- Boost - lists.boost.org

[serialization] Binary iarchive invalid signature
by Gevorg Voskanyan 25 May '10

25 May '10

Hi, Binary iarchive is supposed to throw invalid_signature archive_exception when the archive content doesn't start with the expected signature. However in some cases a different behavior is observed when trying to load from an invalid archive. For example, if the supposed signature size being read ends up a huge number (can easily happen trying to load from an arbitrary file not created with Boost.Serialization), the program spends a few seconds on line 111 of basic_binary_iprimitive.ipp (s.resize(l);) allocating a big chunk of memory to read the signature into, then tries to read that many bytes from the stream, which, assuming the file is not that large, reads less bytes than requested so it all ends up throwing archive_exception(stream_error) at basic_binary_iprimitive.hpp:164. Another example is when the 'signature size' being read happens to be a larger number still, then s.resize(l); throws std::bad_alloc when in fact the user would want to get archive_exception(invalid_signature) instead. I've attached a patch for archive/impl/basic_binary_iarchive.ipp against Boost 1.43 version that fixes the problem for me. Comments? Thanks, Gevorg P.S. Boost 1.43, Win XP SP2, MSVC 8

2 2

[parameter] arg_list.hpp depends on maybe.hpp
by Andrey Semashev 25 May '10

25 May '10

I've been inspecting my code for a possible compilation speedup. The code uses Boost.Parameter and thus my attention shifted to it as well. I noticed that the parameter/aux_/arg_list.hpp header depends on parameter/aux_/maybe.hpp, which appears to be quite heavy due to its usage of Boost.Optional and other headers. From the code it looks like this "maybe" is only related to python binding, and the only thing arg_list.hpp needs from it is the maybe_base class, which is trivial. Is it possible to extract this class to a separate header and remove the dependency on maybe.hpp from arg_list.hpp or am I missing something? The change looks quite harmless. If I'm not missing anything, am I allowed to apply the change?

4 7

[config] Major upgrade to Boost.Config symbol visibility macros
by Beman Dawes 25 May '10

25 May '10

The upgrade to Boost.Config symbol visibility macros that has been under discussion has been committed to trunk. This lays the groundwork for resolving #2144, #2309, #3697. See https://svn.boost.org/trac/boost/changeset/62140 Thanks Jürgen Hunold for his patches and much other work on symbol visibility issues. The changeset also applied the changes to Boost.System to act as a testbed to ensure the changes are correct. That's a concern because any breakage in Boost.Config could break a lot of other stuff. We also want to get the Boost.Config changes right the first time as they kick off rebuilds of everything that uses the config system. The changes have been tested locally on Windows, Linux, and the Mac, and Boost.Filesystem has been retested on those systems as a reality check that unmodified user code continues to work. As trunk regression tests cycle and our confidence builds, we will start to suggest and/or apply changes to other libraries to make use of the new facilities. --Beman

7 9

[ptr_container] clone_ptr library
by raffimoni 25 May '10

25 May '10

Hello everybody, I did not find any boost library that handle with a situation described below, so that I request for your comments. Is there any interest in such a library? ===================== MOTIVATION ===================== Every time we need a class that owns some pointer (so that class is responsible for freeing it), we always need to define destructor, copy constructor and assignment operator. This is example implementation: class SomeClass { Base* ptr_; public: SomeClass( Base* ptr ): ptr_(ptr) {} ~SomeClass() { delete ptr_; } SomeClass(const SomeClass& s); //must be implemented const SomeClass& operator =(const SomeClass& s); //must be implemented }; In order to avoid a need of defining destructor we can use auto_ptr: class SomeClass { std::auto_ptr<Base> ptr_; public: SomeClass( Base* ptr ): ptr_(ptr) {} SomeClass(const SomeClass& s); //must be implemented const SomeClass& operator =(const SomeClass& s); //must be implemented }; But we need still implement copy constructor and assignment operator. Of course we can declare SomeClass noncopyable by deriving from boost::noncopyable to avoid this, but suppose this is not special case class and copying should be allowed. The best elegant solution to this problem is using a clone_ptr: class SomeClass { clone_ptr< Base > ptr_; public: SomeClass( Base* ptr ): ptr_(ptr) {} }; We can think about clone_ptr in two different ways: -as auto_ptr with different copy semantics (this is exactly what clone_ptr is) -as one-element ptr_container clone_ptr frees programmer from managing pointer the same way ptr_container for container of pointers does. So during copying the above class with clone_ptr behaves exactly like this: class SomeClass { //suppose this contains only one pointer boost::ptr_vector< Base > ptr_; public: SomeClass( Base* ptr ) {ptr_.push_back( ptr );} }; Notice that we do not need longer to define destructor, copy constructor and assignment operator. The default behaviour of these methods makes a deep copy of pointer and we should not be bothered by this anymore. clone_ptr uses the same concept of cloning that ptr_containers do, so new_clone function should be defined for Base class, if it is not and clone_ptr< Base > points to some type that derives from Base, an appropriate assert will fail the same way it does with ptr_container. Defining new_clone for Base makes it cloneable for ptr_pointer and clone_ptr simultaneously. ===================== SAMPLE IMPLEMENTATION ===================== class clone_ptr { private: T* ptr_; public: explicit clone_ptr(T* ptr) throw(): ptr_( ptr ) {} ~clone_ptr() { delete ptr_; }; clone_ptr(const clone_ptr<T>& ptr) { ptr_ = new_clone( *ptr.ptr_ ); //default is defined in ptr_container library assert( typeid( *ptr_ ) == typeid( *ptr.ptr_ ) ); } public: void reset( T* ptr = 0) throw() { if (ptr != ptr_) { delete ptr_; ptr_ = ptr; } } T* operator -> () const throw() { return ptr_; } }; Of course this implementation is not complete. Notice the way copy constructor is implemented. ===================== REMARKS ===================== clone_ptr should definitely not be used with stl containers (e.g. stl::vector< clone_ptr< Base > >) because of performance, ptr_containers are the best way to achieve such functionality. I'm looking forward to your feedback. Best Regards, Rafal Moniuszko

