invalid_type_inside_abi exception

[heifner]

The invalid_type_inside_abi is defined as

      FC_DECLARE_DERIVED_EXCEPTION( invalid_type_inside_abi,           abi_exception,
                                    3015004, "The type defined in the ABI is invalid" ) // Not to be confused with abi_type_exception

Note the comment Not to be confused with abi_type_exception. Looks like it has been confused. It should be used only when validating the ABI.
I believe these uses in abi_serializer::_variant_to_binary are incorrect. The exception there should be abi_type_exception.

leap/libraries/chain/abi_serializer.cpp

Line 595 in d2af863

EOS_THROW( invalid_type_inside_abi, "Unknown type ${type}", ("type",ctx.maybe_shorten(type)) );

leap/libraries/chain/abi_serializer.cpp

Line 575 in d2af863

EOS_ASSERT( st.base == type_name(), invalid_type_inside_abi,

leap/libraries/chain/abi_serializer.cpp

Line 356 in d2af863

EOS_ASSERT( s_itr != structs.end(), invalid_type_inside_abi, "Unknown type ${type}", ("type",ctx.maybe_shorten(type)) );

[bhazzard]

This can comfortably slide into 6.0.0 instead of 5.x if necessary... fit into 5.0.1 opportunistically.