December 2013 - Boost-users - lists.preview.boost.org

[test] floating point comparison
by Matthijs Möhlmann 13 Jan '14

13 Jan '14

Hello all, I've the following test program: #define BOOST_TEST_DYN_LINK #define BOOST_TEST_MAIN #define BOOST_TEST_MODULE Master test suite #include <boost/test/unit_test.hpp> BOOST_AUTO_TEST_CASE(TestCase) { BOOST_REQUIRE_CLOSE_FRACTION(45.00f, 45.17f, 0.004f); } When I compile it with the following line: g++ -o main.o -c -Werror -Wall -std=c++11 -fvisibility=hidden -fvisibility-inlines-hidden -DBOOST_SPIRIT_USE_PHOENIX_V3 -DUTEST -g main.cpp g++ -o main main.o -lboost_unit_test_framework I get the following output: Running 1 test case... *** No errors detected If I lower the tolerance to 0.0004f then I'll get the following: Running 1 test case... main.cpp(8): fatal error in "TestCase": difference{0.00376352} between 45.00f{45} and 45.17f{45.1699982} exceeds 0.00039999999 *** 1 failure detected in test suite "Master test suite" Why does it say that the difference is 0.00376352 and not 0.17....... ? Regards, Matthijs Möhlmann

4 4

BOOST.TEST Mixing automatic and manual test registratioI
by Manuel Maier 13 Jan '14

13 Jan '14

Hi, Is it possible to mix manual and automatic test registration in Boost,.Test. Example: BOOST_AUTO_TEST_SUITE( ExampleTestSuite ) BOOST_AUTO_TEST_CASE( testcase1 ) { BOOST_TEST_MESSAGE("Test 1 in Example Test Suite"); BOOST_WARN_EQUAL(1, 1); BOOST_REQUIRE_MESSAGE(5 > 1, "Here is an example message. Something may be wrong"); } BOOST_AUTO_TEST_SUITE_END() Now I want to manually add a testcase to this testsuite. This seems to be important for me, as there is no automatic registration version for BOOST_PARAM_TEST_CASE(...) I also made a manualTestRegstration: test_suite* TestSuiteA = BOOST_TEST_SUITE("test_suiteA"); void manualTestCaseRegistration() { TestSuiteA->add( &test_case ); framework::master_test_suite().add( TestSuiteA ); } Is it possible to add a manual testcase(e.g.: test_case_b) to the ExampleTestSuite? Searching the internet I stumbled over following thread: https://svn.boost.org/trac/boost/ticket/7000 But it seems to me this patch never made it to release(at least until now). In fact( I guess) this could/would be exactly the feature I am looking for. Kind Regards and thanks for help Manuel _____________________________________________ Manuel Maier, B. Eng. Business Unit "Motion" / Research & Development Bernecker + Rainer Industrie-Elektronik Ges.m.b.H. B&R Strasse 1 5142 Eggelsberg Austria Tel: +43 7748 6586 Fax: +43 7748 6586 - 24 Email: manuel.maier(a)br-automation.com http://www.br-automation.com ____________________________________________________________ Firmenbuchnummer: Landesgericht Ried im Innkreis / FN 111651 v

3 3

[boost.test] How to register custom exception for auto test?
by Jörg Becker 13 Jan '14

13 Jan '14

Hi, I'm using the test library of boost 1.49.0 for my unit tests. It works fine so fare, but if my program throws an error, I get the default error message unknown location(0): fatal error in "excption1_test": unknown type, witch hides the original error message. I tried to register an error translator, but without success. Here is what I tried (modified test/test/custom_exception_test.cpp to use auto test with global fixture): #define BOOST_TEST_MODULE "boost test custom_exception_test" #include <boost/test/unit_test.hpp> #include <boost/test/unit_test_monitor.hpp> using namespace boost::unit_test; BOOST_AUTO_TEST_SUITE( custom_exception_testsuite ) struct my_exception1 { explicit my_exception1( int res_code ) : m_res_code( res_code ) {} int m_res_code; }; void throw_my_exception1() { throw my_exception1( 12 ); } void my_exception1_translator( my_exception1 ) { BOOST_TEST_MESSAGE( "Caught my_exception1" ); } class my_config { public: my_config () { unit_test_monitor.register_exception_translator<my_exception1>( &my_exception1_translator ); } }; BOOST_GLOBAL_FIXTURE( my_config ) BOOST_AUTO_TEST_CASE( excption1_test ) { throw_my_exception1(); } BOOST_AUTO_TEST_SUITE_END() Can anybody tell me what I'm missing? Thanks, Jörg

