- Boost-users - lists.boost.org

preprocessor: convert to capitals
by Toon Knapen 10 Apr '03

10 Apr '03

In the ublas forum we're having a discussion on the best way to provide signatures in C for fortran libraries (like BLAS and LAPACK). For linking C with fortran, you sometimes need to add an underscore to the function in C, but on other platforms fortran will store the function-name in capitals. Thus on different platforms, we should preprocess our source to take the platforms' convention into account and we thus would need a macro to convert an identifier automatically to uppercase. Is that possible ? toon -- Toon Knapen http://www.fft.be Free Field Technologies tel:+32 10 45 12 26 16 Place de l'Universite fax:+32 10 45 46 26 1348 Louvain-la-Neuve http://www.femtown.org Belgium

3 5

thread lib using workshop CC on solaris
by jenny_qzhang 10 Apr '03

10 Apr '03

Hi, All I am working on porting an application from MSVC to solaris workshop(the applicatoin is using boost thread lib). I could not locate the makefile or jam file for this compiler. Is there anybody did the same thing before? If you finially made it, can you share your experience, makefile with me? Actually I tried to write a makefile myself, but it reported a lot of errors related to template parameters :(. Thanks very much! Qiu

2 2

ublas matrix slice bug?
by Stephan Puchegger 09 Apr '03

09 Apr '03

Dear ublas developers! (I am using boost 1.30 & gcc 3.2) the following program: #include <boost/numeric/ublas/matrix.hpp> #include <boost/numeric/ublas/io.hpp> int main () { using namespace boost::numeric::ublas; matrix<double> m (4, 4); for (unsigned int i = 0; i < m.size1 (); ++ i) for (unsigned int j = 0; j < m.size2 (); ++ j) m(i, j) = 1; std::cout << m << std::endl; // multiply a slice of the matrix with 2.... project (m, slice (2, 1, m.size1()-2), slice (1, 0, m.size1()-2)) *= 2; // interesting result ... the slice has been multiplied with 4 std::cout << m << std::endl; } gives the following output: [4,4]((1,1,1,1),(1,1,1,1),(1,1,1,1),(1,1,1,1)) [4,4]((1,1,1,1),(1,1,1,1),(1,4,1,1),(1,4,1,1)) Why is the slice multiplied with 4 and not with 2?? Do I make some fundamental mistake? Everything works fine if I am using a matrix_vector_slice. My second (stupid?) question is the following: Is it intentional, that it is not possible to compile: #include <boost/numeric/ublas/matrix.hpp> #include <boost/numeric/ublas/io.hpp> int main () { using namespace boost::numeric::ublas; matrix<double> m (4, 4); for (unsigned int i = 0; i < m.size1 (); ++ i) for (unsigned int j = 0; j < m.size2 (); ++ j) m(i, j) = 4*i+j; std::cout << m << std::endl; swap(row(m,0), row(m,2)); std::cout << m << std::endl; } whereas matrix_row<matrix<double> > l(m, 0); matrix_row<matrix<double> > r(m, 2); swap(l, r); /* swap the rows */ works fine. MTL has support for a similar syntax: swap(rows(m)[0], rows(m)[2]); best regards and thanks for your help! Stephan

3 7

Spirit PDF or something
by Edward Diener 08 Apr '03

08 Apr '03

I would like to print out the Spirit documentation as a whole for offline study. However Boost 1.30 Spirit documentation comes as a series of HTML pages. Is there an easy way to print out the documentation rather than cycling through all of it manually and printing out a page at a time ?

2 3

condition::wait() - Alertable in Win32
by Simon Bailey 08 Apr '03

08 Apr '03

What is the latest thinking on making condition::wait() and condition::timed_wait() "alertable" in the Win32 environment? Please consider this a "me too" request for such a feature! I'm trying to implement a cross-platform (Unix, Windows) implementation of a Proactor pattern that manages asynchronous sockets. It seems that the way to implement this pattern with as much code common to both platforms as possible is to use 1. Overlapped sockets that call back (not post messages) in Windows. 2. Signals in Unix (similar to the Windows call backs but with less- clear semantics in my mind). For this scheme to work, the wait functions need to be alertable in Windows, but the Unix system shouldn't need any changes. >From the research that I have done so far, in my particular case, I would only need wait and timed_wait to be alertable, and it would be appropriate that they were always altertable - I would not need a way to enable or disable this feature.

