How would one efficiently implement a nested structure containing arrays of other nested structures in a language that doesn't support them? -

structure

this main structure need build, creating type tparent (pseudocode, () array declaration):

type ne {     single d()     dword dc     single b     single v     single d }  type l {     ne ns()     dword nc }  type tparent {     l la()     dword lc     single lr }  tparent ntest   // accessed ntest.la(0).ns(0).b = 42 ... 

however, proprietary language has used not feature "arrays of user defined types". nesting supported, apart () array declarations in types above, code valid.

question

what efficient way - algorithmically speaking - implement structure. thought expanding subtypes arrays, , accessing them using second array holds offsets, doubt efficient in way.

all can suggest create array of byte , work work memory in assembly languages. can hide complexity of memory access behind accessor methods like

 single typename_fieldname_get(byte arr(), int offset) {       return convertfombytetosingle(           arr(offset+offset_of_field_name + 0),           arr(offset+offset_of_field_name + 1),           arr(offset+offset_of_field_name + 2),           ....           );  }