3 5

[BOOST]BOOST_AUTO_TEST_CASE_TEMPLATE vs BOOST_AUTO_TEST_CASE
by Hemanth Choudary M V (RBEI/ETA1) 13 Jan '14

13 Jan '14

Hi, Scenario1: Both files are present in the same Project BoostUnitTest. fileA: typedef boost::mpl::list<long long,unsigned char> test_types; BOOST_AUTO_TEST_SUITE(MySuite) BOOST_AUTO_TEST_CASE_TEMPLATE( my_test, T, test_types ) { BOOST_CHECK_EQUAL( sizeof(T), (unsigned)4 ); } BOOST_AUTO_TEST_SUITE_END() fileB: typedef boost::mpl::list<long long,unsigned char> test_types; BOOST_AUTO_TEST_SUITE(MySuite) BOOST_AUTO_TEST_CASE_TEMPLATE( my_test, T, test_types ) { BOOST_CHECK_EQUAL( sizeof(T), (unsigned)4 ); } BOOST_AUTO_TEST_SUITE_END() As shown above, MySuite is extended across 2 files fileA and fileB. The BOOST_AUTO_TEST_CASE_TEMPLATE name is same across the both the files and and the suite name is also same across both the files.. Boost is able to differentiate between the 2 test cases and the test cases are executed. Scenario2: fileA: BOOST_AUTO_TEST_SUITE(MySuite) BOOST_AUTO_TEST_CASE( my_test) { BOOST_CHECK_EQUAL(1,1); } BOOST_AUTO_TEST_SUITE_END() fileB: BOOST_AUTO_TEST_SUITE(MySuite) BOOST_AUTO_TEST_CASE_TEMPLATE( my_test ) { BOOST_CHECK_EQUAL( 1,1 ); } BOOST_AUTO_TEST_SUITE_END() In the scenario2, I get a compile time error (Error 1 error LNK2005: "public: void __thiscall mySuite::insuite2::test_method(void)" (?test_method@insuite2@mySuite@@QAEXXZ) already defined in BoostUnitTest.obj) which basically suggests that one or more multiply defined symbols are found. Why is the behaviour different in these 2 scenarios? How is Boost able to differenciate between the 2 test cases in Scenario1? Mit freundlichen Grüßen / Best Regards, Hemanth Choudary MV RBEI/ETA1

3 2

[Boost.Test] Detail lost when using BOOST_*_NO_THROW
by Adam Nielsen 13 Jan '14

13 Jan '14

Hi again, While developing tests I sometimes find that functions that should not throw exceptions do. Because this is an unexpected situation, I would like to know what exception has been thrown so I know what problem to look for in my test design. Unfortunately the Boost.Test framework only provides a terse message like this: test.cpp(123): fatal error in "fnTest": exception thrown by test(); It doesn't tell you which exception was thrown, nor the reason why. Is there some way to tell Boost.Test to include the exception message (from what()) if the exception derives from std::exception? This would make problem solving much quicker as you would not have to run the tests through a debugger to see which exception is being thrown and why. There is some discussion about this on Stack Overflow[1] but the consensus seems to be the best solution is not to use BOOST_*_NO_THROW. If you don't use these and an exception is thrown, a lot more detail is provided. Assuming this is a limitation in Boost.Test, would it be possible to add this as a feature so that future versions include the same level of detail whether you use BOOST_*_NO_THROW or not? Many thanks, Adam. [1] http://stackoverflow.com/questions/15133259/boost-check-no-throw-how-to-get…

3 2

--run_test and using this to run a subset of Testcases.
by Manuel Maier 13 Jan '14