5 5

[test] [Patch] Unconditional use of non-standard functions.
by John Maddock 25 May '10

25 May '10

Boost.Test has just started using ::fedisableexcept and ::feenableexcept unconditionally - even though these are GLIBC specific extensions - which presumably is why Boost.Test is currently failing so many tests :-( I've just committed a patch to restrict the use of these functions to GLIBC platforms that have the GNU extensions enabled: see https://svn.boost.org/trac/boost/changeset/62194/trunk/boost/test/impl/exec…. Please feel free to change the logic for any other platforms that may have these functions.... but for now, hopefully we can get back to testing Boost again! Cheers, John.

1 0

[fusion] flat fusion sequence of all the node of a type
by vicente.botet 25 May '10

25 May '10

Hi, I want to create a fusion sequence that give access to all the nodes of a struct/class/array that inherit from a given class TO. If we remove this constraint, the fusion sequence can be seen as flattening the tree of nodes in a given order. Once I will have this I can use filter_if<is_base_of<TO,_> > Given class TO; struct W { TO t; }; struct X : TO { // direct TO* to_ptr; char c; }; struct Y : TO { // direct with embeddeds X x; int i; }; struct Z { // with embeddeds Y y[2]; TO t; W w; }; Z z; the flat view corrsponds to z.w.t, z.w, z.t, z.y[0].x.to_ptr, z.y[0].x.c, z.y[0].x, z.y[0], z.y[1].x.to_ptr, z.y[1].x.c, z.y[1].x, z.y[1], z.y, z The filtered view corrsponds to Y z.y[0] X z.y[0].x Y z.y[1] X z.y[1].x TO z.t TO z.w.t If I have fusion sequence struct adaptors for W, X, Y, Z. Is there a simple and run-time efficient way to get this fusion sequence adaptor or I need to do something similar to the struct adaptor? Have others had a similar need already? Best, Vicente P.S. For my personal usage I need only a forward sequence.

1 0

[MSM] compiling MSM with GCC 4.4.1 and -fno-rtti
by Richard Szabo 24 May '10

24 May '10

Hi All I try to compile boost msm in an embedded target with gcc 4.4.1 for performance reasons I want to disable RTTI. I'm having problems because backend is using boost::any which is uses RTTI. Is it possible to remove this dependency ?. or introduce some compile switch like in shared pointers .... BOOST_NO_STD_TYPEINFO Thanks in advance. Richie

1 0

Maintenance Patches and cleaning up trac
by Christopher Jefferson 24 May '10

24 May '10

Having recent been browsing through boost trac I have noticed that it seems to be getting a little clogged with some very simple patches. Examples of simple patches: https://svn.boost.org/trac/boost/ticket/3402 - simple documentation patch, 6 months old, duplicated. https://svn.boost.org/trac/boost/ticket/3669 - add stdint.h to thread, fixes comeau, 6 months old. https://svn.boost.org/trac/boost/ticket/4060 - boost thread support for BOOST_NO_IOSTREAM I notice these 3 are in 'thread', and several others are in 'filesystem'. I do not wish to maintain these libraries myself, and it appears no-one else is coming forward to keep these libraries ticking over on a day-to-day basis. I would however personally be happy (possibly with help) to look at applying these patches, and applying the to trunk, and then release in time, assuming they pass regression testing. I would suggest a policy something like the following: 1) If a patch is considered to be "trivially obvious", mark the patch in trac, by leaving a comment, as a 'maintenance patch'. 2) If no-one disagrees within some period of time (2 weeks? 3?), then the patch is applied to trunk. If it passes regression testing it is then moved to release. Any thoughts on the topic? Chris

8 14

Re: [boost] [Boost-docs] Maintaining boost.org
by David Abrahams 24 May '10

24 May '10

At Tue, 25 May 2010 00:32:02 +0800, Joel de Guzman wrote: > > On 5/24/10 10:41 PM, Rene Rivera wrote: > > > > One idea I had, before Boostcon when the whole where should Boost go > > discussion came up, was to move to using something like the Drupal model > > for sub-projects <http://drupal.org/project/Modules>. Particularly how > > the projects each have their own web area. And also how the groups > > section area of Drupal works <http://groups.drupal.org/>. > > > > Note, I'm not suggesting Drupal (even if I do currently use it). Just > > some of the structure they employ for projects since it does mirror the > > conglomerate project structure Boost has. > > Last November, we switched the Spirit site to Wordpress. > > http://boost-spirit.com/home/ > > After having tried Drupal, Joomla, and a couple of other > CMS, we settled on Wordpress because of its simplicity, > low entry barrier and low maintenance. It's just perfect for > people on the go, like us. Agreed; WP rocks. Consider WPMU (or WP3.0) and/or BuddyPress also. I am using those at http://techarcana.net, although I haven't had time to make everything really nice yet. -- Dave Abrahams Meet me at BoostCon: http://www.boostcon.com BoostPro Computing http://www.boostpro.com

1 0

How to build Boost with C++0x compilers
by vicente.botet 24 May '10

24 May '10

Hi, I don't know how to build Boost with C++0x compilers. Which option must be used with bjam? Should the user.config file be modified? Can someone help me? Best, Vicente

5 6