March 2007 - Boost-users - lists.boost.org

Boost Conference Reminder
by David Abrahams 19 Mar '07

19 Mar '07

Just a quick note to remind you that the official website for BoostCon'07 is up and registration is open: http://www.boostcon.com Space is limited, so register early to be sure you'll get in. See http://www.boostcon.com/registration for registration information. I'm really thrilled at the quality of the program (http://www.boostcon.com/program) we've assembled. On most days I'm going to have a hard time choosing among the simultaneous sessions! I look forward spending a week with you, immersed in Boost, at one of the most beautiful meeting sites in the world (http://www.boostcon.com/location/site) See you at BoostCon! -- Dave Abrahams Boost Consulting www.boost-consulting.com

1 0

[fusion] How to use transform Metafunction?
by Christian Henning 19 Mar '07

19 Mar '07

Hi there, since there is no example in the docs and no test for it. I was wondering how to use fusion's transform metafunction. Here is what I want to do. template <typename T> struct add_vector { typedef std::vector<T> type; }; struct A; struct B; struct C; int main() { typedef fusion::list< A, B, C > types; typedef fusion::transform< types, add_vector<> >::type vector_of_types; return 0; } The compiler is complaining that there are too few parameters for add_vector. In MPL you would just add _1 as the template parameter. But this doesn't work for fusion. So, how do I use it? Thanks, Christian

3 5

[lexical_cast] lexical_cast<int>("1e4") == 1
by n.torrey.pines＠gmail.com 19 Mar '07

19 Mar '07

lexical_cast<int>("1e4") == 1 // because of the way streams work of course I think that in the interests of robustness, and unless a better solution can be found, lexical_cast<> should be secialized for some types. Another solution might be to make sure that the whole string argument got consumed (rejecting things like "123 " )

5 4

which type of smart pointer to use?
by Vaclav 19 Mar '07

19 Mar '07

Hello, When I reset the main pointer to an object, I want all other pointers pointing to that object to be reset as well. Which type of smart pointer should I use? Example: boost::shared_ptr<MySillyClass> mySharedPtr1(new MySillyClass("test")); boost::shared_ptr<MySillyClass> mySharedPtr2 = mySharedPtr1; mySharedPtr1.reset(); if(mySharedPtr2.use_count() > 0) { TRACE("still used\n"); } else { TRACE("not used anymore\n"); } I am getting message "still used". I understand that is the normal behavior of boost::shared_ptr. Is there a smart pointer type that would print the second message? Reason: When I am done with an object, I would reset the main pointer to it and do not want other references to that object to be able to do anything with it. Thanks for advice! Vaclav

3 3

MPL map error on g++
by Santosh Joseph 18 Mar '07

18 Mar '07

My code compiles on visual studio 8, but on g++ 3.3.2 (linux) and with g++ 3.4.4 (cygwin) it chokes with the following error: In function `void convertToScctsOpt()': optionmanager.cpp:119: error: expected init-declarator before "ScctsType" optionmanager.cpp:119: error: expected `, ' or `;' before "ScctsType" optionmanager.cpp:120: error: `ScctsType' undeclared (first use this function) Is some special command line argument needed? thanks for any help. #define BOOST_MPL_CFG_NO_PREPROCESSED_HEADERS #define BOOST_MPL_LIMIT_MAP_SIZE 50 #include <boost/mpl/map.hpp> #include <boost/mpl/at.hpp> #include <boost/mpl/pair.hpp> namespace bm = boost::mpl; typedef bm::map< bm::pair<Type1, MappedType1>, . . . bm::pair<Type41, MappedType41> > TypeToTypeMap; template<typename SoapType> inline void convertToScctsOpt() { typedef bm::at<TypeToTypeMap,SoapType>::type ScctsType; } ____________________________________________________________________________________ Now that's room service! Choose from over 150,000 hotels in 45,000 destinations on Yahoo! Travel to find your fit. http://farechase.yahoo.com/promo-generic-14795097

2 1

Boost.Interprocess shared_memory.hpp
by Jean-Christophe Roux 17 Mar '07

17 Mar '07

Hello, I installed boost-1_35 this morning from cvs, which I believe contains the interprocess library. I took the interprocess doc from the vault. In the doc, there is at the end of the chapter "Shared Memory" a "Simple Example" that I am unable to compile because it contains the line: #include <boost/interprocess/shared_memory.hpp> but that file does not exist in my boost-1_35 folder. In /usr/local/include/boost-1_35/boost/interprocess, I have managed_shared_memory.hpp shared_memory_object.hpp but no shared_memory.hpp Maybe I am missing something obvious... Maybe the library from 1_35 and the docs from the vault are not compatible? Thanks for helping me out here. Jean-Christophe

2 2

[boost-users][quickbook] Trying to get many sections onto one page
by Darren Garvey 17 Mar '07

