June 2006 - Boost-users - lists.preview.boost.org

Help with Tokenizer
by Frank 09 Jun '06

09 Jun '06

I have a string that is delimited with '|' that I want to parse with boost::tokenizer. ie: "label 1|label2||label 3" It all parses ok except that the whitespace within the label is removed. Is there a way to configure the tokenizer to leave whitespace alone? Thanks Frank

2 2

[test] BOOST_CHECK_NOT_EQUAL?
by Ed Johnson 09 Jun '06

09 Jun '06

Hi, Is there a way to get the funcationality of BOOST_CHECK_EQUAL when I want to test that two parameters are not equal to each other? I like the way BOOST_CHECK_EQUAL allow me to see the mismatched runtime values. I have been creating a custom message for BOOST_CHECK_MESSAGE to achieve this functionality, but it would be nicer if there was an automated way to do it. Thanks, Ed Johnson

4 11

[Review][PQS] Review ends
by fred bertsch 09 Jun '06

09 Jun '06

Hello, boosters! Any given day exists for 50 hours, but it does end at some point. Likewise, the PQS review ends at some point. I will arbitrarily say midnight UTC and the timezone of the author. However, it will take me roughly a week to collate all of the reviews and reach an official decision, so don't feel shy about posting or mailing me additional reviews. I thank the reviewers very much for reviewing the library, and I thank Andy Little even more for all of his work responding to the reviews and for writing the library in the first place! -Fred Bertsch

1 0

Problem with serialization of shared_ptr (with example code)
by Tim Taylor 09 Jun '06

09 Jun '06

Hi, I am having some troubles with serialization of shared_ptrs. The basic problem is that when I serialize a shared_ptr, then later deserialize into a new shared_ptr, the use_count of the new pointer is, incorrectly (as far as I can see), 2 rather than 1. I am using boost 1.33.1 and gcc 4.0.1. Here is some example code to demonstrate the problem: ----- // compile with: // gcc -o ptrser-test ptrser-test.cpp -lstdc++ -lboost_serialization-gcc #include <iostream> #include <fstream> #include <boost/archive/text_oarchive.hpp> #include <boost/archive/text_iarchive.hpp> #include <boost/serialization/shared_ptr.hpp> using namespace std; using namespace boost; class Item { public: // Constructors/destructor-------------------------- Item(int n = 0) : _n(n) { cout << "In Item constructor" << endl; }; Item(const Item& other) { cout << "In Item copy constructor" << endl; _n = other._n; }; ~Item() { cout << "In Item destructor" << endl; }; // Data members------------------------------------- int _n; // Methods------------------------------------------ friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & _n; } }; class Bag { public: // Constructors/destructor-------------------------- Bag() { cout << "In Bag default constructor" << endl; }; Bag(shared_ptr<Item> item) : _item(item) { cout << "In Bag shared_ptr constructor" << endl; }; Bag(const Bag& other) { cout << "In Bag copy constructor" << endl; _item = other._item; }; ~Bag() { cout << "In Bag destructor" << endl; }; // Data members------------------------------------- shared_ptr<Item> _item; // Methods------------------------------------------ void printUseCount() const { cout << "Use count: " << _item.use_count() << endl; } friend class boost::serialization::access; template<class Archive> void serialize(Archive& ar, const unsigned int version) { ar & _item; } }; int main(int argc, char **argv) { const char *filename = "bag.ser"; // create a Bag object with a new Item in it Bag *pBag = new Bag(shared_ptr<Item>(new Item(1))); // now serialize the Bag cout << "About to serialize bag..." << endl; ofstream osSer(filename); boost::archive::text_oarchive oa(osSer); const Bag& constBag = *pBag; oa << constBag; osSer.close(); pBag->printUseCount(); // delete the original Bag cout << "About to delete bag 1..." << endl; delete pBag; // create a new Bag object cout << "About to create a new bag..." << endl; pBag = new Bag; pBag->printUseCount(); // and deserialize it into the new Bag object cout << "About to deserialize bag..." << endl; ifstream isSer(filename, std::ios::binary); boost::archive::text_iarchive ia(isSer); ia >> *pBag; isSer.close(); pBag->printUseCount(); // delete the second BSag cout << "About to delete bag 2..." << endl; delete pBag; return 0; } ----- This code produces the following output: ----- In Item constructor In Bag shared_ptr constructor About to serialize bag... Use count: 1 About to delete bag 1... In Bag destructor In Item destructor About to create a new bag... In Bag default constructor Use count: 0 About to deserialize bag... In Item constructor Use count: 2 About to delete bag 2... In Bag destructor In Item destructor ----- The series of constructor and destructor calls is as expected. After the first Bag object is created with a shared_ptr to an Item in it, the use_count of the shared_ptr is 1. When the first Bag object (including the shared_ptr) are deleted, and a new, empty Bag object created, the use_count is now (as expected) 0. However, when the serialization archive is deserialized into the new Bag object, the use_count of the shared_ptr jumps from 0 to 2. Am I doing something silly here, or is this a bug in the shared_ptr serialization code? For interest, the saved archive (bag.ser) produced by this code looks like this: ----- 22 serialization::archive 3 0 0 0 1 2 1 0 0 1 ----- Any help would be much appreciated. Tim

