- Boost - lists.boost.org

[move] explicit moving from a rvalue
by Vicente J. Botet Escriba 27 Mar '12

27 Mar '12

Hi Ion, With the current emulation of move semantics of Boost.Thread using Boost.move, I have a case that I would like to make working in a portable way boost::thread t = boost::thread( MoveOnly()); The library declares the template move constructor from a movable callable as as follows template <class F> explicit thread(boost::rv<F>& f); and MoveOnly is declared as defined below. The problem is a know limitation as the compiler is unable to deduce F as been MoveOnly even if MoveOnly is convertible to boost::rv<MoveOnly>& Jeffrey Lee Hellrung, Jr. made a proposal to try to solve the issue, but his proposal has a performance penality. Here it is a less elegant but efficient enough alternative. The addition of a explicit move conversion as ::boost::rv<MoveOnly>& move() { return *static_cast< ::boost::rv<MoveOnly>* >(this); } allows to the following to work as expected boost::thread t = boost::thread( MoveOnly().move()); While this workaround works when we are using the emulation it doesn't works when rvalue references are supported. I have defined a macro BOOST_EXPLICIT_MOVE that allows to write portable code #if ! defined BOOST_NO_RVALUE_REFERENCES #define BOOST_MOVE(RVALUE) RVALUE #else #define BOOST_MOVE(RVALUE) RVALUE.move() #endif that can be used as boost::thread t = boost::thread( BOOST_MOVE(MoveOnly())); The move() function is not really needed as the macro can hide the needed conversion, but could be preferred by users that don't need to make portable programs. The evident liability is of course the use of the macro. Is there something wrong on this design, something to to improve? The same design can be used on compilers, such as as SunStudio (see https://svn.boost.org/trac/boost/ticket/6222) that prefer the overload of a private copy constructor to an public move constructor. On these kind of compilers the following overload private: MoveOnly(MoveOnly&); public MoveOnly(boost::rv<MoveOnly>&) makes the following to fail at compile time MoveOnly MakeMoveOnly() { return MoveOnly(); } but MoveOnly( MakeMoveOnly(() { return BOOST_MOVE(MoveOnly()); } will work. Do you think that member functions as the explicit move() and a macro as BOOST_MOVE can be added to Boost.Move emulation? Best, Vicente ///////////////////////////// class MoveOnly { #ifndef BOOST_NO_DELETED_FUNCTIONS public: MoveOnly(const MoveOnly&)=delete; MoveOnly& operator=(MoveOnly const&); #else private: MoveOnly(MoveOnly&); MoveOnly& operator=(MoveOnly&); public: #endif MoveOnly() { } #ifndef BOOST_NO_RVALUE_REFERENCES MoveOnly(MoveOnly&&) {} #else #if defined BOOST_THREAD_USES_MOVE MoveOnly(boost::rv<MoveOnly>&) {} MoveOnly& operator=(boost::rv<MoveOnly>&) { return *this; } operator ::boost::rv<MoveOnly>&() { return *static_cast< ::boost::rv<MoveOnly>* >(this); } operator const ::boost::rv<MoveOnly>&() const { return *static_cast<const ::boost::rv<MoveOnly>* >(this); } ::boost::rv<MoveOnly>& move() { return *static_cast< ::boost::rv<MoveOnly>* >(this); } const ::boost::rv<MoveOnly>& move() const { return *static_cast<const ::boost::rv<MoveOnly>* >(this); } #else #error #endif #endif void operator()() { } };

3 5

[quickbook] Controling the layout of table columns in docbook/fo output
by Andreas Pokorny 27 Mar '12

27 Mar '12

Hi, In an earlier thread I mentioned that I am using quickbook to write software design specification/documentation. The whole document grows nicely now. Since more people are contributing, I added howto chapters to the end of the document, including some kind of guided quickbook skeleton for new chapters and the quickbook documentation itself. The quickbook documentation creates warnings with fop: Mar 12, 2012 8:11:46 PM org.apache.fop.events.LoggingEventListener processEvent WARNING: table-layout="fixed" and column-width unspecified => falling back to proportional-column-width(1) (See position 1686:869) Mar 12, 2012 8:11:46 PM org.apache.fop.events.LoggingEventListener processEvent WARNING: table-layout="fixed" and column-width unspecified => falling back to proportional-column-width(1) (See position 1686:905) Mar 12, 2012 8:11:46 PM org.apache.fop.events.LoggingEventListener processEvent WARNING: table-layout="fixed" and column-width unspecified => falling back to proportional-column-width(1) (See position 1896:796) and more. That happens for every table. Inside the documentation about tables there one containing hello world source. There the proportional column width makes the source code leave the table and text area of the page. Is there a yet undocumented way to control column sizes? regards Andreas

2 4

Re: [boost] [utility] Adding noexcept to boost::declval
by Michel Morin 27 Mar '12