17 Mar '07

Hello, I have a situation like this: [section Section 1] ... [endsect] [section Section 2] ... [endsect] and I'd like the generated html to have the layout of an article document-type but for it all to lie on one page. Obviously cut/paste is easy enough, but I'm wondering what the correct way to do this automatically is. Regards, Darren

2 2

Fwd: SQL library with full embedded SQL-syntax
by Lyfar Dmitriy 17 Mar '07

17 Mar '07

Hello, sorry for my long responce. Rene Rivera wrote: > Sure... But we could use that same argument to propose that we write C++ > programs by writing an abstraction on top of C++ compilers in Python so > that we can remove the incompatibilities in the various C++ > implementations. Having had the displeasure of dealing with PL/SQL there > is one serious drawback to the language embedded approach, it ties ones > to the capabilities of the abstraction. May be you right, but I think that that comparison is incorrect, because of as far as I know SQL-standart doesn't include many useful features from various databases, opposite, the C++ standart, which is describes the whole language :) The main goal of the abstract layer for that library is, as Thorsten said, is to provide "portable" SQL for at least a simple queries. There is no goal to emulate features of each db for others at the first time. > Jorge Lodos wrote: > Sorry, I don't see how any of this applies -- just because the SQL is a string > doesn't mean it comes from an untrusted source. And, programmers that don't > validate input from untrusted sources deserve what they get.... I'd like to draw an intermediate conclusions: 1) The SQL library should provide a backdoor so you can still use string queries if you need to be non-portable or cutting egde etc. 2) Embedded SQL commands is type-safe, but strings are flexible and more useful in the 'real world' 3) It's hard to write one big abstract layer for all databases with single functionality and many people doesn't need in it usually. 4) ...? Phil Endecott wrote: > Well, an interface is not much use without some sort of implementation > behind it. Possible implementations for a "generic relational > interface" include an SQL database, on-disk data structures implemented > using something like stxxl, and in-memory data structures made > persistent using serialisation. A measure of a good generic relational > interface could be that it works well with all of these implementations. > > Thinking just about SQL backends, the basic question is whether we want to > > (a) hide the SQL, making the database accesses look more like C++ > function calls and data structures, or > > (b) make the SQL explicit, making the C++ code look more like SQL. > > I definitely vote for (a), and the possibility of moving an application > to something like stxxl is one of my motivations for that. If you interested in such libraries -- take a look at database template library(dtl) also. What do you think about containers overhead? Does the (a) variant means that the parsing of the result of query is something like parsing an Variant type? Is it always opportune? The people who wants to have type-safety of queries and platform independence will select the (b) variant, I hope :) ----- Best Regards, Dmitriy Lyfar

1 0

Bug in boost.filesystem?
by Tom Bachmann 17 Mar '07

17 Mar '07

-----BEGIN PGP SIGNED MESSAGE----- Hash: SHA1 Hi, Is the following code expected to be portable? namespace fs = boost::filesystem; fs::rename (current_p, skin_p / current_p.leaf ()); Where current_p and skin_p are fs::path, where current_p is valid and skin_p is the result of dereferencing a directory iterator (that was not at behind-end position at time of dereferencing). For current_p = foo/10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg skin_p = foo/bar (directory) the code causes an exception: boost::filesystem::path: invalid name "10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg" in path: "10NfRaHD1AAICPMKZIC50FvmaHw==.medium.jpg" (that is, I believe, at the point where a temporary path object is constructed from current_p.leaf ()). This appears quiet strange to me, as I never mention the file in the code, and I'd expect the library to treat such oddities. The following code works: fs::rename (current_p, skin_p / fs::path (current_p.leaf (), fs::native)); Is this the expected behavior? - -- - -ness- -----BEGIN PGP SIGNATURE----- Version: GnuPG v1.4.6 (GNU/Linux) iD8DBQFF+/vPvD/ijq9JWhsRAhMdAKCDyLsi0EKzhUnVcM3GjKopqnpcCQCbBK4y vkNablkmw69y3xj6ulSNdvs= =izVN -----END PGP SIGNATURE-----

2 2

Boost path to use for VC++ 8?
by Jeff Dunlap 17 Mar '07

17 Mar '07

Hello, I am new to VC++ (ver 8) and would really like to use Boost! This is what I have done: 1) I unzipped the boost download to c:\boost_1_33_1 2) At the command prompt, I executed: c:\boost_1_33_1\>bjam "-sTOOLS=VC-8_0" install bjam ran fine and a new directory was created: c:\boost\include\boost-1_33_1\boost c:\boost\lib I had assumed that using bjam's install parameter would modify whatever the path settings for the VC IDE are supposed to be, but that does not appear to happen. There are alot of paths and I want to make sure that I use the right one. What are the paths that I am supposed to include in VC++ when using the bjam param 'install' above? Secondly, I was thinking that maybe the 'stage' param might be a better bjam choice since the boost libraries are built in a common directory, thus making it easier to use. Are there drawbacks to using boost this way? Thank you very much

2 2