3 3

[Review][PQS] Physical Quantities System review (May 31 through June 9)
by fred bertsch 09 Jun '06

09 Jun '06

The review of Andy Little's Physical Quantities System begins today, May 31, 2006, and continues through June 9, 2006. Download: pqs_3_1_0 is the Official Boost Review version of pqs. (Please note that pqs_3_1_0 has a bug in one header. This can be either patched according to the latest release notes, "relnotes.txt" available from http://tinyurl.com/7m5l8 or alternatively download pqs_3_1_1, which is basically equivalent to pqs_3_1_0, except the bug has been fixed there. Description PQS (Physical Quantities System) is designed to replace the use of floating point types for modelling physical quantities in C++ programs. Advantages include automated dimensional analysis checking, automatic unit conversions, self documentation of code and uniform mechanisms for dealing with physical quantities overall, allowing better interoperability, accuracy and repeatability among engineering, physics and trade applications. The library will ultimately consist of various types for modelling quantities in different situations. The t1_quantity, which is the first to be implemented, performs all its dimensional analysis checks as well as parts of its unit conversion calculations, at compile time rather than runtime, for maximum efficiency. PQS is designed with a firm commitment to the SI system, as being the internationally agreed preferred system of units. A reasonable selection of common quantities and units is included in the review version for demonstration purposes. The intent is to expand on this to include all the common quantities mentioned in Appendix B.8 and B.9 of the NIST SI document: http://physics.nist.gov/cuu/pdf/sp811.pdf The library also defines a selection of physical constants again intended as a flavour of the constants that would be included in a full library implementation as outlined in: http://physics.nist.gov/cuu/pdf/chart1.pdf Review questions: Please always explicitly state in your review, whether you think the library should be accepted into Boost. You might want to comment on the following questions: - What is your evaluation of the design? - What is your evaluation of the implementation? - What is your evaluation of the documentation? - What is your evaluation of the potential usefulness of the library? - Did you try to use the library? With what compiler? Did you have any problems? - How much effort did you put into your evaluation? A glance? A quick reading? In-depth study? - Are you knowledgeable about the problem domain? -Fred Bertsch

4 4

Re: [Boost-users] Boost.Signals: strange bug in example program
by Pranckevicius, Tadas 09 Jun '06

09 Jun '06

-----Original Message----- From: Sebastian Redl [mailto:sebastian.redl@getdesigned.at] Sent: 09 June 2006 16:22 To: boost-users(a)lists.boost.org Subject: Re: [Boost-users] Boost.Signals: strange bug in example program Pranckevicius, Tadas wrote: >I tried to run this example and I got the same results, it prints 8 on >Microsoft Visual studio 2003 (7.1) as well. > >I did try to debug it and ended up in "algorithm" file: > >After adding cout: > >template<class _FwdIt> inline > _FwdIt _Max_element(_FwdIt _First, _FwdIt _Last) > { // find largest element, using operator< > _DEBUG_RANGE(_First, _Last); > _FwdIt _Found = _First; > cout<<*_Found<<endl; > if (_First != _Last) > for (; ++_First != _Last;) { > cout<<*_First <<endl; > if (_DEBUG_LT(*_Found, *_First)) > _Found = _First; > } > return (_Found); > } > > >My Visual Studio.Net 2003 does not even have this function anywhere in >the include directory. What precise version are you using? Have you >replaced your standard library? >Sebastian Redl Sorry for the confusion I have debugged in VS 2005 In VS2003 it is: C:\Program Files\Microsoft Visual Studio .NET 2003\Vc7\include\algorithm And it looks like: template<class _FwdIt> inline _FwdIt max_element(_FwdIt _First, _FwdIt _Last) { // find largest element, using operator< _FwdIt _Found = _First; if (_First != _Last) for (; ++_First != _Last; ) if (*_Found < *_First) _Found = _First; return (_Found); } And gives the same result... thanks

1 0

Re: [Boost-users] [multi index].[composite key] Index creation atrun-time.
by JOAQUIN LOPEZ MU?Z 09 Jun '06

