May 2004 - Boost - lists.preview.boost.org

Re: [boost] Re: static sized strings
by Reece Dunn 25 May '04

25 May '04

Rob Stewart wrote: >From: "Reece Dunn" <msclrhd(a)hotmail.com> > > > > What I am interested in is a comparison between > >* I think that data() would be better named "buffer." The This would go with the renaming you suggest for functions that are the same in the base and implementation classes (hiding issue). >* Should fixed_string support zero-length strings? You can have zero-length strings in my implementation (this is what fized_string() defaults to). What you *can't* have is a zero-*capacity* string, which is the same as CharT[ 0 ]. >* I'm not sure that there's enough value in > const_iter_offset()/iter_offset() versus begin()/end() (const > and non-const) in the derived class. You eliminate four > trivial functions, but you may eliminate opportunities for > optimizations. (I'm thinking there may be derived types that > have a better way of computing end and that returning the > result of adding zero to a pointer is slower than simply > returning the pointer.) This is true. I will probably expand the interface to the four begin/end functions. >* fixed_string doesn't have a const overload of at(). I shall fix this. > Calling fixed_string's at() "at" seems wrong. You should take the tack > of requiring a derivate of detail::basic_string_impl to provide > uniquely named, implementation functions that basic_string_impl > uses to provide the public interface. You could even suggest > that such functions be made private with the basic_string_impl > specialization made a friend. Your current approach will > generate many "hiding" warnings, too. I will rename the conflicting names. >* fixed_string's at() shouldn't test the length and throw an > exception since detail::basic_string_impl::at() does it and > detail::basic_string_impl::operator[] calls it. fixed_string tests against capacity as a measure against [] usage that would cause buffer overflow, e.g. fixed_srting< 5 > str; str[ 10 ] = 'a'; // ok: exception vs char str2[ 5 ]; str[ 10 ] = 'a'; // foobar >* fixed_string::swap() is missing. >* detail::basic_string_impl::get_allocator() expects > string_type::get_allocator() but that isn't listed in the > requirements levied on the derived type and isn't defined in > fixed_string. These are currently missing becase (1) swap requires a deep copy and haven't implemented that yet; (2) allocators are not used by fixed_string (no need to allocate any data) so it is not implemented (I am considering if/how it should be implemented). NOTE: it should compile despite these missing from fixed_string unless you attempt to call those functions. >* I wouldn't expect fixed_string's StringPolicy to govern how to > determine the length of a string passed to a mf as a CharT *. It does not. string_type::length() returns the value of the string held in the class (return a variable value, defer to container, use traits_type, etc.). traits_type::length( const CharT * ) calculates the length of a string passed to it, wher traits_type is a char_triats compatible interface. >* If I kept looking, I'm sure I'd find many more things, but you > are more interested in philosophical differences, so I'll stop > here. Look at the end of the message for my summary. I am actually interested in both. The feedback is appreciated and I will use it to improve the code. >I think John's approach is better, only because it does less type >indirection. It relies on the more ordinary pvf approach to the >core functionality. Otherwise, the two libraries provide the >same functionality (or can when complete). To simplify my design, I could implement basic_string_impl using the policy approach that flex_string uses. The problem with this is that it requires a usage like: flex_string< fixed_string< 10, char > > str; so in order to get the fixed_string< n > syntax, I would need two classes instead of just the one: this was something I was trying to move away from, but the approach I use has several problems: [1] the code is less optimal (7.5k vs 5k on MS VC 7.1) in my approach, due to the indirection used; the flex_string policy approach would allow compilers to aggressively inline. [2] the implementation functions are exposed in the interface; a policy approach would allow the implementation to be hidden because it can be derived from it like flex_string. [3] in my current approach, I can't use typedefs from the StringPolicy in basic_string_impl (they are not defined yet); seperate classes will solve this. There may be others to this list. So I am going to change to the flex_string approach and see what happens :). Regards, Reece _________________________________________________________________ Express yourself with cool new emoticons http://www.msn.co.uk/specials/myemo