13 Jan '14

Hi I am looking for help considering running sets of Testcases/Testsuites. First off, is there a possibility to do a negative selection. i.e. tell boost.test to NOT run certain Testcases/testsuites? Further is it possible to run sets of Testcases(which may reside in different testsuites)? Assume following test tree: MASTER suite1 case1 case2 suite2 testcase1 testcase2 suite3 subsuite testcase1 Now lets say I wanna run testcase case1, testcase2 and the subsuite testsuite. Is this even possible? Kind Regards and thanks a lot for help Manuel Maier _____________________________________________ Manuel Maier, B. Eng. Business Unit "Motion" / Research & Development Bernecker + Rainer Industrie-Elektronik Ges.m.b.H. B&R Strasse 1 5142 Eggelsberg Austria Tel: +43 7748 6586 Fax: +43 7748 6586 - 24 Email: manuel.maier(a)br-automation.com http://www.br-automation.com ____________________________________________________________ Firmenbuchnummer: Landesgericht Ried im Innkreis / FN 111651 v

5 6

[Boost.Test] Unable to perform init at startup
by Adam Nielsen 12 Jan '14

12 Jan '14

(Apologies for the last message, I am using an unfamiliar keyboard and accidentally pressed the 'send' keyboard combination.) Hi all, I'm having a bit of a problem trying to figure out how to perform some class init while manually registering my tests. I want to run a set of tests which all share the same class instance (this is a factory class that creates the objects to be tested.) I am creating a number of global class instances and using boost::bind in the constructor to register the tests with Boost.Test. This works fine, except that I need to create the factory class before the tests can be run. In case the factory class fails, I want to use some of the Boost.Test checks in the initialisation function. However when I do this, I get this runtime error: terminate called after throwing an instance of 'std::runtime_error' what(): can't use testing tools before framework is initialized I am assuming this means I need to wait with the construction of my factory class until Boost.Test is ready, but I am unsure how to do this. I am #defining BOOST_TEST_MODULE, and when I try to declare a function called init_unit_test() as suggested in the docs it tells me the function already exists. If I remove the BOOST_TEST_MODULE #define, then it complains that main() is not implemented. I am a bit confused because I do not understand the reason for having init_unit_test() if you cannot implement it. If you choose not to use the Boost auto function then it seems you must implement your own main(), so I would expect your init code would go there, so there is no need for the init_unit_test() function? Is there some other way not mentioned in the docs to manually register tests (which are class member functions) but where the class can call Boost.Test functions in its constructor? Many thanks, Adam.

2 2

[ASIO] Async TCP/UDP server design help
by Terry Lazoki 02 Jan '14

02 Jan '14

I'm having hard time understanding the boost asio design and how i can use it to tackle my problem. My problem: Designing a tcp/udp server that is time efficient. 1. User sends packet through tcp 2. Server receives packet and processes it to build a large tree (n > 20000) 3. Server then parses the new tree and sends a summary packet through udp to another client Conditions/Limitations: - Has to be time efficient, receiving, processing and sending has to be done as fast as possible. - Processing the packet is heavy and can cause delays if its done in the same thread - Parsing the tree is also a little heavy when sending it through udp, can cause delays too Design 1 (2 threads): Thread 1: ASIO TCP and UDP. TCP receives packets and adds it to a circular buffer. While at the same time, processes the shared tree and sends packet through udp. Thread 2: Server grabs packet from circular buffer and processes it Some problems here: - I dont understand the async part here. What happens when tcp receives a packet, if the current thread is continuously parsing the tree and sending packets through udp. Do i need to separate the udp part to another thread? - I have 3 parts of the program that i need to run concurrently, well its really 2 parts and receiving packets through tcp. Any help is appreciated. Thanks,

3 2

Mysterious early data termination using filtering_iostreams zlib_compressor
by Kenneth Adam Miller 31 Dec '13