27 Mar '12

Vicente J. Botet Escriba wrote >> Summary: >> Defining declval<T>() as add_lvalue_reference<T>::type is wrong; >> add_lvalue_reference should be removed! >> >> > There is something that disturb me by returning just T, doesn't this forces > the type T to CopyConstructible? Nope. T needs not to be CopyConstructible (Note that declval can be used only in unevaluated contexts.) > The idea to use T& instead was to avoid > this. You shouldn't do this, because this transforms rvalues to lvalues. > I have rolled back at revision 77562. Thanks, Vicente! Minor issue: `#include <boost/type_traits/add_lvalue_reference.hpp>` is not removed yet. Regards, Michel

2 1

[filesystem] Heads up; filesystem in trunk unstable while V2 removed
by Beman Dawes 26 Mar '12

26 Mar '12

To ensure that history is preserved, the removal of version 2 will be done as three commits. Filesystem in trunk will be unstable during until all three commits are completed. Hopefully this will take less than an hour. I'll post a notice when complete. --Beman

3 7

[gsoc]
by Andrew Sutton 26 Mar '12

26 Mar '12

It looks like there will be no summer of Code for Boost this year. For whatever reason, Google has decided not to accept Boost as a mentoring organization. Andrew

10 18

Re: [boost] [utility] Adding noexcept to boost::declval
by Michel Morin 26 Mar '12

26 Mar '12

(Forgot to send to the list; now back to the list again.) Vicente J. Botet Escriba wrote: > OK, I will rollback it and analyze the breaking change. The important thing is that declval<T>() should preserve the constness/non-constness and the rvalueness/lvalueness of T. Essentially, it's OK to define declval<T>() as simply T. Then, why declval<T>() is defined as add_rvalue_reference<T>::type? This is because * When declval<T>() is defined as T, we cannot use abstract types as T. * When declval<T>() is defined as T&&, we cannot use void as T. The definition with add_rvalue_reference<T>::type also allows to use incomplete types as T. In C++03, it is not possible to construct rvalues of abstract types or incomplete types. So, the correct way to define declval<T>() in C++03 is to just use T. And this is the definition of declval<T>() in the previous version of boost::declval when BOOST_NO_RVALUE_REFERENCES is defined. Summary: Defining declval<T>() as add_lvalue_reference<T>::type is wrong; add_lvalue_reference should be removed! Regards, Michel

1 0

Re: [boost] [gsoc]
by Andrzej Krzemienski 26 Mar '12

26 Mar '12

> From an outside point of view, i do agree. I don't see a lot of Boost GSOC > projects completed. And that's really disapointing. > In my fileds of interests, i would love to see some libraries integrated > that deals with : > - large integer / large fixed point computation > - json parser/ writer > - xml parser/ writer like pugiwml > - crossplatform async file io for boost asio (not only for windows) > - xmpp/sip stack Boost does provide parsers and writers for JSON and XML files. See http://www.boost.org/doc/libs/1_49_0/doc/html/property_tree.html. Regards, &rzej

2 1

how to mark a test "fail*"
by lcaminiti 25 Mar '12

25 Mar '12

Hello all, Expected regression test failures are marked with "fail*" and give an explanation for why the failure is expected: http://www.boost.org/development/tests/trunk/developer/output/Sandia-intel-… How do I add an explanation for a new expected failure? Is there some documentation on this process? For example, emulation_this and emulation_this_tpl are expected to fail on MSVC8 and the user should use native type-of for those-- how do I add a link with such explanation the regression test table? http://www.boost.org/development/tests/trunk/developer/output/VC8%20jc-bell… http://www.boost.org/development/tests/trunk/developer/output/VC8%20jc-bell… Thanks a lot! --Lorenzo -- View this message in context: http://boost.2283326.n4.nabble.com/boost-how-to-mark-a-test-fail-tp4503985p… Sent from the Boost - Dev mailing list archive at Nabble.com.

2 1

[mpi] all_reduce and MPI_IN_PLACE like feature
by Alain O Miniussi 25 Mar '12

25 Mar '12

Hi, MPI_Allreduce offers the possibility to use the same buffer for the send and recv data: http://www.open-mpi.org/doc/v1.4/man3/MPI_Allreduce.3.php#toc8 It seems that boost's all_reduce does not allow such a feature. First, I would like to know if the previous statement is correct and, if yes, if it was just overlooked or if there was a reason for that? Just in case, I filed a ticket regarding that issue, along with a tentative patch, in https://svn.boost.org/trac/boost/ticket/6704 Regards, Alain

1 0

Using C++11 lambdas to replace Phoenix in MPL expressions
by Augustus Saunders 25 Mar '12

25 Mar '12

I use Phoenix to essentially turn code into types which can then be manipulated through MPL. I mostly use this to conditionally select algorithms or to include/exclude code at compile time, based on some logic written with MPL. I was interested in writing the code fragments using the new C++11 lambda expressions as it gives a more natural way to write code. While you can get the type of a lambda expression with decltype, the problem I've run into is instantiating the resultant type in order to actually get the function to execute. Lambdas have no default constructor, which makes sense, as there would be no way to deal with the capture variables. So I'm wondering if anybody can shed some light on manually instantiating lambdas. I would think there is a way (possibly shady) to do this, at least for lambdas with no capture. Then passing around lambda types as template parameters would be useful, which I think would lead to many interesting results. I only get the digest, so it I would appreciate it if you CC me on any response. Thanks- Augustus

3 6