2 1

Re: [boost] Re: [Test] Intel 8 seems to need workaround
by JOAQUIN LOPEZ MU?Z 25 May '04

25 May '04

----- Mensaje original ----- De: Gennadiy Rozental <gennadiy.rozental(a)thomson.com> [...] > How presence of unrelated class foo affects access in class AA? That's the funny thing about it. In your code, from where I came up with this, the error with p_stages_amount pops out due to the presence of other, unrelated objects of type class_property. > Is it important that method is called foo also? > No, it is not relevant. You might want to play with the snippet using the online Comeau compiler at: http://www.comeaucomputing.com/tryitout/ Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

1 0

Re: [boost] Re: static sized strings
by Reece Dunn 25 May '04

25 May '04

Rob Stewart wrote: >Sure, but you've both jumped off on a tangent unrelated to the >purpose of the original query. See my reply to Dave's message. I agree (thanks for the info, though :)). What I am interested in is a comparison between my implementation at: boost-sandbox/boost/fixed_string and John Nagle's implementation at: http://www.animats.com/source if possible. Regards, Reece _________________________________________________________________ It's fast, it's easy and it's free. Get MSN Messenger today! http://www.msn.co.uk/messenger

5 4

Fwd: Bug#245893: Boost.Python looks broken on debian/arm
by Domenico Andreoli 25 May '04

25 May '04

hello there, this is the the last important bug debian packages of Boost 1.31.0 have. this is copy of the message i already sent to debian-arm ml and Boost.Python ml with no response. i'm forwarding it here, on boost developers ml, as last chance to settle this out and finally make 1.31.0 reach debian sarge. it may also look like a problem in MPL. cheers domenico ----- Forwarded message from Domenico Andreoli <cavok(a)debian.org> ----- Date: Mon, 26 Apr 2004 23:46:10 +0200 From: Domenico Andreoli <cavok(a)debian.org> To: c++-sig(a)python.org, debian-arm(a)lists.debian.org Cc: 245893(a)bugs.debian.org Subject: Bug#245893: Boost.Python looks broken on debian/arm hi all, debian arm autobuilder fails to build the only two applications in main archive which make use Boost.Python, they are kdeedu [0] and regina-normal [1]. i'm trying to understand where the chain of errors reported in debian bug #245893 [2] really starts, problem is that i still have no simple program to reproduce this bug and not arm machine to play with. i'm asking to debian-arm mailing list some help reproducing the bug and to Boost.Python mailing list some help understanding the outcome. here follows what i think is the culprit error... /usr/include/boost/python/detail/def_helper.hpp: At global scope: /usr/include/boost/python/detail/def_helper.hpp: In instantiation of `boost::mpl::apply1<boost::python::detail::is_reference_to_member_function_pointer<boost::mpl::_1>, const char* const&>': /usr/include/boost/python/detail/def_helper.hpp:78: instantiated from `boost::python::detail::tuple_extract_base_select<boost::tuples::tuple<const char* const&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, boost::python::default_call_policies, boost::python::detail::keywords<0>, const char*, void (boost::python::detail::not_specified::*)(), boost::tuples::null_type, boost::tuples::null_type>, boost::python::detail::is_reference_to_member_function_pointer<boost::mpl::_1> >' /usr/include/boost/python/detail/def_helper.hpp:173: instantiated from `boost::python::detail::tuple_extract<boost::tuples::tuple<const char* const&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, boost::python::default_call_policies, boost::python::detail::keywords<0>, const char*, void (boost::python::detail::not_specified::*)(), boost::tuples::null_type, boost::tuples::null_type>, boost::python::detail::is_reference_to_member_function_pointer<boost::mpl::_1> >' /usr/include/boost/python/detail/def_helper.hpp:173: instantiated from `boost::python::detail::default_implementation_extract<boost::tuples::tuple<const char* const&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, const boost::python::detail::not_specified&, boost::python::default_call_policies, boost::python::detail::keywords<0>, const char*, void (boost::python::detail::not_specified::*)(), boost::tuples::null_type, boost::tuples::null_type> >' /usr/include/boost/python/detail/def_helper.hpp:173: instantiated from `boost::python::detail::def_helper<const char*, boost::python::detail::not_specified, boost::python::detail::not_specified, boost::python::detail::not_specified>' /usr/include/boost/python/class.hpp:304: instantiated from `boost::python::class_<T, X1, X2, X3>& boost::python::class_<T, X1, X2, X3>::def(const char*, F) [with F = void (PythonOutputStream::*)(), T = PythonOutputStream, X1 = boost::noncopyable, X2 = boost::python::detail::not_specified, X3 = boost::python::detail::not_specified]' pythoninterpreter.cpp:89: instantiated from here /usr/include/boost/python/detail/def_helper.hpp:78: error: no class template named `apply' in `struct boost::python::detail::is_reference_to_member_function_pointer<boost::mpl::_1> ' any hint? thanks domenico [0] http://buildd.debian.org/fetch.php?&pkg=kdeedu&ver=4%3A3.2.2-1&arch=arm&sta… [1] http://buildd.debian.org/fetch.php?&pkg=regina-normal&ver=4.1.1-1&arch=arm&… [2] http://bugs.debian.org/245893 ----- End forwarded message ----- -----[ Domenico Andreoli, aka cavok --[ http://people.debian.org/~cavok/gpgkey.asc ---[ 3A0F 2F80 F79C 678A 8936 4FEE 0677 9033 A20E BC50