31 Dec '13

I'm using intel's pin tool (pintool.org) and the boost library's filtering_stream in order to compress data before it goes to a file. I already have functioning code that does that, but at the very end of the execution of the target, right before the PIN_AddFiniFunction callback gets executed the stream of bytes in the compressed version ends (I assume this because the decompressed version is generated by subsequently opening another filtering_stream and using a zlib_decompressor and reading the entire compressed file out). I don't know why the stream ends early... This makes getting help kind of hard, because the files that are being produced are hundreds of megabytes, and differ by just a little kb. So even the diff is large, but I think the salient point is that the data included in the compressed/decompressed version is the same aside from the end. To make this easy for anybody helping, I've included a link to my repo where I've created a reproducible mini example, which should be able to be downloaded, built, and run without any hassel on the latest ubuntu: https://github.com/KennethAdamMiller/Boost-PinTool (It's a drop in replacement for debug trace; you just drop the appropriate files into the DebugTrace folder in your PIN_ROOT/source/tools/, then type make. There are very clear instructions! :) ) link to boost's filtering_stream: http://www.boost.org/doc/libs/1_54_0/libs/iostreams/doc/classes/filtering_s… After you build and run the tool, you should see debugtrace.out and decompressed.xml (the xml part I'm too lazy to change even as I type; it's just text). Go to the bottom of each, you can see that debugtrace.out ends with the line containing "eof" properly, but that the decompressed.xml does not... Can anybody tell me what the issue is? I'm quite certain that everything *should* have went into the FilteringStream that went into the ofstream... please help, I have no idea why it's terminating this way...

1 0

Matrix Vector Multiplication in MPI (Coordinate Storage Format)
by ozhan fenerci 30 Dec '13

30 Dec '13

I wrote a serial code for a matrix-vector multiplication in a Coordinate Storage Format(COO). COO method has already made a big speed-up. My matrix is in a tridiagonal form. I want to parallelize the code and see how the parallelizm improves the code. My matrix could grow up to 2.000.000*2.000.000. Any efficient parallelization approach doesn't come to my mind (I have written some basic mpi programs but not much more. So my experience is limited.) Could you give me some headstart? Thanks. Özhan Fenerci #include <boost/numeric/ublas/banded.hpp> #include <boost/numeric/ublas/io.hpp> #include <vector> #include <iostream> #include "mpi.h" #include <stdio.h> int main(int argc, char *argv[] ) { using namespace boost::numeric::ublas; using std::cout; using std::endl; int numprocs, rank, chunk_size, i,j; int max, mymax,rem; int totalNonZeros; //totalNonZeros of TriDiagonal Matrix MPI_Status status; //b=A*x banded_matrix<double> A (10, 10, 1, 1); //vector<double> x(10); //vector<double> b(10); int* x = new int[10]; int* b = new int[10]; /* Initialize MPI */ MPI_Init( &argc,&argv); MPI_Comm_rank( MPI_COMM_WORLD, &rank); MPI_Comm_size( MPI_COMM_WORLD, &numprocs); //MPI_Status status; if (rank == 0) { for (signed i = 0; i < signed (A.size1 ()); ++ i) { for (signed j = std::max (i - 1, 0); j < std::min (i + 2, signed (A.size2 ())); ++j) A (i, j) = i + j; x[i]= 1; } } if(rank == 0) { cout << A <<endl<<endl; totalNonZeros = 0; for (signed i = 0; i < signed (A.size1 ()); ++ i) { for (signed j = std::max (i - 1, 0); j < std::min (i + 2, signed (A.size2 ())); ++ j) if( A(i,j) !=0) totalNonZeros++; } } cout<<"Number ot NonZeros "<<totalNonZeros<<endl; if(rank == 0) { int * row_idx = new int[totalNonZeros]; // row index array int * col_idx = new int[totalNonZeros]; // col index array int * val = new int[totalNonZeros]; //val array int index=0; for (signed i = 0; i < signed (A.size1 ()); ++ i) { for (signed j = std::max (i - 1, 0); j < std::min (i + 2, signed (A.size2 ())); ++ j) if( A(i,j) !=0) { row_idx[index] = i; col_idx[index]=j; val[index] = A(i,j); index++; } } //Clear out result for ( int k=0; k<A.size1(); ++k) { b[k]= 0; } //Matrix-vector Multiply for (int index=0 ; index <totalNonZeros;++index){ b[row_idx[index]] = b[row_idx[index]]+ val[index]*x[col_idx[index]]; cout<<b[3]<<endl; } //delete row_idx; // delete col_idx; //delete val; } MPI_Finalize(); return 0;

2 1