09 Jun '06

----- Mensaje original ----- De: Prashant Thakre <prashant.thakre(a)gmail.com> Fecha: Jueves, Junio 8, 2006 8:46 pm Asunto: Re: [Boost-users] [multi index].[composite key] Index creation atrun-time. > One final suggestion :) regarding the choice of index_container in > terms of performance, > memory requirements, serialization support and insert/lookup > opreations considering that std::vector<const std::string> will be the > basis for this container : > 1.Composite keys and hashed indices. > 2.Hashed indices. Excuse me, but I don't get what your suggestion or question is. Would you mind rephrasing or elaborating some more? Thank you. > Thanks a lot for clarifying all the doubts. It is a pleasure. > -- > regards, > Prashant Thakre > Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

3 3

Re: [Boost-users] Boost.Signals: strange bug in example program
by Pranckevicius, Tadas 09 Jun '06

09 Jun '06

>-----Original Message----- >From: Benjamin Winkler [mailto:Benjamin.Winkler@RWTH-Aachen.de] >Sent: 07 June 2006 22:20 >To: boost-users(a)lists.boost.org >Subject: [Boost-users] Boost.Signals: strange bug in example program >Hi, >I just downloaded the newest boost library 1.33.1 and compiled the >"maximum.cpp" example >from the Signals library, which goes like this: >#include <algorithm> >#include <iostream> >#include <boost/signals/signal2.hpp> >template<typename T> >struct maximum { > typedef T result_type; > template<typename InputIterator> > T operator()(InputIterator first, InputIterator last) const > { > if (first == last) > throw std::runtime_error("Cannot compute maximum of zero elements!"); > return *std::max_element(first, last); > } >}; >int main() >{ > boost::signal2<int, int, int, maximum<int> > sig_max; > sig_max.connect(std::plus<int>()); > sig_max.connect(std::multiplies<int>()); > sig_max.connect(std::minus<int>()); > sig_max.connect(std::divides<int>()); > > std::cout << sig_max(5, 3) << std::endl; // prints 15 > > return 0; >} >As stated, the correct result would be 15 but the output really is 8. >However, when I print the values via > std::copy(first, last, std::ostream_iterator<T>(std::cout, " ")); >I get "8 15 2 1 ", which means that the values are calculated correctly. >This error occurs with Microsoft Visual Studio 2005 (8.0) as well as MinGW >3.4.2. >I even tried replacing std::max_element() with my own function, which >expectedly >didn't help. >Seems like a bug to me, but I can't track it down. Could it be some issue >with >copying the iterator? >Benjamin Winkler Hello, I tried to run this example and I got the same results, it prints 8 on Microsoft Visual studio 2003 (7.1) as well. I did try to debug it and ended up in "algorithm" file: After adding cout: template<class _FwdIt> inline _FwdIt _Max_element(_FwdIt _First, _FwdIt _Last) { // find largest element, using operator< _DEBUG_RANGE(_First, _Last); _FwdIt _Found = _First; cout<<*_Found<<endl; if (_First != _Last) for (; ++_First != _Last;) { cout<<*_First <<endl; if (_DEBUG_LT(*_Found, *_First)) _Found = _First; } return (_Found); } It print's 8 15 2 1.66667 This is expected, but if I add to cout *_Found Like this: cout<<*_First<<" "<<*_Found<<endl; Results are: 8 8 8 8 8 8 8 So (_DEBUG_LT(*_Found, *_First) is always comparing the same numbers. Something is wrong with overloaded operator *. Smart people please help!!! Thanks