2 2

Where is pool::get_requested_size() ??
by Martin Pärtel 25 May '04

25 May '04

This function can be found in documentation of pool: size_type get_requested_size() const; I searched for it in the header file in the latest 1.31.0 release and in the main CVS and was not able to find it. A google search turned up a post on the Boost mailing list dated March 2003! http://lists.boost.org/MailArchives/boost/msg45833.php A fix was promised in the reply but as far as I can see it is still not fixed (I may be wrong). Does anyone know anything about this? Thanks

1 0

RE: [boost] STL algorithm for containers (was Re: Re: boost/test/ detail/algorithm.hpp: request forpromotionand/or fast review)
by Rozental, Gennadiy 25 May '04

25 May '04

> Gennadiy Rozental wrote: > > Why don't we provide container versions of all stl algorithms: > > > > template<typename C,typename P> > > inline > > for_each( C const& c, P pred) // or non const ref?? > > { > > // unwrap here used to allow by_ref( c ) usage for > mutable predicates > > std::for_each( begin( unwrap( c ) ), end( unwrap( c ) ), P ); } > > This depends who do you mean by "we" ;-) There's a bunch of container > algorithms in the sandbox: > > http://tinyurl.com/yv6q3 > > and I use them quite a lot. > > - Volodya Oh! I missed it somehow. Is there plans for review request? I would second it. Gennadiy.

3 3

class properties: request for promotion and/or fast review.
by Gennadiy Rozental 25 May '04

25 May '04

Hi, I am using simple class properties facility for some time and found it quite useful. Actually it's nothing fancy - simple helper classes to eliminate repetitive setter/accessor methods (and no need for language extensions). Essentially if you need simple readwrite property with trivial access methods in you class you would instead of writing: T member; void set_member( T const& ) {...} T const& get_member() {...} would write: readwrite_property<T> member; Now any time you need to get a member value you would write T t = obj.member; while if you need to update the value: obj.member.value = t; It's not more nor less safe as usual get/set scheme. If you need readonly object property that get initialized in constructor and never updated you could use readonly_property template for example like this: class node { public: readonly<std::string> p_name; .... }; Now everywhere (almost) you could just use n.p_name. To get an access to name member functions you could use operator-> like this: n.p_name->size(); And nobody would be able to update readonly property value. Finally if you need to provide a write access for readonly property you could use another form of readonly_property template. Unfortunately during language limitations (inability to declare template parameter as friend) you would need to use macro, for example like this: class foo { public: BOOST_READONLY_PROPERTY( int, (foo)(Friend2) ) p_value; ... }; Now only class foo and Friend2 have a write access to the value of the property p_value; If boosters find this functionality worth reusing I am ready to prepare document page describing it's usage. If we don't find this header eligible for review I would like to ask then for permission to at least put these templates in boost namespace (currently boost::unit_test), for it inconvenience for me to keep it in deep namespace for rest of my current development. Header you could see here: http://tinyurl.com/38mcm Regards, Gennadiy.

