- Boost - lists.boost.org

is_abstract< A<int> > fails with gcc 3.4.1
by Christoph Ludwig 29 Aug '04

29 Aug '04

FYI: According to the regression tests, gcc 3.4.1 supports boost::is_abstract<T>. Unfortunately, that does not hold if T is a class template specialization, as exhibited by the following program: cludwig@lap200:~/C++/gcc3.4/tmp> cat ./is_abstract_test.cc #include <iostream> #include <iomanip> #include <boost/type_traits/is_abstract.hpp> template<typename T> class A { virtual void f() = 0; }; class B { virtual void f() = 0; }; int main() { std::cout << std::boolalpha; std::cout << "is_abstract< A<int> >::value == " << boost::is_abstract< A<int> >::value << '\n'; std::cout << "is_abstract< B >::value == " << boost::is_abstract< B >::value << '\n'; } cludwig@lap200:~/C++/gcc3.4/tmp> ./is_abstract_test is_abstract< A<int> >::value == false is_abstract< B >::value == true This causes problems, e.g., when using the serialization library with class templates. I entered the problem in the gcc bug database. It's PR#17232. Christoph -- http://www.informatik.tu-darmstadt.de/TI/Mitarbeiter/cludwig.html LiDIA: http://www.informatik.tu-darmstadt.de/TI/LiDIA/Welcome.html

1 0

problem with small width CRCs
by Bert Klaps 29 Aug '04

29 Aug '04

Hi, I ran into a problem using the CRC library for SDH LCAS control words (3-bit CRC) and SDH Trail Trace Identifier messages (7-bit CRC). For CRC widths smaller than character width the theoretical and optimized CRC computers produce different results for non-reflected input. Gcc produces this warning: /usr/local/boost/boost_1_31_0/boost/crc.hpp:581: warning: right shift count >= width of type The problem is that for the table based computer the msb of the remainder needs to be compared to the msb of the input byte. For most popular CRC parameters that's a right shift, but for CRC widths smaller than a character this should be a left shift... Reflected input requires comparing lsb's which doesn't involve any shifting. I've attached a patch that solved the problem for me. Regards, Bert Klaps /usr/local/boost/boost_1_31_0/boost/crc.hpp: In static member function `static unsigned char boost::detail::crc_helper<Bits, false>::index(typename boost::uint_t<Bits>::fast, unsigned char) [with unsigned int Bits = 3]': /usr/local/boost/boost_1_31_0/boost/crc.hpp:928: instantiated from `void boost::crc_optimal<Bits, TruncPoly, InitRem, FinalXor, ReflectIn, ReflectRem>::process_block(const void*, const void*) [with unsigned int Bits = 3, typename boost::uint_t<Bits>::fast TruncPoly = 3, typename boost::uint_t<Bits>::fast InitRem = 0, typename boost::uint_t<Bits>::fast FinalXor = 0, bool ReflectIn = false, bool ReflectRem = false]' /usr/local/boost/boost_1_31_0/boost/crc.hpp:947: instantiated from `void boost::crc_optimal<Bits, TruncPoly, InitRem, FinalXor, ReflectIn, ReflectRem>::process_bytes(const void*, unsigned int) [with unsigned int Bits = 3, typename boost::uint_t<Bits>::fast TruncPoly = 3, typename boost::uint_t<Bits>::fast InitRem = 0, typename boost::uint_t<Bits>::fast FinalXor = 0, bool ReflectIn = false, bool ReflectRem = false]' /usr/local/boost/boost_1_31_0/boost/crc.hpp:907: instantiated from `void boost::crc_optimal<Bits, TruncPoly, InitRem, FinalXor, ReflectIn, ReflectRem>::process_byte(unsigned char) [with unsigned int Bits = 3, typename boost::uint_t<Bits>::fast TruncPoly = 3, typename boost::uint_t<Bits>::fast InitRem = 0, typename boost::uint_t<Bits>::fast FinalXor = 0, bool ReflectIn = false, bool ReflectRem = false]' /usr/local/boost/boost_1_31_0/boost/crc.hpp:973: instantiated from `void boost::crc_optimal<Bits, TruncPoly, InitRem, FinalXor, ReflectIn, ReflectRem>::operator()(unsigned char) [with unsigned int Bits = 3, typename boost::uint_t<Bits>::fast TruncPoly = 3, typename boost::uint_t<Bits>::fast InitRem = 0, typename boost::uint_t<Bits>::fast FinalXor = 0, bool ReflectIn = false, bool ReflectRem = false]' /usr/lib/gcc-lib/i686-pc-linux-gnu/3.3.3/include/g++-v3/bits/stl_algo.h:157: instantiated from `_Function std::for_each(_InputIter, _InputIter, _Function) [with _InputIter = __gnu_cxx::__normal_iterator<unsigned char*, std::vector<unsigned char, std::allocator<unsigned char> > >, _Function = boost::crc_optimal<3, 3, 0, 0, false, false>]' lcas.cpp:44: instantiated from here /usr/local/boost/boost_1_31_0/boost/crc.hpp:581: warning: right shift count >= width of type *** crc.hpp.orig Tue Aug 10 14:10:34 2004 --- crc.hpp Tue Aug 10 14:57:06 2004 *************** *** 527,532 **** --- 527,556 ---- did_init = true; } + #ifndef BOOST_NO_TEMPLATE_PARTIAL_SPECIALIZATION + // Align the msb of the remainder to a byte + template < std::size_t Bits, bool RightShift > + class remainder + { + public: + typedef typename uint_t<Bits>::fast value_type; + + static unsigned char align_msb( value_type rem ) + { return rem >> (Bits - CHAR_BIT); } + }; + + // Specialization for the case that the remainder has less + // bits than a byte: align the remainder msb to the byte msb + template < std::size_t Bits > + class remainder< Bits, false > + { + public: + typedef typename uint_t<Bits>::fast value_type; + + static unsigned char align_msb( value_type rem ) + { return rem << (CHAR_BIT - Bits); } + }; + #endif // CRC helper routines template < std::size_t Bits, bool DoReflect > *************** *** 555,561 **** // Compare a byte to the remainder's highest byte static unsigned char index( value_type rem, unsigned char x ) ! { return x ^ ( rem >> (DoReflect ? 0u : Bits - CHAR_BIT) ); } // Shift out the remainder's highest byte static value_type shift( value_type rem ) --- 579,587 ---- // Compare a byte to the remainder's highest byte static unsigned char index( value_type rem, unsigned char x ) ! { return x ^ ( DoReflect ? rem : ! ((Bits>CHAR_BIT)?( rem >> (Bits - CHAR_BIT) ) : ! ( rem << (CHAR_BIT - Bits) ))); } // Shift out the remainder's highest byte static value_type shift( value_type rem ) *************** *** 578,584 **** // Compare a byte to the remainder's highest byte static unsigned char index( value_type rem, unsigned char x ) ! { return x ^ ( rem >> (Bits - CHAR_BIT) ); } // Shift out the remainder's highest byte static value_type shift( value_type rem ) --- 604,610 ---- // Compare a byte to the remainder's highest byte static unsigned char index( value_type rem, unsigned char x ) ! { return x ^ remainder<Bits,(Bits>CHAR_BIT)>::align_msb( rem ); } // Shift out the remainder's highest byte static value_type shift( value_type rem )

