- Boost-users - lists.boost.org

Re: [Boost-users] [1.33][ptr_container] bug in erase() in associative containers?
by JOAQUIN LOPEZ MU?Z 24 Aug '05

24 Aug '05

----- Mensaje original ----- De: Thorsten Ottosen <nesotto(a)cs.aau.dk> Fecha: Miércoles, Agosto 24, 2005 6:05 pm Asunto: Re: [Boost-users] [1.33][ptr_container] bug inerase() inassociativecontainers? > > "JOAQUIN LOPEZ MU?Z" <joaquin(a)tid.es> wrote in message > news:1553a314f2df.14f2df1553a3@tid.es... > > > size_type erase( const key_type& x ) // nothrow > > > { > > > BOOST_ASSERT( !this->empty() ); > > Why the BOOST_ASSERT? Do you require that the container > be non-empty for erase to be callable? Am I missing something? > > >>>>>> > > Hm...I couldn't find anything on this in the standard. Can you > locate the > description for map<K,V>::erase( const K&) ? > In 23.1.2 [lib.associative.reqmts], the table has an entry for the expression "a.erase(k)" saying: "erases all the elements in the container with key equivalent to k. returns the number of erased elements." No precondition is stated that "a" should be nonempty, neither in the table itself nor in the general requisites for the table (23.1.2/7), hence "a" is allowed to be empty. I've taken a closer look at your code and think there are some more of these overzealous asserts: * The following in associative_ptr_container.hpp iterator erase( iterator first, iterator last ) // nothrow { BOOST_ASSERT( !this->empty() ); requires again that the container be nonempty. There's no such precondition in the std for the expression "a.erase(q1,q2)": the only requisite in 23.1.2/7 is that [q1, q2) be a valid range, which is compatible with "a" being empty if q1==q2==a.end() (see 24.1/7, see also DR151 which addresses an issue with the original erroneous requirement that q1 be dereferenceable). * Same problem as above in reversible_ptr_container.hpp. HTH, Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

Re: [Boost-users] [bjam] [v1] How to copy the generated binaries from a dependent project?
by Ames Andreas 24 Aug '05

24 Aug '05

Hi, many thanks for your help. Rene Rivera wrote: > Hm, for 1.33 I see a "stage" target in the signals/build/Jamfile. You are right. What doesn't fit my use case is that, at least I believe so, I can't specify the destination directory there. It always copies the binaries to toplevel/stage/lib. My statement was misleading here, sorry. > Similarly for BBv1 it's: > > stage my_stage > : <dll>@boost/libs/signals/build/boost_signals > <lib>@boost/libs/signals/build/boost_signals > ; > > It requires that you add definition of "boost" to your Jamrules file: > > project boost : _path_to_boost_ ; Aah, that's exactly what I needed to get started. With a bit of additional boilerplate it does what I want. cheers, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | andreas . ames AT comergo . com

1 0

[Test] Asserting a failure
by Mike 24 Aug '05

24 Aug '05

Hi. Is there a way inside a test case using the Unit Testing Framework to specify that a particular function call ought to fail? For example, I have a function that asserts when passed invalid input. In my test case I want to pass it an invalid parameter and check that the assert is triggered. I'm looking for a macro called BOOST_CHECK_FAILURE (or something like that), but cannot find one. All I see are macros that assert successes, and those that assert exceptions. But my particular function does not throw, it simply "assert()"s. Any ideas? Thanks, Mike

2 1

Re: [Boost-users] [bjam] [v1] How to copy the generated binaries from a dependent project?
by Ames Andreas 24 Aug '05