5 18

RE: [boost] Bang for the buck (reducing warnings)
by Rozental, Gennadiy 24 May '04

24 May '04

> Would someone like to commit this fix, please? Ok. Will do tonight. Gennadiy.

1 0

Bang for the buck (reducing warnings)
by Marshall Clow 24 May '04

24 May '04

I found that adding one line to "boost/libs/test/src/unit_test_log.cpp" reduced the number of warnings in the Mac regression test by 80% Here's the change: >diff -r1.22 unit_test_log.cpp >318a319 > > case invalid_log_level: Here's the warning (over and over again!): /Users/marshall/boost/boost/libs/test/src/unit_test_log.cpp:320: warning: enumeration value `invalid_log_level' not handled in switch Would someone like to commit this fix, please? -- -- Marshall Marshall Clow Idio Software <mailto:marshall@idio.com> I want a machine that thinks I'm more important than it is, and acts like it. -- Eric Herrmann

1 0

[random] missing seed() method
by Mario Ruetti 24 May '04

24 May '04

Hi I sent this message already some month before. I made some corrections to the boost random library because some generators don't fulfil the wg-21 specifications about the seed() method. I add the missing seed() methods to the following generators * linear_congruential * linear_feedback_shift * additive_combine * inversive_congruential * discard_block * shuffle_output Greetings Mario Index: linear_congruential.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/linear_congruential.hpp,v retrieving revision 1.15 diff -r1.15 linear_congruential.hpp 219c219 < void seed(int32_t x0) { lcf.seed(cnv(x0)); } --- > void seed(int32_t x0=1) { lcf.seed(cnv(x0)); } Index: linear_feedback_shift.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/linear_feedback_shift.hpp,v retrieving revision 1.7 diff -r1.7 linear_feedback_shift.hpp 83c83 < void seed(UIntType s0) { assert(s0 >= (1 << (w-k))); value = s0; } --- > void seed(UIntType s0 = 341) { assert(s0 >= (1 << (w-k))); value = s0; } Index: additive_combine.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/additive_combine.hpp,v retrieving revision 1.5 diff -r1.5 additive_combine.hpp 62a63,68 > void seed() > { > _mlcg1.seed(); > _mlcg2.seed(); > } > Index: inversive_congruential.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/inversive_congruential.hpp,v retrieving revision 1.6 diff -r1.6 inversive_congruential.hpp 63c63 < void seed(IntType y0) { value = y0; if(b == 0) assert(y0 > 0); } --- > void seed(IntType y0 = 1) { value = y0; if(b == 0) assert(y0 > 0); } Index: discard_block.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/discard_block.hpp,v retrieving revision 1.7 diff -r1.7 discard_block.hpp 52a53 > void seed() { _rng.seed(); _n = 0; } Index: shuffle_output.hpp =================================================================== RCS file: /cvsroot/boost/boost/boost/random/shuffle_output.hpp,v retrieving revision 1.6 diff -r1.6 shuffle_output.hpp 65a66 > void seed() { _rng.seed(); init(); } -- "Sie haben neue Mails!" - Die GMX Toolbar informiert Sie beim Surfen! Jetzt aktivieren unter http://www.gmx.net/info

2 1