Implement a function `solution(operations, values, seed)` that simulates a multiset of integers supporting duplicates. The multiset must support these operations in expected O(1) time: - `insert(val)`: Add one occurrence of `val`. Return `True` if `val` was not already present in the multiset, otherwise return `False`. - `remove(val)`: Remove one occurrence of `val` if it exists. Return `True` if an element was removed, otherwise return `False`. - `getRandom()`: Return one stored element chosen from all stored occurrences, so duplicates count multiple times. For deterministic judging, model the structure exactly as follows: 1. Keep a dynamic array `data` of all stored occurrences. 2. For each value, keep an index list of all positions where that value appears in `data`. 3. When `remove(val)` is called and multiple occurrences exist, remove the occurrence whose array index is stored at the end of `val`'s index list. 4. During deletion, swap the last array element into the removed slot if needed, and update all bookkeeping in O(1). The per-value index lists do not need to stay sorted. For reproducible `getRandom()` results, do not use a library RNG. Instead maintain an integer `state`, initially equal to `seed`. - If the collection is empty, `getRandom()` returns `None`. - Otherwise update `state = (state * 1103515245 + 12345) % (2**31)`. - Return `data[state % len(data)]`. Return a list containing the result of every operation in order.