February 2004 - Boost-users

[Graph] What can I do with null_vertex?
by Douglas Gregor 20 Feb '04

20 Feb '04

There are no semantics for null_vertex. I assume I can copy it, but can I compare it against other vertex descriptors via == and !=? Is it always different from "real" vertex descriptors? Doug

2 1

Determining offset of regex syntax errors
by Keith MacDonald 20 Feb '04

20 Feb '04

Members of class reg_expression know where they are up to while compiling, but don't include that information when they throw. It can be pretty daunting to get the "Invalid regular expression" error message, with no indication of where to start looking, so I'd like to be more helpful. Given the character offset from the start of the RE, I can put the cursor there. Any suggestions as to how to obtain that information? Thanks, Keith MacDonald

2 3

Redirecting the detailed test report
by Hermann Oliveira Rodrigues 20 Feb '04

20 Feb '04

Hi, Recently we have decided to use the Boost Test libraries to provide automated tests to our project. To help with the test result visualization, we created an utility to generate HTML pages from Boost test reports (as generated by -- report_level=detailed). In the begining of implementation, the amount of tests was small and we could just cut-and-paste the test report to a file used as the input of out in-house HTML generator. Now, the amount of tests increased and cutting-and-pasting the report become a pain. So the question: How can we send the test report to a separated file? We used unit_test_log::instance().set_log_stream( stream ) in the definition of test_suite* init_unit_test_suite( int argc, char* argv[] ) to redirect the test log and it worked fine. However, unit_test_result::instance().detailed_report( stream ) can not be used in the same way. We can not simply redirect the test output to a file (using ">") because our application prints lots of information during execution and is pretty hard to change this behaviour. Thanks in advance.

2 2

Overloading New and Delete to track memory in shared_ptr.
by Joshua Little 19 Feb '04

19 Feb '04

Hi again, Well I'm making pretty good progress in learning proper way to use shared_ptr's but I'm still a little nervous about leaking memory, especially if I manage to do something I'm not supposed to do and don't recognize that a memory leak may be occuring. I've used the overloading new/delete before to help track memory leaks and I was wondering if that would be possible with shared_ptrs. I tried my old header files with and saw tons of leaks but I think its because my New that I use logs the creation to a Tracer class but since the Delete happens in the shared_ptr, its not using the overloaded Delete that removes stuff from the tracer when its deleted. I've looked around and found some examples of making a custom deleter to pass into the shared ptr when creating a new shared_ptr which should work but I have some worries about it. I'm not quite sure what the recommended way of doing this would be. About the only thing I can think of would be to have two different deleters which would be conditionally compiled something like this : //Overloaded New to log allocation to a tracer, theres a define to add the file and line number to the new call #ifndef NDEBUG void * operator new (unsigned int size, char const * file, int line) { void * p = new (size); if (Tracer::Ready) NewTrace.Add (p, file, line); return p; } struct deleter { void operator() (void * p) const { if (Tracer::Ready) NewTrace.Remove(p); delete p; } } #else struct deleter { void operator() (void * p) const { delete p; } } #endif //NewTrace is an extern'd object in the Main.cpp file So then I'd pass in the deleter to the shared_ptr and then whichever deleter was defined would be passed into the shared_ptr. I'm kinda cobbling together a few different things I've seen but I think that with this I could include the header file in the .cpp file of any class that I want to track the memory of and then just make sure I send in the deleter when I reset or make a new shared ptr. For instance in my LoggerFactory.cpp file : #include "../include/Logger.hpp" #include "../include/LoggerFactory.hpp" #include "../include/dbnew.hpp" LoggerMap LoggerFactory::loggers; boost::shared_ptr<Logger> LoggerFactory::getInstance(std::string& logger) { boost::shared_ptr<Logger>& TheLoggerRef = loggers[logger]; if(!TheLoggerRef.get()) TheLoggerRef.reset(new Logger(logger),deleter()); return TheLoggerRef; } So then the dbnew would define the deleter to use for the Logger object, if it was a debug build it would also create the overloaded new operator, define the _FILE_ and _LINE_ macros to send into the new New operator. And if its not the debug build then essentially the normal deleter is given to the shared_ptr and everything procedes as normal. Is this a good way to go about adding memory leak checking? I've looked for a new memory leak checkers like valgrind but thats only available on Linux and I'm stuck on windows.. and all the memory checkers I've been able to find only support VC++ or borland and I use gcc. So I would like a nice simple way that I can drop right in, which is what I'm going for with the overloading new and the custom deleter. Joshua.

2 1

Concurrency support based on Boost.Threads
by Slawomir Lisznianski 19 Feb '04

19 Feb '04

License: Public Domain URL: http://www.rhapsodia.org Depends on: Boost, STL Code maturity: Beta Target platform: Any with C++ compiler. C++ compilers verified to work on: GNU g++ 3.2.x, Visual C++ 6 Service Pack 5 Description: Rhapsodia is a collection of header files implementing commonly used classes in concurrent C++ programming. Library mainly consists of implementations of: - Channel (bounded, unbounded and priority queue), - Barrier (multi-party synchronization) - Executor (replacement for direct use of thread creation and management API). Your critiques, suggestions or usage questions are much welcomed. Slawomir Lisznianski; [ www.rhapsodia.org ]

1 0

RE: [Boost-users] Mathematical expressions with containers
by Hurd, Matthew 19 Feb '04

