- Boost - lists.boost.org

Boost 1.33.1 Beta released
by Douglas Gregor 23 Nov '05

23 Nov '05

Boost 1.33.1 Beta ----------------- Boost 1.33.1 Beta is now available at http://www.boost.org. This is a bug fix release to 1.33.0, containing updates to existing libraries and additional computer support. Updated Libraries * Any Library: Cast to reference types introduced in 1.33.0 is now documented on any_cast documentation page. * Config Library: Don't undef BOOST_LIB_TOOLSET after use. * Boost.Python: o The build now assumes Python 2.4 by default, rather than 2.2 o Support Python that's built without Unicode support o Support for wrapping classes with overloaded address-of (&) operators * Smart Pointer Library: Fixed problems under Metrowerks CodeWarrior on PowerPC (Mac OS X) with inlining on, GNU GCC on PowerPC 64. * Regex Library: Fixed the supplied makefiles, and other small compiler specific changes. Refer to the regex history page for more information on these and other small changes. * Iostreams Library: Improved the interface for accessing a chain's components, added is_open members to the file and file descriptor devices, fixed memory-mapped files on Windows, and made minor changes to the documentation. * Functional/Hash Library: Fixed the points example. * Multi-index Containers Library: Fixed a problem with multithreaded code, and other minor changes. Refer to the library release notes for further details. * Graph Library: Fixed a problem with the relaxed heap on x86 Linux. Fixed problems with cuthill_mckee_ordering and king_ordering producing no results. * Signals Library: Fixed problems with the use of Signals across shared library boundaries. * Thread library: read_write_mutex has been removed due to problems with deadlocks. Supported Compilers Boost is tested on a wide range of compilers and platforms. Since Boost libraries rely on modern C++ features not available in all compilers, not all Boost libraries will work with every compiler. The following compilers and platforms have been extensively tested with Boost, although many other compilers and platforms will work as well. For more information, see the regression test results. New for this release: Support for building with the newest STLport-5.0 was added. The support includes building with MinGW Runtime 3.8 plus STLport-5.0 improved to support wide character operations. Apple GCC 4.0, HP Tru64 C++, and Microsoft Visual C++ 8.0 are supported platforms. We have added an experimental autoconf-like configure script for Unix-like systems: run configure --help for more information. * Apple GCC 3.3, 4.0 on Mac OS X. * Borland C++ 5.6.4 on Windows. * GNU C++ 2.95.3 (with and without STLport), 3.2.x., 3.3.x, 3.4.x, 4.0.x on Windows, Linux and Solaris. * HP C++ for Tru64 UNIX 7.1. * Intel C++ 8.1, 9.0 on Windows, Linux. * Metrowerks CodeWarrior 8.3, 9.4, 9.5 on Mac OS X and Windows. * Microsoft Visual C++ 6.0 (sp5, with and without STLport), 7.0, 7.1, 8.0. Note: Boost does not support the non-standard "Safe" C++ Library shipping with Visual C++ 8.0, which may result in many spurious warnings from Boost headers and other standards- conforming C++ code. To suppress these warnings, define the macro _SCL_SECURE_NO_DEPRECATE. Acknowledgements Douglas Gregor managed this release. A great number of people contributed their time and expertise to make this release possible. Special thanks go to Aleksey Gurtovoy and Misha Bergal, who managed to keep the regression testing system working throughout the release process; David Abrahams, Beman Dawes, Aleksey Gurtovoy, Bronek Kozicki, Rene Rivera and Jonathan Turkanis for greatly improving the quality of this release; Rene Rivera for the new Boost web page design; and Zoltan "cad" Juhasz for the new Boost logo.

6 7

build boost with other standard Lib (std)
by Renato Tegon Forti 23 Nov '05

23 Nov '05

Hi All, I need build boost whit other std, have any way to do this? I need this because my system use STD provided by RogueWave, than I need that boost is build with this std. If have way to do this, what is the steps? Thanks to all -- :o)

2 1

[Boost-bugs] [ boost-Bugs-1364416 ] mistake in documentation of condition
by SourceForge.net 23 Nov '05

23 Nov '05

Bugs item #1364416, was opened at 2005-11-22 22:57 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1364416&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: threads Group: None Status: Open Resolution: None Priority: 5 Submitted By: Nobody/Anonymous (nobody) Assigned to: Mike Glassford (glassfordm) Summary: mistake in documentation of condition Initial Comment: Hello, the documentation (http://www.boost.org/doc/html/condition.html) of class condition contains a mistake: instead of template<typename ScopedLock, typename Pred> bool timed_wait <http://www.boost.org/doc/html/condition.html> (ScopedLock&, Pred); it should be template<typename ScopedLock, typename Pred> bool timed_wait <http://www.boost.org/doc/html/condition.html> (ScopedLock&, const boost::xtime <http://www.boost.org/doc/html/xtime.html> & , Pred); so long, Oliver ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1364416&group_… ------------------------------------------------------- This SF.Net email is sponsored by the JBoss Inc. Get Certified Today Register for a JBoss Training Course. Free Certification Exam for All Training Attendees Through End of 2005. For more info visit: http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

