heterogeneous tree container
Hello, I have need for a heterogeneous tree container that holds different types (all of them primitive, int, double...) except for the leaf nodes which would hold a POD. The depth of the tree is N, known at runtime, and constant through all the paths, the actual types of each node are only known at runtime as well. Each level holds the same type. The number of children at 1 level, is variable from 1 parent to the next. root (no data here) | | | type1 type1 type1 | | type2 type2 ... typeN typeN .... typeN There are no paths from root to leaf with a depth different from N. typeN is a struct containing only primitive types, known a compile-time, and instances of it are the result of calling a complex function with arguments of type1.... type_N-1. Note that I will generate the [1, N-1] part of the tree sequentially, but then call the complex function in parallel. I have looked at the ntree.hh but it seems to be a homogenous container. Any reusable container out there for this purpose, or is ntree reusable somehow? Regards,
Any reusable container out there for this purpose, or is ntree reusable somehow? Maybe you could use Boost.Any in the tree.
Best regards,
Mau.
2010/7/8 Hicham Mouline
Hello,
I have need for a heterogeneous tree container that holds different types (all of them primitive, int, double...) except for the leaf nodes which would hold a POD.
The depth of the tree is N, known at runtime, and constant through all the paths, the actual types of each node are only known at runtime as well.
Each level holds the same type. The number of children at 1 level, is variable from 1 parent to the next.
root (no data here)
| | |
type1 type1 type1
| |
type2 type2
...
typeN typeN .... typeN
There are no paths from root to leaf with a depth different from N.
typeN is a struct containing only primitive types, known a compile-time, and instances of it are the result of calling a complex function with arguments of type1.... type_N-1.
Note that I will generate the [1, N-1] part of the tree sequentially, but then call the complex function in parallel.
I have looked at the ntree.hh but it seems to be a homogenous container.
Regards,
_______________________________________________ Boost-users mailing list Boost-users@lists.boost.org http://lists.boost.org/mailman/listinfo.cgi/boost-users
-- Mau
participants (2)
-
Hicham Mouline
-
Mauricio Gomes