You have to do several runs. The depth is the number of runs, you have to do.
In every run you have to switch the direction. First run from left, next run from right. Next left...
Every run has these rules:
You start at the first number (from the direction).
Take for every number x the remainder of the pision by the number of still available elements (from
this position!) to have the number for the next decision.
If the remainder-value is smaller than 3, take this number x (NOT the remainder-Value) direct
for the new array and continue with the next number.
If the remainder-value (e.g. 3) is greater than 2, take the next remainder-value-number (e.g. 3)
elements/numbers (inclusive the number x, NOT the remainder-value) as a sub-array in the new array.
Continue with the next number/element AFTER this taken elements/numbers.
Every sub-array in the array is independent and is only one element for the progress on the array.
For every sub-array you have to follow the same rules for unflatten it.
The direction is always the same as the actual run.
Array: [4, 5, 1, 7, 1] Depth: 2 -> [[ 4, [ 5, 1, 7 ] ], 1]
Steps:
First run: (start from left side!)
The first number is 4. The number is smaller than the number of remaining elements, so it is the remainder-value (4 / 5 -> remainder 4).
So 4 numbers (4, 5, 1, 7) are added as sub-array in the new array.
The next number is 1. It is smaller than 3, so the 1 is added direct to the new array.
Now we have --> [[4, 5, 1, 7], 1]
Second run: (start from right side!)
The last number (first from other side) is 1. So the 1 is added direct to the new array.
The next element is the sub-array. So we use the rules for this.
2a.The last number is 7. There are 4 elements in the array. So for the next decision you have to
take the remainder from 7 / 4 -> 3. So 3 numbers (5, 1, 7) are added as sub-array in the
new array.
2b.Now there is the 4 and only one element last in this array. 4 / 1 -> remainder 0. It is smaller
than 3. So the 4 is added direct to the new array.
Now we have --> [[ 4, [ 5, 1, 7 ] ], 1]
