Recursive Function return value learning

Analyze the following code

Int dp (int I) {Int & Ans = d [I]; // declare a reference for this table item to simplify read/write operations on it. If (ANS> 0) return ans; ans = 1; for (Int J = 0; j <n; j ++) {If (G [I] [J]) {int TMP = dp (j); ans = ans> TMP + 1? Ans: TMP + 1;} return ans ;}

This is a dynamic planning under a Dag. Is a rectangle nested DP function. The reason for sticking this code is to repeatedly understand how this ANS is used in this recursive function. The for statement block is used to select the maximum value of DP (j) in the current row. ans then acts as Max here.