[prev in list] [next in list] [prev in thread] [next in thread] 

List:       boost-users
Subject:    [Boost-users] [format][patch][1.33.1][SunPro11] missing static_cast
From:       Marc Mutz <marc () klaralvdalens-datakonsult ! se>
Date:       2006-04-03 12:00:26
Message-ID: 200604031400.35972.marc () klaralvdalens-datakonsult ! se
[Download RAW message or body]


Hi,

SunPro (tested: v11, with current patches), needs the attached patch to compile
the following code:

--code--
using boost::format;
using boost::str;

void foo( const std::string & );

format fmt( "host=%1%,port=%2%" );
const std::string host = ...;
const unsigned short port = ...;
foo( str( fmt % host % port ) );
--/code--

--- /opt/boost/cc/1.33.1/include/boost-1_33_1/boost/format/alt_sstream_impl.hpp~      \
                Sun Feb 27 03:06:57 2005
+++ /opt/boost/cc/1.33.1/include/boost-1_33_1/boost/format/alt_sstream_impl.hpp Mon \
Apr 24 18:07:09 2006 @@ -245,7 +245,7 @@
                     add_size /= 2;
                 if(0 < add_size) {
                     new_size += add_size;
-                    newptr = alloc_.allocate(new_size, is_allocated_? oldptr : 0);
+                    newptr = static_cast<Ch*>(alloc_.allocate(new_size, \
is_allocated_? oldptr : 0));  }

                 if(0 < prev_size)


Thanks,
Marc

PS: Regarding the API: I don't much like the fact that format::operator%
changes the format instance. I would have liked to write "static const format"
in the code snippet above. Feels more natural ("do as the ints do"). I see where
the design comes from (speed optimization), but speed shouldn't dictate design.
Expression templates could be used to speed up the process.

-- 
Marc Mutz -- marc@klaralvdalens-datakonsult.se, mutz@kde.org
Klarälvdalens Datakonsult AB, Platform-independent software solutions


[Attachment #3 (application/pgp-signature)]

[prev in list] [next in list] [prev in thread] [next in thread]