- Boost-users - lists.boost.org

boost::posix::time_duration
by Michael Thurston 02 Nov '07

02 Nov '07

[boost][date_time] Hi there, I have a question about the boost::posix::time_duration class. It has accessors such as: - hours() - minutes() that return a data type of long. I would have expected it to return a datatype of boost::int64_t (or boost::int32_t) because could there be a risk that the return value would cause an overflow on an 8 bit embedded platform. Regards, Michael

1 0

C++ & Enterprise Solutions
by Stefan Kok 02 Nov '07

02 Nov '07

Dear List If this is not the correct list to post to. Then please forgive my ignorance. I have been admiring the work of Boost and its contributors for some time now. My questions / comments : 1) Would their be any benefit to the community, developing a specification for C++ similar to SUN's J2EE and JEE specification ? I do realize this would require a mammoth effort as well as commitment in terms of resources. 2) Is it possible to establish such a project under the open source banner ? Given the commitment and resources required. 3) Would this be reinventing the wheel ? I know SUN & Java got there first (the J2EE thing). But C++ does have some very powerful features which makes it unique in some ways, templates for one. Thank you in advance. Stefan.

1 0

[Boot.Pool] Opinions about inconsistent pointer types
by David Baird 01 Nov '07

01 Nov '07

Hi, I was looking through the Pool code and noticed that it presents inconsistent pointer types to the user ("void *" versus "char *"). For example, if the user wants to implement his/her own user allocator, they have to write it like, e.g. (using "char *" as the pointer type): struct default_user_allocator_new_delete { typedef std::size_t size_type; typedef std::ptrdiff_t difference_type; static char * malloc(const size_type bytes) { return new (std::nothrow) char[bytes]; } static void free(char * const block) { delete [] block; } }; However, other classes return "void *" as their pointer types, e.g. singleton_pool: template <typename Tag, unsigned RequestedSize, typename UserAllocator, typename Mutex, unsigned NextSize> struct singleton_pool { static void * malloc() { pool_type & p = singleton::instance(); details::pool::guard<Mutex> g(p); return p.p.malloc(); } } Is it really important that the types be different? I just think it might be more elegant to use "void *" in more places - it would make more things the same and be similar to e.g. "std::malloc/free". -David

1 0

Serialization of C dynamic data
by Lorenzo Cavallaro 31 Oct '07

31 Oct '07

Hi, I'm trying to serialize the following piece of code (a data structure whose object size is dynamically allocated and it should use the correct size/len which has been serialized). typedef struct data { char *payload; size_t len; } data_t; vector<data_t *> v; So I added the corresponding serialize function to the structure definition void serialize(Archive & ar, const unsigned int file_version) { ar & len; for (int i = 0; i < len; i++) ar & payload[i]; } Moreover, I added a default constructor which is invoked by the serialization library whenever the object has to be unserialized data::data() { /* no error checking */ payload = (char *)calloc(1, X); len = X; } Now, of course X is fixed. I'd like to reuse the serialized 'size' (ar & size) as argument to the calloc to be able to create an object with the right size instead of having to approximate (exceeding, of course) its real dimension. Do you have any idea how to get it work the way I'd like to, please? -- if it's possible, of course. Nevertheless to say that with a right X everything works fine. I was thinking to iterate over all the elements, after the unserialization step is over, to reallocate payload using the right len... but it sucks :-) TIA, bye Lorenzo -- Lorenzo `Gigi Sullivan' Cavallaro <sullivan(a)security.dico.unimi.it> GPG key at http://security.dico.unimi.it/~sullivan/sullivan.asc Until I loved, life had no beauty; I did not know I lived until I had loved. (Theodor Korner) See the reality in your eyes, when the hate makes you blind. (A.H.X)

2 1

[Boost users] [interprocess]: static_pointer_cast on offset_ptr
by Lars Hagström 31 Oct '07

31 Oct '07

Hi, In offset_ptr.hpp the static_pointer_cast<> &co are commented out. Why is this? I need something of the kind to do casts from a void* to a bool*, for example. Is there something I am missing? Lars

3 4

REGEXP - Serializing regexp ?
by Jacques-Olivier Goussard 31 Oct '07

31 Oct '07

Hi I'm trying to use the boost-regexp package and I'm facing an issue with this. I'm basically applying a replace algo on a list of strings using a huge list of regular expressions. It seems that the boost-regexp API allows me only to either load all regular expressions in memory (i.e creating the regexp at process start) or load one regexp at a time but re-compile it each time. Is there a way to save to disk the compiled version of the regexp (so that the regexp package would not have to re-create the statemachine) ? If not, what would be the best approach for doing this ? Thanks /jog

4 9

Re: [Boost-users] Regex: Invalid preceding regular expression, problem
by Brian Krahmer 31 Oct '07

31 Oct '07

> Nope there's a bug in the options you are using - boost::regex::extended > enables the "POSIX extended" flavour of regular expressions - and these > don't support non-greedy repeats. Use boost::regex::perl if you want to use > these (since you are using Perl-style regexes). > > HTH, John. > That helps very much. Thanks for clearing that up for me, John. brian

1 0

Preprocessor lib question -- how to fold a macro that produces comma in its output across a sequence?
by Yu Di 31 Oct '07

31 Oct '07

Hi, I am trying to use the preprocessor lib to generate something like Cons<a, Cons<b, Cons<c, Cons<d, Nil> > > > from (a) (b) (c) (d) I tried something like #define START(x) Cons<x,Nil> #define ADD(s,state,x) Cons<x,state > #define GENERATE(x) BOOST_PP_SEQ_FOLD_RIGHT(ADD,START(BOOST_PP_SEQ_HEAD(BOOST_PP_SEQ_REVERSE(x))),BOOST_PP_SEQ_POP_BACK(x)) but immediately got into trouble with the commas. The best I can do is to add parentheses around the content of START and ADD, but that generated (Cons<a, (Cons<b, (Cons<c, (Cons<d, Nil>) >) >) >) which is not valid. Is there any way I can achieve the desired output? Thanks, Di, Yu 10.30 __________________________________________________ Do You Yahoo!? Tired of spam? Yahoo! Mail has the best spam protection around http://mail.yahoo.com

2 1

FILE boost\preprocessor\arithmetic\dec.hpp
by question boost 31 Oct '07

31 Oct '07

Hi, I' ve the following error message: FILE C:\Documents and Settings\src\boost\preprocessor\arithmetic\decc.hpp LINE 17 boost/preprocessor/config/config.hpp no such file or directory So I look at decc.hpp, it contains the line # include <boost/preprocessor/config/config.hpp> and I look at the directory C:\Documents and Settings\src\boost\preprocessor\config, it contains the file config.hpp IS ANYONE GOING TO TELL ME WHAT IS THE PROBLEM WITH THOSE BOOST FILES? THANK YOU VERY MUCH David _________________________________________________________________ Vous êtes plutôt Desperate ou LOST ? Personnalisez votre PC avec votre série TV préférée ! http://specials.divertissements.fr.msn.com/SeriesTV.aspx

1 0

graph python bindings : unpickle hangs
by John Reid 31 Oct '07

31 Oct '07

Hi, Unpickling an empty boost.graph seems to hang. E.g. In [1]: import boost.graph as bgl In [2]: g=bgl.Graph() In [3]: import cPickle In [4]: cPickle.dumps(g) Out[4]: 'cboost.graph._graph\nGraph\np1\n(tRp2\n((dp3\nI0\n(l(d(dtb.' In [5]: cPickle.loads(cPickle.dumps(g)) *** never returns *** I'm using the latest SVN revision of bgl-python. John.

1 0