19 Feb '04

> Russell Hind > Subject: [Boost-users] Mathematical expressions with containers > > Currently I'm doing more and more mathematical calculations on > containers, and wondered if there is any library out there that allows > containers to be placed in expressions without having to use > transform/for_each etc. > > e.g. > > std::vector<std::complex<double> > V1; > std::vector<std::complex<double> > V2; > double C1, C2; > > ... > > std::vector<std::complex<double> > Result; > > Result = C1 + V2 / (V1 * C2); > > while would be equivalent to (assuming Result has been resized) > > (I'm currently using a mixture of transforms and spirit::phoenix > (because lambda doesn't work with my compiler)) > > transform(V1.begin(), V1.end(), V2.begin(), > Result.begin(), > C1 + phoenix::arg2 / (phoenix::arg1 * C2)); > > Or something to this effect that will aid with this. > > Cheers > > Russell Seems to my quick glance you are asking for some vector operations. Have you looked at boost's ublas? Regards, Matt Hurd. IMPORTANT: The information contained in this email and/or its attachments is confidential. If you are not the intended recipient, please notify the sender immediately by reply and immediately delete this message and all its attachments. Any review, use, reproduction, disclosure or dissemination of this message or any attachment by an unintended recipient is strictly prohibited. Neither this message nor any attachment is intended as or should be construed as an offer, solicitation or recommendation to buy or sell any security or other financial instrument. Neither the sender, his or her employer nor any of their respective affiliates makes any warranties as to the completeness or accuracy of any of the information contained herein or that this message or any of its attachments is free of viruses.

3 2

Any one used boost with comeau c++ compiler for visual c++.net 2003?
by Ben Polen 18 Feb '04

18 Feb '04

To the community; I'm thinking of geting comeau computing's c++ frontend for visual c++.net 2003. I was wondering if anyone has experience using those two programs in conjunction with the boost libraries. In viewing the regression test, I was also wondering why Comeau doesn't seem to compile at a rate of 100% with boost. Is that due to Comeau unconformance, boost unconformance or problems with the c++ backend program, vc 7.1, gcc 3.2 ect. Anyone have any clues? Thanks Ben _________________________________________________________________ Stay informed on Election 2004 and the race to Super Tuesday. http://special.msn.com/msn/election2004.armx

1 0

RE: [Boost-users] Trouble linking regex library on Solaris machine
by Ben Hutchings 18 Feb '04

18 Feb '04

Justin Adam Miller <millerj(a)wam.umd.edu> wrote: > Howdy. As I am only interested in using the regex library right now, I > followed the instructions for building that library only, using dmake. > Compilation went fine. When I tried using the library in a little test > program with the command: > > g++ regextest.cxx -I./boost_1_31_0 > -L./boost_1_31_0/libs/regex/build/sunpro -lboost_regex -o > regextest.cxx > > I get many undefined symbol errors. If requested, I can provide the > entire error output. It looks like you are trying to use the wrong library. SunPro refers to the Sun C++ compiler and tools whereas you are using GNU C++. You have to build the library and application with the same compiler.

2 1

Trouble linking regex library on Solaris machine
by Justin Adam Miller 18 Feb '04

18 Feb '04

Howdy. As I am only interested in using the regex library right now, I followed the instructions for building that library only, using dmake. Compilation went fine. When I tried using the library in a little test program with the command: g++ regextest.cxx -I./boost_1_31_0 -L./boost_1_31_0/libs/regex/build/sunpro -lboost_regex -o regextest.cxx I get many undefined symbol errors. If requested, I can provide the entire error output. Justin Miller

2 2

RE: [Boost-users] Re: Semaphore and barriers for boost-threads?
by Ben Hutchings 18 Feb '04

18 Feb '04

Craig Rodrigues <rodrigc(a)crodrigues.org> wrote: > It is possible to implement semaphore's with mutexes and > condition variables. I looked at the source code for > FreeBSD 5.2's implementation of semaphores to see how this was > done: > > http://www.freebsd.org/cgi/cvsweb.cgi/~checkout~/src/lib/libpthread/thre ad/thr_sem.c > > I even took a whack at writing one based on Boost (it may be buggy): > > #include <boost/thread.hpp> > #include <boost/thread/mutex.hpp> > > struct mysem { > mysem(unsigned int count): count_(count) {} > > void post() { > boost::mutex::scoped_lock l(sem_mutex_); > ++count_; > sem_cond_.notify_all(); notify_one() should be safe and more efficient here, because any of the waiters can be satisfied by the change you have made. Use notify_all() where waiters are waiting for different conditions on the same condition variable. (I'm not sure I expressed that exactly correctly - can anyone improve on that?) > } > > void wait() { > boost::mutex::scoped_lock l(sem_mutex_); > for(;;) { > > if( count_ == 0 ) sem_cond_.wait(l); This is wrong - you must always use while rather than if to test the inverse of the condition you are waiting for because it is possible that the condition is no longer true when wait() returns. (There is also a function template for wait() that tests a predicate for you.) It is even possible on some systems for wait() to return even though no-one called notify_one() or notify_all(). > else { --count_; break; } > } > } > > private: > unsigned int count_; > boost::mutex sem_mutex_; > boost::condition sem_cond_; > }; > > > What kinds of things can go wrong if I use this class? <snip> The other potential problem I see is overflow of count_, but it is exceedingly unlikely that you will be able to create enough threads to cause that.

1 0