my app, but when I try to open an archive made by the pervious version it throws unregistered_class for one of my classes. The both versions use boost 1.47, and the differences between them seem to me absolutely unrelated to serialization (as far as I can see). Unfortunately I can't provide here code snippets, so my question is theoretical: what could break the backward compatibility in such a way? What should I look into? Thanks.

2 4

[shared_ptr] implicit conversion to bool problem
by Steckmann, Sven 19 Oct '11

19 Oct '11

Dear list, I have a problem with the implicit conversion to bool in the shared_ptr class. I want to use operator overloading for an easy to use interface. I have 2 Instances of a class with a common Interface and I want to define an operator-class (OpAnd of type K) which can handle the operator. By compiling this little test below, I got: g++ -I/home/steckmann/proj/libs/boost -O0 -g3 -Wall -c -fmessage-length=0 -MMD -MP -MF"test.d" -MT"test.d" -o "test.o" "../test.cpp" ../test.cpp: In function 'int main()': ../test.cpp:49: error: ambiguous overload for 'operator&&' in 'a && b' ../test.cpp:49: note: candidates are: operator&&(bool, bool) <built-in> ../test.cpp:40: note: boost::shared_ptr<K> operator&&(boost::shared_ptr<K>, boost::shared_ptr<K>) This is quite clear to me, but how can I avoid this without touching the shared_ptr? Thanks for any comments, Sven #include <boost/shared_ptr.hpp> using namespace boost; class K { public: virtual bool value() = 0; }; class Value :public K { public: Value( bool i ) : var( i ) {} bool value() { return var; } bool var; }; class OpAnd : public K { public: OpAnd( boost::shared_ptr<K> l, boost::shared_ptr<K> r ) :l(l), r(r) {} bool value() { return l->value() && r->value(); } boost::shared_ptr<K> l; boost::shared_ptr<K> r; }; boost::shared_ptr<K> operator&&( boost::shared_ptr<K> l, boost::shared_ptr<K> r) { return shared_ptr<K>( new OpAnd(l, r ) ); } int main() { boost::shared_ptr<Value> a( new Value(true) ); boost::shared_ptr<Value> b( new Value(false) ); boost::shared_ptr<K> c = a && b; // Now use c here for some stuff c->value(); // returns false b->var = true; c->value(); // returns true return 0; }

2 1

Why does the creating of distributed adjacency list cost so much time?
by Yan 19 Oct '11

19 Oct '11

Hi all, I'm learning to use the distributed adjacency list and distributed dijkstra SP-algorithm. I use some test data contains about 3,800,000 edges and 2,800,000 vertices. I use a pair of iterators to pass the data to the distributed adjacency list, like that: Graph g(reader.begin(), reader.end(), reader.weight_begin(), reader.num_vertices(), pg, dist); "reader" is an user-defined class which supplys required data (edges, vertices, weights). "pg" is defined like that "mpi_process_group pg;" "dist" is supplied by the reader like that "graph::metis_distribution dist = reader.distribution();". It is based on a metis partition file coordinating the test data. I run 2 processes on my notebook computer(Intel core2 2.0G, Ram 2G, WinXPsp3). I also test the same data using a sequential bgl dijkstra SP-algorithm on this computer. The result is that: Parallel bgl: Creating of Graph g costs *10080 seconds*; dijkstra computing costs 28.281 seconds. Sequential bgl: Creating of Graph g costs *4.563 seconds*; dijkstra computing costs 1.703 seconds. As you see, when using distributed adjacency list, it costs so much time to create the graph. Is it nomal? I have done some more comparison between 2-processes parallel dijkstra on one core2 computer and sequential dijkstra on the same computer, using different scale data set. Always, the sequential bgl dijkstra costs less time than the 2-processes parallel dijkstra. What's the problem? I have read the boost document on "boost_1_46_1\libs\graph_parallel\doc\html\dijkstra_shortest_paths.html", which shows a good performance and speedup in the charts. Doesn't that mean less time costing than the sequential program? Any advice would be appreciated. Thanks, Yan -- View this message in context: http://boost.2283326.n4.nabble.com/Why-does-the-creating-of-distributed-adj… Sent from the Boost - Users mailing list archive at Nabble.com.

2 2

condition variable predicate and lambda
by Christopher Pisz 19 Oct '11

19 Oct '11

I want to do something simple, but implementing it seems awefully complex. All I want to do is wait for a specified amount of time for a member variable of my class to equal a certain value. Can someone please help me with an example? I've tried several that came from Googling and read the documentation, but I get compile errors no matter what I try. Currently the compiler is complainging "term does not evaluate to a function taking 0 arguments" Here is the problem in all its simplicity: class A { public: enum STATE { HAPPY = 0, SAD, ANGRY, NUM_STATES }; Foo(boost::posix_time::millisec & timeout) { // I need to wait to return from this until the state is HAPPY } private: STATE state_; };

3 4

Wave doesn't recognize c++0x keywords ?
by 邓尧 19 Oct '11

19 Oct '11

Hi, I'm new to Boost.Wave. I followed the "quick start", but found that wave take the C++0x keyword "decltype" as an identifier instead of a keyword. The macro "BOOST_WAVE_SUPPORT_CPP0X" is defined as "1". Boost version: 1.47.0 Any suggestions ? Thanks. -Yao

2 2

Re: [Boost-users] Building boost with stlport?
by Brian Vandenberg 18 Oct '11

18 Oct '11

> STLport is not used much nowadays. Used to use it with msvc-6 and 7.1, but > those are loooong gone here. Keep in mind that msvc is not the only place someone might consider using a more portable STL implementation. Where I work we're using Sun Studio; at least in old implementations their support for STL was anything but standards compliant -- in particular, people were doing god-awful things that never should have passed peer review; stuff that stlport or gcc puke on. Using stlport in older versions of Sun Studio is all but necessary if you want to use boost. More recent releases of Sun Studio are less troublesome, but in our case we're porting between Solaris and Linux; the implementations of STL between SS for Solaris & SS for Linux appeared to have different behaviors for some things, and if we port to gcc there'll no doubt be issues with poorly written code that worked in SS that doesn't work under gcc. In short, I wouldn't discount stlport as an ancient, no-longer-used piece of software; for those of us trying to port/improve older and more fragile code, stlport is a boon. -Brian

2 1

Chrono System time Conversion to Since Epoch not working
by Michael Chrostowski 18 Oct '11

18 Oct '11

Hi, I'm using boost 1.47 on Visual Studio 2008 and cannot seem to get the time since epoch from the Chrono library. At first I noticed that boost::chrono::system_time::now() does not have to be represented in time since epoch format so I don't believe that's an issue. However, I tried calling boost::chrono::system_time::now().time_since_epoch() and I get the same return result as if I was only referencing now(). That is, it is not time since epoch. Calling the static method to_time_t() DOES return proper time since epoch but it is in seconds (as time_t generally is) and is not what I'm looking for. Michael Chrostowski

2 1

boost atomic query
by Sean Farrow 18 Oct '11

18 Oct '11

Hi: I'm just starting to use boost atomic. I need to protect an std::map. If I had: boost::atomic<std;:map<int, int> > test; Would I need to protect this with a mutex, or is the fact that the variable is atomic enough? Cheers SEean.

2 1