Boost pointer containers compile errors, runtime errors, and compiler warnings
by Axter 22 Nov '05

22 Nov '05

http://www.boost.org/libs/ptr_container/doc/ptr_container.html I'm currently in the middle of writing an article making the comparison between the above boost pointer containers and the following clone smart pointers: http://code.axter.com/copy_ptr.h http://code.axter.com/cow_ptr.h I'm performing my test using boost_1_33_0, and VC++ 7.1 compiler. I'm comparring vector<copy_ptr<T> > to boost::ptr_vector<T> I want to make sure that my article is accurate and complete, so I'm posting this question. When trying to compile using copy constructor for ptr_vector, I get a compile error. boost::ptr_vector<Shape> CopyViaContainerCopyConstructor(ContainerOfOriginalShapes); //Will not compileContainerType boost::ptr_vector<Shape> CopyViaContainerCopyConstructor = ContainerOfOriginalShapes; //Will not compile Is this by design, and if so, why doesn't the boost pointer container have a copy constructor like the STL containers have? Another standard STL container interface that fails to compile using ptr_container is push_back via dereferenced iterator. Example: for(it = ContainerOfOriginalShapes.begin();it != ContainerOfOriginalShapes.end();++it) CopyViaContainerForLoopIterator.push_back(*it); Is this also by designed, and if so, why? I also notice that the follow code fails at runtime. boost::ptr_vector<Shape> CopyViaContainerForLoopOp(ContainerOfOriginalShapes.size()); for(i = 0;i < ContainerOfOriginalShapes.size();++i) CopyViaContainerForLoopOp[i] = ContainerOfOriginalShapes[i]; However, if I do the following, then it works: //Try creating the right size for the container via push_back for(i = 0;i < ContainerOfOriginalShapes.size();++i) CopyViaContainerForLoopOp.push_back(Shape::AbstractFactory(SourceData[i][0], SourceData[i][1])); //Again test copy via for loop using operator[] for(i = 0;i < ContainerOfOriginalShapes.size();++i) CopyViaContainerForLoopOp[i] = ContainerOfOriginalShapes[i]; Is there a problem with the way ptr_vector gets constructed when size argument is used for constructor with out a default value? When using the VC++ 7.1 compiler, I notice I get a compile warning "resolved overload was found by argument-dependent lookup" See full warning in following link: http://code.axter.com/boost_ptr_containers/compilerwarnings.txt Warning for following code: http://code.axter.com/boost_ptr_containers/BoostPolymorphicContainerTest.cpp http://code.axter.com/boost_ptr_containers/shape.h Is this a problem with VC++ 7.1 giving a false warning, or with the ptr_vector code design? I also welcome any additional input that may be useful for the article. Thank you

2 1

[test] please ignore
by Thorsten Ottosen 22 Nov '05

22 Nov '05

-Thorsten

1 0

Possibility of GLR parser combinators? Spirit-2?
by Korcan Hussein 22 Nov '05

22 Nov '05

I was briefly reading GLR Parsing in Haskell and found the library based off the paper HaGLR a GLR parser combinator framework and it got me thinking. With the advent of Phoenix-2 (very nice work guys) and FC++ being worked on (to be re-reviewed for boost in the not to distant future?) there could be a possibility of direct & easy port of the haskell code to C++ using either of these libraries as it's foundation and built & used in a simillar fashion to Spirit ala expression templates (maybe an idea for Spirit 2 even). There might be an issue with Phoenix-2 (i don't know enough about the subject to know for sure) since HaGLR is based on the idea of parser combinators i'd assume it takes a simillar approach to the Parsec library using monads, from what i gather from Phoenix-2 docs there currently is no plains to support monads. So what do you guys think? :-)

5 6

[Boost-bugs] [ boost-Bugs-1364007 ] [Regex] Splitting string: last empty token is "compressed"
by SourceForge.net 22 Nov '05

