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

[iterator] reverse_iterator from iterators that return references to data in self.
by Seb Martel 15 Jul '06

15 Jul '06

Hello, given something like: class ProxyIterator : public boost::iterator_facade<ProxyIterator, Costly, boost::random_access_traversal_tag> { public: ProxyIterator(FactoryOfCostly* pF, int i ) : m_pFactory( pF ), m_index( i ) {} private: friend class boost::iterator_core_access; void decrement() {--m_index;} // ... and others Costly& dereference() const { m_value = m_pFactory->MakeCostlyAtIndex( m_index ); return m_value; } private: FactoryOfCostly* m_pFactory; int m_index; mutable Costly m_value; }; I cannot use this ProxyIterator with a boost::reverse_iterator adaptor because the implementation of reverse_iterator's dereference dereferences a temporary: // from reverse_iterator.hpp typename super_t::reference dereference() const { return *boost::prior(this->base()); } and in ProxyIterator, this dereference returns a reference to an internal member, m_value, that will immediately vanish out of existence. Bummer. What if reverse_iterator carried as member the decremented iterator instead of decrementing to a temporary on a dereference? In the current reverse_iterator implementation, decrement must behave properly on an end iterator, but not on a begin iterator. Can a begin iterator be decremented and still used for comparison purposes? -seb

2 1

force inline
by Artem Alimarine 15 Jul '06

15 Jul '06

Hi all, Is there a macro in boost for something like MSVC __forceinline. Currently I have in the project: #if defined __GNUC__ #define FORCE_INLINE __attribute__((always_inline)) #define NO_INLINE __attribute__((noinline)) #elif defined _MSC_VER #define FORCE_INLINE __forceinline #define NO_INLINE __declspec(noinline) #else //#error "no inlines" #define FORCE_INLINE inline #define NO_INLINE #endif I am wondering whether I can get rid of it and use boost for it? Artem

3 3

boost::lambda performance
by Aleksey Loginov 15 Jul '06

15 Jul '06

Where I can find benchmark sources for boost::lambda?

1 0

Re: [Boost-users] Macro to contruct/allocate a shared_ptr?
by David Klein 15 Jul '06

15 Jul '06

boost-users-bounces(a)lists.boost.org wrote: > Hi Dave. That code went over the top of my head. Do you mind > giving me an example of the usage? > > On 7/14/06, David Klein <d.klein(a)ascaron.com> wrote: >> [snip] >> >>> I just have a problem with having the user manage the memory. >>> ie I'd rather do the following: >>> >>> shared_ptr<Test> p(1, 'a', "hello"); >>> or: >>> shared_ptr<Test> p = newptr<Test>(1, 'a', "hello"); >>> >>> instead of: >>> >>> shared_ptr<Test> p(new p(1, 'a', "hello")); >>> >>> but i guess that's a limitation of the language. to get >> >> hi, >> >> have a look at shared_new.zip in the vault. >> >> http://tinyurl.com/khd8v >> >> -- >> HTH dave well, i just tried to compile the shared_new_test.cc, but it failed using vc71 and vc8 (boost 1.33.1). anyway, its quite similiar to the code Roman Perepelitsa posted. a short example using his newptr: struct X { X( int _x, double _y, const std::string & _z ) { ... } } shared_ptr<X> x = newptr<X>(42, 42.0, "hello world");

2 1

stlport?
by bringiton bringiton 14 Jul '06

14 Jul '06

i'm porting my app so that it now uses stl and boost. can someone please give advise on what stl i should be using? i hear about stlport all the time, but not sure why it is used? ie is it better that the version that comes with the compilers? im using ms vc++ 6 at the moment and plan to upgrade to 7.1. should i use the default stl, or install stlport?

3 5

Spirit based shell command interpreter
by George M. Garner Jr. 14 Jul '06

14 Jul '06

Has anyone implemented a shell command interpreter using the boost spirit library? Yes, I am aware of jam/bjam. There is also a calculator example available on the web and the process doesn't seem all that difficult. I just want to see whether this specific application has already been done (before I reinvent the wheel). Regards, George.

1 0

Re: [Boost-users] Boost Test usage
by Matt Schuckmann 14 Jul '06

14 Jul '06

Actually right now we are using 1.32.0, I've downloaded 1.33.1 but I'll have to convince the rest of my group that it's worth it to update to 1.33.1 or I maybe able to use 1.33.1 just for the test library, I don't know yet. IMO Example 3 doesn't really show building a multilevel hierarchy it just shows how to derive from test_suite and add that to the master test_suite. Yes you could create multiple test_suites and add them together in interesting ways but what does that buy you? Now that I see the detailed report level I see some use for hierarchy's, maybe the documentation should mention this and provide some example output. It took me a while to find the command line documentation you might consider making it stand out more in the documentation and maybe even add a standard "-h --help" option so that the test executables are self documenting, if your using boost::program_options this is pretty easy. Thanks for quick feedback I'll see what I can do. I really hate creating a bunch of executables (especially in MS .NET 2003) but I also don't really want to always run all my tests I'll watch for version 1.35. Thanks Matt S.