24 Aug '05

Hi, Vladimir Prus wrote: > I know your message talks about V1, but with V2 this is a simple as: > > install install-boost-libs-you-want > : <path-to-boost>/libs/signals/build//boost_signals > : <location>some-directory > ; > > for multiple variants: > > install install-boost-libs-you-want > : <path-to-boost>/libs/signals/build//boost_signals/<variant>release > <path-to-boost>/libs/signals/build//boost_signals/<variant>debug > : <location>some-directory > ; Thanks for the hint. I have tried it out with the Boost.Buildv2 that ships with 1.33.0 but unfortunately there were some problems: 1) examples/echo doesn't build at all 2) boost toplevel doesn't build because of an issue in iostreams 3) boost/libs/signals builds only a single subvariant and I couldn't figure out how to build all the others 4) contrary to v1 when I build in a subproject (like signals) v2 seems to always complain about missing python, ICU etc. Unless these are grave failures on my part (I've followed the documentation on boost.org which seems to be a bit outdated) I will presumably stay with v1 and more so because Rene Rivera provided me the clue I needed to get going. cheers, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | andreas . ames AT comergo . com

1 0

[1.33][ptr_container] bug in erase() in associative containers?
by Malte Clasen 24 Aug '05

24 Aug '05

Hi, the erase() method (defined in associative_ptr_container.hpp, line 104) seems a bit strange: size_type erase( const key_type& x ) // nothrow { BOOST_ASSERT( !this->empty() ); iterator i = find( x ); // nothrow if( i == this->end() ) // nothrow return 0; // nothrow this->remove( i ); // nothrow return this->c_private().erase( i.base() ); // nothrow } http://cvs.sourceforge.net/viewcvs.py/boost/boost/boost/ptr_container/detai… My compiler (vc++7.1) complained about the unknown find() when I called erase() on ptr_map. "find" is not declared neither in this class nor in its base class. The closest valid call I could think of would be this->c_private().find( x ) . But there seems to be a nicer solution: size_type erase( const key_type& x ) // nothrow { BOOST_ASSERT( !this->empty() ); return this->c_private().erase( x ); // nothrow } This compiles fine and does just what I want it to do, even nothrow is guaranteed (I checked the vc++ reference for std::map). Now I wonder why erase() is implemented differently in Boost 1.33. Is there anything I overlooked? Btw, is there a rationale for the declaration of insert()? std::pair<iterator,bool> insert( key_type& k, value_type x ); http://cvs.sourceforge.net/viewcvs.py/boost/boost/boost/ptr_container/ptr_m… Passing key_type by non-const reference caused some unaesthetic constructs in my code because return values (temporaries) cannot be used directly. Malte

2 2

Re: [Boost-users] [1.33][ptr_container] bug in erase() inassociativecontainers?
by JOAQUIN LOPEZ MU?Z 24 Aug '05

24 Aug '05

----- Mensaje original ----- De: Thorsten Ottosen <nesotto(a)cs.aau.dk> Fecha: Martes, Agosto 23, 2005 11:22 pm Asunto: Re: [Boost-users] [1.33][ptr_container] bug in erase() inassociativecontainers? [...] > > > > size_type erase( const key_type& x ) // nothrow > > { > > BOOST_ASSERT( !this->empty() ); > > iterator i = find( x ); // nothrow > > if( i == this->end() ) // nothrow > > return 0; // nothrow > > this->remove( i ); // nothrow > > return this->c_private().erase( i.base() ); // nothrow > > } Why the BOOST_ASSERT? Do you require that the container be non-empty for erase to be callable? Am I missing something? Joaquín M López Muñoz Telefónica, Investigación y Desarrollo