22 Nov '05

Bugs item #1364007, was opened at 2005-11-22 09:41 Message generated for change (Tracker Item Submitted) made by Item Submitter You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1364007&group_… Please note that this message will contain a full copy of the comment thread, including the initial issue submission, for this request, not just the latest update. Category: regex Group: None Status: Open Resolution: None Priority: 5 Submitted By: Nobody/Anonymous (nobody) Assigned to: John Maddock (johnmaddock) Summary: [Regex] Splitting string: last empty token is "compressed" Initial Comment: toto_lists(a)hotbox.ru Last empty token is "compressed" when using regex_token_iterator for splitting strings. BCB 5.64, boost 1.33.0 See the example, compare it with string_algo splitting: #include <boost\regex.hpp> #include <string> #include <vector> #include <iostream> #include <boost/algorithm/string/finder.hpp> #include <boost/algorithm/string/find_iterator.hpp> void with_stringAlgo() { using namespace std; vector<string> result; string to_split("&|&_field2_&|&_field3_&|&"); boost::split_iterator<string::iterator> i(to_split.begin(),to_split.end(),boost::first_finder("&|&")),j; for(;i!=j;++i) { result.push_back(boost::copy_range<std::string>(*i)); } cout<<"size is "<<result.size()<<endl; for(int i=0;i<result.size();++i) cout<<result[i]<<endl; } void with_regex() { using namespace std; using namespace boost; using namespace boost::regex_constants; string s("&|&_field2_&|&_field3_&|&"); boost::regex r("&\|&");//use &|& as delimiter boost::sregex_token_iterator i(s.begin(),s.end(),r,-1, match_default),j; // vector<string> v; copy(i,j,back_inserter(v)); // cout<<"size is "<<v.end()-v.begin()<<endl; copy(v.begin(),v.end(),ostream_iterator<string>(cout,"\n")); } int main() { with_stringAlgo(); with_regex(); } ---------------------------------------------------------------------- You can respond by visiting: https://sourceforge.net/tracker/?func=detail&atid=107586&aid=1364007&group_… ------------------------------------------------------- This SF.Net email is sponsored by the JBoss Inc. Get Certified Today Register for a JBoss Training Course. Free Certification Exam for All Training Attendees Through End of 2005. For more info visit: http://ads.osdn.com/?ad_id=7628&alloc_id=16845&op=click _______________________________________________ Boost-bugs mailing list Boost-bugs(a)lists.sourceforge.net https://lists.sourceforge.net/lists/listinfo/boost-bugs

1 0

[iterator] indirect_iterator->
by Orjan Westin 22 Nov '05

22 Nov '05

I have a question on the indirect_iterator, and what related standards say. I assume this class is primarily aimed for use in algorithms, which is fine. An algorithm typically dereference an iterator to do its thing, be it comparing or passing the dereferenced iterator on to a function or functor. This is great. But it doesn't handle direct usage well. If I have an indirect_iterator i for, say, a pointer to std::string, I cannot treat it as any other iterator to string and empty the string with this call: i->clear(); because the pointer operator is not changed. Instead you have to dereference it and use it as a reference: (*i).clear(); It seems strange to me that one way of getting hold of the iterated object is provided, but not the other. I couldn't find anything about this in a brief look through the standard, but I would expect it to want iterators yield the same results for both formats. Does anybody know if it does say anything on this subject, and why operator->() wasn't included in indirect_iterator? Cheers, Orjan

3 5

[smart pointers] Proposal to add some small utilities
by Ion Gaztañaga 22 Nov '05

22 Nov '05

Hi to all, While developing Shmem I wrote a smart pointer to represent relative pointers and now I'm experimenting with a multi-segment shared memory pointer to create growing multi-segment shared memory containers and allocators. Both pointers are based on boost smart pointer's interface (shared_ptr, etc...). In Shmem and possibly other generic libraries the developer wants pointer-independent code, so that the pointer type is a template parameter and the same code is used for relative, reference counted, intrusive, raw, or multisegment pointers. One of such uses, in Shmem, is for example, a memory allocator (a first fit, best fit algorithm) or an intrusive container. While developing pointer-independent code, however, I've seen that some features are present in boost (for example, get_pointer() function) and other features are missing, and I would like to propose them: 1. Overloads of static_pointer_cast, dynamic_pointer_cast, const_pointer_cast overloads for raw pointers, so that we can use pointer_xxx_cast in a generic way. 2. Add reinterpret_pointer_cast to complete cast family. 3. A protocol to obtain a pointer of the same type from any pointer. For example, if you want to generically use a pointer (for example, raw or relative pointer) in a class and you want to declare another pointer to other type (raw or relative, depending on the source): //typedef typename boost::pointer_to_other //<GenericPointer, NewType>::type MyNewPointer; //A new pointer to float values typedef typename boost::pointer_to_other <SmartPtr, float>::type MyFloatPointer; We should define how this protocol is implemented in boost smart pointers (for example, using overloading or a STL-like rebind mechanism) so that every boost smart pointer developer knows how to make its pointer type protocol-aware. Shmem uses rebind-like mechanism now, but the mechanism is not important. This way, a class can receive a void pointer like a template parameter and can be pointer-independent: template <class VoidPtr> class memory_manager { struct memory_block { //... }; typedef typename boost::pointer_to_other <VoidPtr, memory_block>::type BlockPtr; BlockPtr mp_free_list; }; I think these utilities are simple to develop and quite useful. What do smart pointer experts think? Regards, Ion

2 4

Boost regression notification (2005-11-22 [RC_1_33_0])
by Douglas Gregor 22 Nov '05

22 Nov '05

Boost regression test failures ------------------------------ Report time: 2005-11-22T14:05:07Z This report lists all regression test failures on release platforms. Detailed report: http://engineering.meta-comm.com/boost-regression/CVS-RC_1_33_0/developer/i… 4 failures in 2 libraries: multi_array (1) serialization (3) |multi_array| assign: vc-8_0 |serialization| test_list_ptrs_xml_warchive_dll: vc-8_0 test_non_default_ctor2_xml_archive: vc-8_0 test_shared_ptr_text_warchive_dll: vc-8_0

1 0