4 4

Triple change to Base-from-member
by Daryle Walker 28 Aug '04

28 Aug '04

I applied three changes for this library to CVS. (Three separate commits were done.) 1. Implemented a scheme to use the Pre-processor library to generate the various constructor templates. This allows the maximum number of constructors to be set at compile-time. It's a macro constant, so other headers can copy the exact number of constructors too. Changed "base_from_member.hpp" and "base_from_member.html". Shouldn't cause any problems unless your header can't work with the Pre-processor library. 2. Moved the default of the second template argument of base_from_member to the primary header. This removes the last component of "utility_fwd.hpp". Changed "base_from_member.hpp" and "base_from_member.html", and removed "utility_fwd.hpp". Shouldn't cause any problems unless your header directly #included "utility_fwd.hpp", which my check didn't pick up. 3. Forgot to update the copyright year list of "base_from_member.hpp" to add 2004. Changed that. Shouldn't cause any (additional) problems. -- Daryle Walker Mac, Internet, and Video Game Junkie darylew AT hotmail DOT com

1 0

[type_traits] "member_function_pointer_traits" ?
by Tobias Schwinger 28 Aug '04

28 Aug '04

Dear Boosters, what is the reason for the absence of "member_function_pointer_traits" (or alike) in the type_traits library ? To be precise: is it a design decision or just something not yet implemented ? In the latter case I hereby volunteer for implementing it. I suggest an interface similar to the one of function_traits but with two additional constants for 'const' and 'volatile' qualification of the pointee method, because there is no member function type without the pointer thus making it impossible to use remove_pointer / is_[qualifier] or pass 'Class::member_function' (without the '&', that is) as an argument of parametrized type to a template function. Best regards, Tobias