1 0

Building (runtime-link-static) thread lib on vc7
by Taras Tielkes 07 Apr '03

07 Apr '03

Hi, Some newbie boost thread lib questions: 1) I used to build the 1.29 thread library with the runtime-link-static\threading-multi build configuration. This would yield (in release mode) a 467 KB import library in the destination directory. When I build boost 1.30 (with runtime-link-static\threading-multi), the target directory for the thread library contains: -a boost_thread.dll of 116 KB -boost_thread.lib import libarary of 27KB Programs compile against the import library have a runtime dependency on the boost thread dll (as I would expect from the size of the import .lib) What's the right way to compile a static-link, multi-threaded boost.threads library (using the vc7 toolset) in 1.30? 2) What happended to the threadmon.dll? Is it gone in 1.30? Thanks for any pointers..

1 0

RE: [Boost-Users] Re: Please, please help!!!
by scleary＠jerviswebb.com 07 Apr '03

07 Apr '03

> From: agiatojon [mailto:JonAgiato@nyc.rr.com] > > Thanks a lot for the info. I saw that but did not know exceptions > were disabled in some compilers by default, which as you suggested, I > found out this one does. I enabled them, and now it compiles. > > I have another question for you if I may. I attempted to simply > stream the numbers to the standard output but it complains that there > is no operator defined for this object. Is there a file I can include > that contains the overloaded output operator functionality? I am sure > this must be available, I am just not sure of its location. I know > that ublas uses the file "IO" in order to bring those definitions in. Make sure you're doing: std::cout << dev(); and not: std::cout << dev; If that isn't the problem, could you post a code sample? -Steve

2 1

specializing new lexical_cast
by Tom Matelich 07 Apr '03

07 Apr '03

I'm going to embark on it, but I thought I'd ask in parallel. One of the things about Terje's proposed lexical_cast was a lot of documentation of how to specialize it. That doesn't seem to be addressed in 1.30.0's version. I see in the documentation "Where non-stream-based conversions are required, lexical_cast is the wrong tool for the job and is not special-cased for such scenarios." I understand this sentiment, and I tried to search for discussions of this on the mailing list, but I did not conquer ASPN's search tool. The reason I'm asking is that we did some profiling and strtod kicked lexical_cast's butt for string -> double. I currently use lexical_cast in some template'd xml code, so I could overload my function, but I'd have to do it in about 5 spots due to an evolving interface. I'd like to just do it once for lexical_cast. I suppose I could use Terje's implementation, but I've stuck with boost canon for a long time now intentionally. Any suggestions? ----------------------------------------------------------------------- DISCLAIMER: Information contained in this message and/or attachment(s) may contain confidential information of Zetec, Inc. If you have received this transmission in error, please notify the sender by return email. -----------------------------------------------------------------------

4 4

New BGL warning in 1.30.0
by Steven Solie 07 Apr '03

07 Apr '03

I'm now getting the following warning from BGL 1.30.0: /gg/boost/boost/pending/property.hpp:71: warning: sorry: semantics of inline function static data `const bool match' are wrong (you'll wind up with multiple copies) /gg/boost/boost/pending/property.hpp:71: warning: you can work around this by removing the initializer The code that generates the above warning looks something like this: Territory* t1 = new Territory(...); Vertex vertex = add_vertex(m_graph); put(vertex_name, m_graph, vertex, t1); // generates warning The graph declaration looks like this: typedef adjacency_list<vecS, vecS, undirectedS, property<vertex_name_t, Territory*, property<vertex_distance_t, unsigned long> > > Graph; I don't have any problems with BGL 1.29.0 and I don't understand what the warning is trying to tell me. Any ideas? I'm using GCC 2.95.3 with STLport. --Steven

2 1

Spirit - rule generating functions
by brockpeabody 07 Apr '03

07 Apr '03

Does anyone know the best way to return a rule from a function by value? I know that the following does not work, but what's the best workaround? struct calculator : grammar<calculator> { template <typename ScannerT> struct definition { definition(const calculator&) { r = make_complex_rule('A') | make_complex_rule('B'); } typedef rule<ScannerT> rule_type; rule_type const& start() { return r;} private: rule_type make_complex_rule(char c) { return c >> real_p; //boom } rule_type r; }; }; my current workaround is the following nastiness: //inside struct calculator::definition<> rule_type& anonymous_rule() { anonymous_rules.push_back( boost::shared_ptr<rule_type>(new rule_type)); return *anonymous_rules.back(); } rule_type& make_complex_rule(char c) { return anonymous_rule() = c >> real_p; } std::vector<boost::shared_ptr<rule_type> > anonymous_rules; Is there a better way? Also, I've got an easy one... Why is a grammar passed by const reference to its definition? Does Spirit make copies of the grammars internally? I ask because otherwise the grammar class seems like the ideal place to store state information (like numbers on a calculator's stack). Thanks ahead of time! Brock Peabody brock.peabody(a)npcinternational.com

3 2