Accessing data in hashed_unique in boost::multi_index

Question

I'm trying for the first time to work with boost's multi-index and I can't seem to be able to wrap my head around all the code I see on the web.

First of all: My aim is to have a container with an enum as the key for direct access as well as to be able to iterate over it based on the original order of insertion.

I have defined my boost elements like so:

struct VarMapEle {
    SolutionVariableNames first;
    uint    second;
};

struct var_tag {};
struct rand_tag {};

typedef multi_index_container<
    VarMapEle,
    indexed_by< 
        random_access<tag<rand_tag>>, // this index represents insertion order
        hashed_unique<tag<var_tag>, member<VarMapEle, SolutionVariableNames,     &VarMapEle::first>>
        >
> VariableMap;

How can I do either of the tasks I mentioned before?

Answer 1

A multi_index_container can be thought of as a bunch of containers (in your example, one std::vector-like --random_acces-- and the other similar to a std::unordered_set --hashed_unique--) which happen to act on the same underlying collection of elements. Each "container" or index is accessed with get<tag>() or get<n>()where n is the 0-based order of the index as specified in the indexed_by section. So, to iterate over the elements in their order of insertion you need to access index #0 and use it as you'd do with a std::vector:

for(const VarMapEle& e:m.get<rand_tag>()){
    std::cout<<e.first<<","<<e.second<<"\n";
}

Similarly, lookup facilities are provided by index #1:

auto it=m.get<var_tag>().find(SolutionVariableNames{1});
std::cout<<it->second<<"\n";