2 1

program_options range checking?
by Fredrik Hedman 24 Aug '05

24 Aug '05

Hello, I would like to be able to specify different sanity checks on the input data using the program_options framework. This should be a common enough requirement, but I can not find any example that exactly fits the bill. To be more specific I have a solution that kind of works, but I think it needs a bit to much typing. Here is what I have now: /*! Check for values greater than min. */ template <typename T> class value_gt { public: value_gt(std::string const & msg, T min) : msg_(msg), min_(min) {} void operator()(T const & v) const { if (v > min_) return; ostringstream os; os << msg_ << " to small. Must be larger than " << min_ << " . Found " << v; throw pgmopt::invalid_option_value(os.str()); } private: std::string msg_; T min_; }; void doSetProgram(Options & options) { options.add_options() ("help,h", "show this message") ("density", pgmopt::value<double>()->notifier( value_gt<double>("density", 1.0)), "initial density") ("particles", pgmopt::value<unsigned>()->notifier( value_gt<unsigned>("particles", 100)), "number of particles") ; } I don't like repeating the type (see code above) and would rather like to have access to the name of the option in the function that does the check. What would be a better solution? Adding to the validators idea or? -- F.

2 2

[bjam] [v1] How to copy the generated binaries from a dependent project?
by Ames Andreas 24 Aug '05

24 Aug '05

Hi, I'm using some of boost's libs in my own projects. Therefore I like to build certain (sub)variants of these libraries from the according library subdirectories to avoid the (quite large) overhead when building from the toplevel boost directory with the --with-<lib> option. Take Boost.Signals as an example. When I build in boost/libs/signals/build using bjam, the subproject (not only signals) seems not to support the stage target so that the resulting binary/ies are build somewhere in boost/bin. Now I would like to create a Jamfile outside the boost source tree which does two things: 1) it builds selected (sub)variants of boost_signals, perhaps using some accordingly named NOTFILE targets, 2) it let's me copy the resulting binary/ies (i.e. libs and dlls) to a specified directory. Unfortunately I seem not to be able to figure out how to do this with bjam, especially how to get the full path of the built binary/ies. I would really like to avoid hardcoding these paths. So any help is appreciated. TIA, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | andreas . ames AT comergo . com

3 2

[bjam] [v1] Minor issue with vc-7_1
by Ames Andreas 24 Aug '05

24 Aug '05

Hi, when I use bjam -sTOOLS=vc-7_1 "-sBUILD=debug <debug-store>database" lib I get LNK4044 linker warning because the /PDB command line switch is not supported (but ignored) for archive targets. I suppose the vc-Archive action in msvc-tools.jam shouldn't use PDB_LINKFLAG because I can't see why it should be needed. cheers, aa -- Andreas Ames | Programmer | Comergo GmbH | Voice: +49 69 7505 3213 | andreas . ames AT comergo . com

1 0

Re: [Boost-users] regex_replace with file streams
by louis 24 Aug '05

24 Aug '05

Thanks for the help. My platform is Win32 and I was able to create a class: class mystring: public std::iterator<std::bidirectional_iterator_tag, char > mystring class opens the input file and maps it into memory using: CreateFileMapping MapViewOfFile and I pass this into regex_replace. It seems to work - any thoughts? Thanks again Louis > -----Original Message----- > From: boost-users-bounces(a)lists.boost.org > [mailto:boost-users-bounces@lists.boost.org] On Behalf Of John Maddock > Sent: Tuesday, August 23, 2005 11:08 AM > To: boost-users(a)lists.boost.org > Subject: Re: [Boost-users] regex_replace with file streams > >>I want to use regex_replace to operate on a potentially large input >>file and output the results to another file. So, I can't load the >>entire file into a buffer in memory and pass it in as is done in the >>examples. I can't seem to figure out how to pass fstream or some other >>file I/O into regex_replace (ie create the proper iterator). > > Take a look at the second example in > http://www.boost.org/libs/regex/doc/partial_matches.html > > Unfortunately if you want to do a search and replace in an "infinite" text, > then you're not going to be able to use regex_replace. Instead you'll have > to enumerate through the matches and copy unmatched sections to output, and > then call match_results::format on matched sections to get the text to > output for that section. > > HTH, John. > > _______________________________________________ > Boost-users mailing list > Boost-users(a)lists.boost.org > http://lists.boost.org/mailman/listinfo.cgi/boost-users >

2 1