5 7

RE: [boost] Re: [1.32 Plan] Stalled forever?
by Rozental, Gennadiy 28 Aug '04

28 Aug '04

> Gennadiy Rozental writes: > > I would like to second that. > > > > Two month ago I was told that I had to freeze any development yet > > another month ago. Now we still nowhere and I still couldn't move a > > finger. > > 1) Release or not, any major new development should happen on > the branch. I could do my own development anywhere, even keep files on my hard drive. How am I supposed to test it? > 2) At least please take care of the unmarked failures in > Boost.Test before > claiming that you've done your part for the release. Does it say anywhere that I am required to? All the failures either expected or too minor to jump to fixes now and disturb the release. I had enough unexpected headache with things I did decide to fix > > I understand that it may be caused by book preparation. But > should we > > necessarily need to use namely this release? > > Yes. We have to use a coherent set of sources that contains > the new version of the MPL, and this is what this release is > going to provide us with, in particular. We may've published this release month ago with MPL changes. And publish another one (may be patch release) month from now with new MPL. > > > > Let at least set a policy for open information: if for > whatever reason > > release needs to be postponed lets announce it and set a new date. > > There *was* announcement of the release delay. The new date is > going to be set as soon as MPL is in the main trunk. I don't believe "release postponed until further notice" is good enough. The same time new schedule needs to be proposed and agreed upon (or after discussion we may decide to go ahead with release on previously defined terms, ignoring issues causing release to postpone). Please, understand that I am not trying to find responsible. But I also don't find this uncertainty acceptable. Regards, Gennadiy.

3 2

[boost.thread] Thread local storage on Windows
by Roland 28 Aug '04

28 Aug '04

As everyone is currently busy preparing for the 1.32 release, I do not want to clutter up the list with discussion of new things too much. However, recently I posted some suggestions for improving the windows implementation of tls. Because noone yet had time to comment on it, (because there are lot of other things to be done) I decided to also post a reworked version to http://codeproject.com/useritems/tls.asp Anyone interested in my proposal, please refer to the above article, instead of going to the boost file section. I did this in the hope to get some feedback that could help to find out whether these improvements are worth considering to be implemented to the boost library as well. Regards, Roland

1 0

preprocessor question
by Arkadiy Vertleyb 28 Aug '04

28 Aug '04

Hi all, I am trying to find a way to implement something like a preprocessing-time map. As far as I understand, the task is trivial if "keys" contain only one token: #define FIND(x) BOOST_PP_CAT(FIND_, x) #define FIND_key1 val1 #define FIND_key2 val2 #define FIND_key3 val3 And then: FIND(key1) // evaluates to val1 FIND(key2) // evaluates to val2 FIND(key3) // evaluates to val3 However, this approach obviously brakes when keys contain more than one token, like "key 1" (or, to be closer to the real problem, "unsigned int"). So I am wonderring if there is any facility in the preprocessor (or in the Boost.Preprocessor) that would allow me to do this. Thanks in advance, Arkadiy

4 4

[boost.move] move_ptr
by Thorsten Ottosen 28 Aug '04

28 Aug '04

Hi, I just wanted to hear if anybody had already made a move_ptr I could use? br Thorsten

9 24

Re: MPL vector size limit
by Arkadiy Vertleyb 28 Aug '04

28 Aug '04

Sorry to repeat this question, but I think I am missing somehing here: Do I understand correctly that it's not possible in the current version of MPL to use mpl::vector of a size larger than 50? If this is not correct, what exactly do I have to do to use, for example, mpl::vector97<...>? If this _is_ correct, I would like to make a feature request for it to be increased, since the limit of 50 doesnt seem to be high enough (I am making this statement in conjunction with the need of my typeof emulation lib, but I am kind of surprized this question hasn't yet popped up in other contexts, or has it?). In particular, isn't it possible to just get it fully controlled by BOOST_MPL_LIMIT_VECTOR_SIZE without any limitations whatsoever? Thanks, Arkadiy

2 1

Suggestion for a new class TextTableFormat
by AviadR 28 Aug '04

28 Aug '04

Hi, I am (shamefully) a newbie. I want to suggest a tiny library I've written that allows you to create "tables" of textual strings. What I mean is supposed you want to output rows of data that look like this FIELD1 (tab) FIELD2 (tab) .... FIELD-N Because many times the actual data that is written to the fields are of different lengths, You might get your output like this Field1_data field2_data field3_data Another_field1_data another_field2_data another_field3_data ... Which is not very readble or conveniet. The library I am proposing fixes this dire situation by defining a format, which states for each field: 1. its 'width' or how many columns it takes 2. its length, how many characters can it hold altogether. Data is written to a buffer using a special class which relies on the Format information. * Lines within a field are broken down in order for the column invariant to be preserved. * if data written to a field is too long for the field, the output of the field ends with "..." to indicate that the data was truncated. * the classes uses NO dynamic memory whatsoever and are very fast. If this is somehow interesting to someone I can spend the time 'boostfying' it And maybe factor some of the features stated above into Strategy classes. So how about it? I am attaching the relevant code here. IT WILL NOT COMPILE because it currently relies on My IO library, my string classes etc. of course if it is deemed interesting I will change it to use boost's classes. Thanks AviadR mailto:aviadr@pob.huji.ac.il

1 0