[BGL] adjacency_list with hash_setS: bug ?

Is there any example using an adjacency_list with hash_setS for EdgeList ? I think there is a mistake in adjacency_list.hpp:21: // ====================== #if !defined BOOST_NO_HASH #include <hash_set> // instead of <ext/hash_set> ? #endif // ====================== I'm using gcc 4.0.2 with boost 1.33.0 and I can't compile a simple example using adjacency_list with hash_setS: // ====================== # if defined __cplusplus # if defined __GNUC__ && __GNUC__ >= 3 # include <ext/hash_set> # include <ext/hash_map> # if __GNUC__ >= 4 || __GNUC_MINOR__ >= 2 namespace std { using ::__gnu_cxx::hash; using ::__gnu_cxx::hash_set; using ::__gnu_cxx::hash_map; } # endif # else # include <hash_set> # include <hash_map> # endif /* defined __GNUC__ and __GNUC__ == 3 */ # endif /* defined __cplusplus */ #define BOOST_HAS_HASH 1 // for boost::hash_setS #include <boost/config.hpp> #include <boost/graph/adjacency_list.hpp> using namespace boost; int main() { typedef adjacency_list<hash_setS, vecS, bidirectionalS> Graph; Graph g; } // ========================== -- Johan

Johan Oudinet wrote:
Is there any example using an adjacency_list with hash_setS for EdgeList ?
I think there is a mistake in adjacency_list.hpp:21: // ====================== #if !defined BOOST_NO_HASH #include <hash_set> // instead of <ext/hash_set> ? #endif // ======================
I'm using gcc 4.0.2 with boost 1.33.0 and I can't compile a simple example using adjacency_list with hash_setS:
Yep, this is a bug. When hash_setS support was added to the Graph library, most compilers had it in <hash_set>. It's since moved to different places on different compilers. Would you mind filing a bug in the Sourceforge bug tracker? Doug

On 11/8/05, Douglas Gregor <doug.gregor@gmail.com> wrote:
Johan Oudinet wrote:
I'm using gcc 4.0.2 with boost 1.33.0 and I can't compile a simple example using adjacency_list with hash_setS:
Yep, this is a bug. When hash_setS support was added to the Graph library, most compilers had it in <hash_set>. It's since moved to different places on different compilers.
Ok, thanks for the reply.
Would you mind filing a bug in the Sourceforge bug tracker?
Done. -- Johan
participants (2)
-
Douglas Gregor
-
Johan Oudinet