has_binary_operator.hpp error
Hello list. I just upgraded my system to the newest version of boost, i.e., 1.48.0. Boost compiled just fine. However, I recompile my applications (that link to boost) and I get the following error: ****************************************************************** usr/local/include/boost/type_traits/detail/has_binary_operator.hp:50: Parse error at "BOOST_JOIN" ****************************************************************** Any suggestions? Regards, Panagiotis Foteinos
Hi, Panagiotis Foteinos writes:
Hello list.
I just upgraded my system to the newest version of boost, i.e., 1.48.0. Boost compiled just fine.
However, I recompile my applications (that link to boost) and I get the following error:
****************************************************************** usr/local/include/boost/type_traits/detail/has_binary_operator.hp:50: Parse error at "BOOST_JOIN" ******************************************************************
Any suggestions?
Regards, Panagiotis Foteinos
I could reproduce this. It occurs when building CGAL 3.8. It is not a compiler error, instead it is an error of Qt generating the moc-file (the "moc" program of Qt). For some reason, boost is included in this file where moc is run. As you see the last step is [ 29%] Generating __/__/include/CGAL/Qt/moc_GraphicsItem.cxx which is not a compiler step. I suppose that this either needs a fix in CGAL, or in Qt in the moc generator. However, I don't know a quick fix for this... Regards, Lothar
I suppose that this either needs a fix in CGAL, or in Qt in the moc generator. However, I don't know a quick fix for this...
Regards, Lothar
In case it helps the OP, I just proposed a patch to CGAL: https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2011-11/msg00087.html Mark
Thank you guys; yes, the error occurs when compiling CGAL. This conversation seems more relevant inside CGAL's mailing list. Regards, Panagiotis On Thu, Nov 17, 2011 at 7:16 PM, Mark Brand <mabrand@mabrand.nl> wrote:
I suppose that this either needs a fix in CGAL, or in Qt in the moc generator. However, I don't know a quick fix for this...
Regards, Lothar
In case it helps the OP, I just proposed a patch to CGAL:
https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2011-11/msg00087.html
Mark
_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users
I just upgraded my system to the newest version of boost, i.e., 1.48.0. Boost compiled just fine.
However, I recompile my applications (that link to boost) and I get the following error:
****************************************************************** usr/local/include/boost/type_traits/detail/has_binary_operator.hp:50: Parse error at "BOOST_JOIN"
This file is new to this release. Then you probably did not include it yourself in your "old" code. So I think that a simple fix would be to replace somewhere in your code: #include <boost/type_traits.hpp> // this includes everthing from type_traits by the list of #include you really need: #include <boost/type_traits/header1.hpp> #include <boost/type_traits/header2.hpp> #include <boost/type_traits/header3.hpp> ... Regards, Frédéric
Frédéric Bron <frederic.bron <at> m4x.org> writes:
****************************************************************** usr/local/include/boost/type_traits/detail/has_binary_operator.hp:50: Parse error at "BOOST_JOIN"
This file is new to this release. Then you probably did not include it yourself in your "old" code. So I think that a simple fix would be to replace somewhere in your code:
#include <boost/type_traits.hpp> // this includes everthing from type_traits ...
Regards, Frédéric
I am having exactly the same problem, compiling boost+Qt sources with 1.48 release of boost. What might be of interest for everyone having this problem is that in my case I don't explicitly specify #include <boost/type_traits.hpp> in code. It might be used in some other boost headers that I have plenty in my code. I have to fallback to 1.47. I think that this issue will affect large population of boost users, since many are using Qt.
I am having exactly the same problem, compiling boost+Qt sources with 1.48 release of boost. What might be of interest for everyone having this problem is that in my case I don't explicitly specify #include <boost/type_traits.hpp> in code. It might be used in some other boost headers that I have plenty in my code.
It would help if somebody could give the output of the preprocessor. With g++, just add -E option and compile one failing file. This may help understand the error with BOOST_JOIN. Is there #undef BOOST_JOIN somewhere? Frédéric
Frédéric Bron wrote:
I am having exactly the same problem, compiling boost+Qt sources with 1.48 release of boost. What might be of interest for everyone having this problem is that in my case I don't explicitly specify #include<boost/type_traits.hpp> in code. It might be used in some other boost headers that I have plenty in my code. It would help if somebody could give the output of the preprocessor. With g++, just add -E option and compile one failing file. This may help understand the error with BOOST_JOIN. Is there #undef BOOST_JOIN somewhere? Frédéric
Hi Frédéric, It's not the C++ preprocessor, but Qt's moc that chokes on BOOST_JOIN. The workaround I proposed to the CGAL list at https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2011-11/msg00087.html just instructs the moc preprocessor to skip the #include that introduces BOOST_JOIN. Mark
It's not the C++ preprocessor, but Qt's moc that chokes on BOOST_JOIN. The workaround I proposed to the CGAL list at https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2011-11/msg00087.html just instructs the moc preprocessor to skip the #include that introduces BOOST_JOIN.
OK, I do not know this moc. I understand it reads the files and tries to understand the syntax which fails for some reasons here. So there is nothing to do in boost then. It is strange that it appears now because BOOST_JOIN was already used in other files in 1.47.0. Frédéric
Hi Frédéric, On Saturday, 19. November 2011 06:06:11 Frédéric Bron wrote:
It's not the C++ preprocessor, but Qt's moc that chokes on BOOST_JOIN. The workaround I proposed to the CGAL list at https://lists-sop.inria.fr/sympa/arc/cgal-discuss/2011-11/msg00087.html just instructs the moc preprocessor to skip the #include that introduces BOOST_JOIN.
OK, I do not know this moc. I understand it reads the files and tries to understand the syntax which fails for some reasons here. So there is nothing to do in boost then.
Well, we've been bitten by this one, too. The way to solve this is to file a bug report at Qts bugtracker so they can improve the moc. In the meantime client code just has to put #ifndef Q_MOC_RUN around the critical places.
It is strange that it appears now because BOOST_JOIN was already used in other files in 1.47.0.
Well, the moc has improved a lot in Qt4, but nevertheless has only a parser for a limited subset of c++ (and a limited preprocessor.) So this is no real surprise, at least for me :-) Yours, Jürgen -- Dipl.-Math. Jürgen Hunold | IVE mbH Software-Entwickler | Lützerodestraße 10 Tel: +49 511 897668 33 | 30161 Hannover, Germany Fax: +49 511 897668 29 | http://www.ivembh.de juergen.hunold@ivembh.de | | Geschäftsführer: Sitz des Unternehmens: Hannover | Univ.-Prof. Dr.-Ing. Thomas Siefer Amtsgericht Hannover, HRB 56965 | PD Dr.-Ing. Alfons Radtke
participants (6)
-
Frédéric Bron -
Jürgen Hunold -
Lothar -
Mark Brand -
Panagiotis Foteinos -
qplace