2 1

unit_test library on linux
by gelbeiche 09 Jun '06

09 Jun '06

Hi, I tried to use boost unittest lib at my linux box (SuSE 10.0 gcc version 4.0.2 ). I'm able to compile #include using boost::unit_test::test_suite; // most frequently you implement test cases as a free functions void free_test_function() { BOOST_CHECK( 2 == 1 ); } test_suite* init_unit_test_suite( int argc, char* argv[] ) { test_suite* test= BOOST_TEST_SUITE( "Example" ); test->add( BOOST_TEST_CASE( &free_test_function )); return test; } and to link against boost_unit_test_framework. make all-am make[1]: Entering directory `/home/pberg/boost_example_unittest' if g++ -DHAVE_CONFIG_H -I. -I. -I. -I/usr/local/include -g -O2 -MT main.o -MD -MP -MF ".deps/main.Tpo" -c -o main.o main.cpp; \ then mv -f ".deps/main.Tpo" ".deps/main.Po"; else rm -f ".deps/main.Tpo"; exit 1; fi g++ -I/usr/local/include -g -O2 -lboost_unit_test_framework -o sharedpointer main.o /usr/lib/gcc/i586-suse-linux/4.0.2/../../../../i586-suse-linux/bin/ld: warning: libstdc++.so.5, needed by /usr/local/lib/libboost_unit_test_framework.so, may conflict with libstdc++.so.6 make[1]: Leaving directory `/home/pberg/boost_example_unittest' However at execution time I get: porschberg(pberg):/home/pberg/boost_example_unittest> ./sharedpointer *** glibc detected *** free(): invalid pointer: 0x0804ba10 *** Abgebrochen I "valgrinded" the executable and here is the output: *** errors detected in test suite "Example"; see standard output for details ==30533== ==30533== Mismatched free() / delete / delete [] ==30533== at 0x1B904A69: free (vg_replace_malloc.c:152) ==30533== by 0x1BC2FC80: operator delete(void*) (in /usr/lib/libstdc++.so.5.0 .7) ==30533== by 0x1B95840B: boost::unit_test::test_suite::~test_suite() (in /usr /local/lib/libboost_unit_test_framework.so.1.33.1) ==30533== by 0x1B9407FA: boost::unit_test::framework_impl::~framework_impl() (in /usr/local/lib/libboost_unit_test_framework.so.1.33.1) ==30533== by 0x1B93DF5E: (within /usr/local/lib/libboost_unit_test_framework. so.1.33.1) ==30533== by 0x1BA9D313: exit (in /lib/tls/libc-2.3.5.so) ==30533== by 0x1BA87EA7: __libc_start_main (in /lib/tls/libc-2.3.5.so) ==30533== by 0x8049120: ??? (start.S:119) ==30533== Address 0x1BC5D908 is 0 bytes inside a block of size 52 alloc'd ==30533== at 0x1B9046D3: operator new(unsigned) (vg_replace_malloc.c:132) ==30533== by 0x8049201: init_unit_test_suite(int, char**) (main.cpp:12) ==30533== by 0x1B93EEBD: boost::unit_test::framework::init(int, char**) (in / usr/local/lib/libboost_unit_test_framework.so.1.33.1) ==30533== by 0x1B94CF93: main (in /usr/local/lib/libboost_unit_test_framework .so.1.33.1) ==30533== ==30533== Mismatched free() / delete / delete [] ==30533== at 0x1B904A69: free (vg_replace_malloc.c:152) ==30533== by 0x1BC2FC80: operator delete(void*) (in /usr/lib/libstdc++.so.5.0 .7) ==30533== by 0x1B940927: boost::unit_test::framework_impl::~framework_impl() (in /usr/local/lib/libboost_unit_test_framework.so.1.33.1) ==30533== by 0x1B93DF5E: (within /usr/local/lib/libboost_unit_test_framework. so.1.33.1) ==30533== by 0x1BA9D313: exit (in /lib/tls/libc-2.3.5.so) ==30533== by 0x1BA87EA7: __libc_start_main (in /lib/tls/libc-2.3.5.so) ==30533== by 0x8049120: ??? (start.S:119) ==30533== Address 0x1BC5D9E8 is 0 bytes inside a block of size 44 alloc'd ==30533== at 0x1B9046D3: operator new(unsigned) (vg_replace_malloc.c:132) ==30533== by 0x804929E: init_unit_test_suite(int, char**) (unit_test_suite.hp p:222) ==30533== by 0x1B93EEBD: boost::unit_test::framework::init(int, char**) (in / usr/local/lib/libboost_unit_test_framework.so.1.33.1) ==30533== by 0x1B94CF93: main (in /usr/local/lib/libboost_unit_test_framework .so.1.33.1) ==30533== ==30533== ERROR SUMMARY: 7 errors from 7 contexts (suppressed: 24 from 2) ==30533== malloc/free: in use at exit: 9643 bytes in 6 blocks. ==30533== malloc/free: 24 allocs, 19 frees, 11593 bytes allocated. ==30533== For counts of detected errors, rerun with: -v ==30533== searching for pointers to 6 not-freed blocks. ==30533== checked 210328 bytes. ==30533== ==30533== LEAK SUMMARY: ==30533== definitely lost: 19 bytes in 1 blocks. ==30533== possibly lost: 0 bytes in 0 blocks. ==30533== still reachable: 9624 bytes in 5 blocks. ==30533== suppressed: 0 bytes in 0 blocks. ==30533== Use --leak-check=full to see details of leaked memory. I looked through the mailing list archives with no luck. Could someone give a hint ? Thomas

2 2

[ptr_container] ptr_multimap insert and transfer functions
by Konstantin Litvinenko 09 Jun '06

09 Jun '06

Hello, All! I have a couple of questions about ptr_container library. Specificaly 1) Why ptr_multimap_adapter::insert has non const key_type argument? 2) Why ptr_multimap_adapter::transfer( iterator object, ptr_multimap_adapter& from ) has void return type instead of iterator? PS I use 1.33.1 but this is true for 1.34 too. With best regards, Konstantin Litvinenko.

2 8