1 0

A question about shared_ptr
by Zhuo Hao 14 Jul '06

14 Jul '06

In file <boost/shared_ptr.hpp>, I find the following function: /begin quote template<class Y, class D> shared_ptr(Y * p, D d): px(p), pn(p, d) { detail::sp_enable_shared_from_this( pn, p, p ); } /end quote If I use it as follows:"boost::shared_ptr<Investment> pInv(0,getRidOfInvestment)",it will not compile.But when I use:"boost::shared_ptr<Investment> pInv(static_cast<Investment*>(0),getRidOfInvestment)",it can be compiled. The constructor's first parameter is Y*,i.e.,a pointer to Y.And 0 can be interpreted as a pointer to Y,too. Why can't zero be treated as a pointer to Y here? I will be appreciative if anyone can give me some suggestions.Thank you! hzhuo July,14th

3 2

Re: [Boost-users] serialization-lib add user info to XML format
by Eduardo Ferro Aldama 14 Jul '06

14 Jul '06

Hi! Thank very much for the help I am study the lib implementation, and trying to extend xml_oarchive (as you suggest)... Bye >I don't know that I can offer you a good solution. But here is what I can >come up with. > >First a couple of relevant facts about the serialization library. > >a) generally it saves in the archive only that information absolutely >required >to be able to recover the data. This is basad on design/efficiency >considerations. > >b) In some cases it stores "optional" information. That it is, the archive >may be written to save and recover this information. Normally these >are just stubbed out as no ops. > >c) In xml archives, information about object and class id and their >references is included in the xml. I'm don't remember whether its done >in the "optional" types or not. > >c) class name was not considered as one of these optional types > i) it wasn't deemed necessary > ii) it is not always available. Many types (e.g. primitives) are >serialized >without calling the library code. For these cases the "optional" data >is not called. > >So back to your situation. > >What is really needed for you is the following: > >a) some sort reflection for C++ so that you can write something like >t.name(). >b) derive from xml_oarchive to invoke t.name() and insert it into >the xml stream output. Examples of deriving from one archive >type to produce another are in the documentation. > >So the problem is "reduced" to making/finding something like t.name() >which returns a text string which can be inserted into the output stream. > >This already exists on one form - the RTTI implementation uses >type_id(t).name to return a pointer to a name. This may be sufficient >for your needs. The only problem with it is that > >a) the name will be mangled to a unique C++ name. This may not always >be user friendly. >b) the name will be mangled differently for each compiler. > >If you can work with the above limitations - perhaps by massaging the >mangled name to something readable - this can probably be made to work >for you. > >If not - you can use the name passed by BOOST_CLASS_EXPORT which >was "invented" to address the issues above. The only problem is that >BOOST_CLASS_EXPORT might not bedefined for all your types. > >You should have enough here to craft a solution to your problem. > >Good Luck > >Robert Ramey > Ce message ?lectronique et tous les fichiers attach?s sont confidentiels et destin?s exclusivement ? l'usage de la personne ? laquelle ils sont adress?s. Si vous n'?tes pas le destinataire de ce message, merci d'avertir ou de le retourner ? son ?metteur et d?truire ce message ?lectronique et tous les fichiers attach?s de votre syst?me informatique. La publication, l'usage, la distribution, l'impression ou la copie non autoris?e de ce message et des attachements qu'il contient sont strictement interdits. This e-mail and any attachment are confidential and intended solely for the use of the individual to whom it is addressed. If you are not the intended recipient, please telephone or email the sender and delete this message and any attachment from your system. Unauthorized publication, use, dissemination, forwarding, printing or copying of this e-mail and its associated attachments is strictly prohibited. Este correo electr?nico y todos los ficheros adjuntos son confidenciales y destinados exclusivamente al uso de la persona a la cual han sido remitidos. Si Usted no es el destinatario del mensaje, agradecemos advierta al remitente y elimine el mensaje y sus adjuntos de su sistema. La publicaci?n, distribuci?n, impresi?n o copia no autorizada de este mensaje y de sus adjuntos queda estrictamente prohibida.

1 0

Installation on Macintosh
by John Christopher 14 Jul '06

14 Jul '06

Hello I am trying to install boost on a macintosh (Tiger). I downloaded the boost-jam-3.1.13-1-macosxppc.tgz file and uncompressed it in a folder: /Developer/Libraries/boost there I installed bjam and in the Terminal I typed: ./bjam "-sTOOLS=darwin" install everything things run smoothly until I get a lot of error messages like: ...skipped <@boost!include!boost-1_33_1!boost!wave!util>symbol_table.hpp for lack of <directory-grist>/usr/local/include/boost-1_33_1/boost/wave/util... ...skipped <@boost!include!boost-1_33_1!boost!wave!util>time_conversion_helper.hpp for lack of <directory-grist>/usr/local/include/boost-1_33_1/boost/wave/util... ...skipped <@boost!include!boost-1_33_1!boost!wave!util>transform_iterator.hpp for lack of <directory-grist>/usr/local/include/boost-1_33_1/boost/wave/util... ...skipped <@boost!include!boost-1_33_1!boost!wave!util>unput_queue_iterator.hpp for lack of <directory-grist>/usr/local/include/boost-1_33_1/boost/wave/util... and the installation process ends with ...failed updating 8 targets... ...skipped 3649 targets... ...updated 709 targets... Any idea on what I am doing wrong? Thanks in advance

1 1