Warb

Task

Gigel, a renowned computer scientist (nicknamed the "shining star of Romanian computer science"), wants to color a beautiful tree using $m$ colors, encoded by natural numbers from $1$ to $m$.

After assigning each color $c$ a weight $w_c$ and each node $u$ a value $req_u$, Gigel stated that a coloring of the tree nodes is good if each node $u$ meets one of the following conditions:

1. Node $u$ has color $1$; or
2. If node $u$ has color $c > 1$, then it has exactly $req_u$ neighbors with color $c - 1$.

Gigel also defines the value of such a coloring as $w_{c_1} + w_{c_2} + \ldots + w_{c_n}$, where $c_u$ represents the color of node $u$.

Help Gigel determine the highest value of a good coloring!

Input data

The first line of the input file warb.in contains two natural numbers $n$ and $m$ — the number of nodes in the tree and the number of colors, respectively.
The second line contains $m$ natural numbers $w_1, w_2, \ldots, w_m$ — the weights of the $m$ colors.
The third line contains $n$ natural numbers $req_1, req_2, \ldots, req_n$ — the values associated with the $n$ nodes.
The next $n-1$ lines contain $2$ natural numbers $u$ and $v$, indicating that there is an edge in the tree between nodes $u$ and $v$.

Output data

On the first line of the output file warb.out, print a single natural number, the highest value of a good coloring.

Constraints and clarifications

• $1 \leq n \leq 10^5$
• $1 \leq m \leq 100$
• $1 \leq w_i \leq 10^4$
• $0 \leq req_i \leq$ degree of node $i$
• It is guaranteed that the graph in the input file is a tree (a connected acyclic undirected graph).
• To obtain points for a particular subtask, at least one submitted source code will have to pass all the tests of that subtask.

Example 1

warb.in

3 2
1 2
1 1 1
1 2
1 3

warb.out

5

Explanation

For the first example, an optimal coloring is $c = [1, 2, 2]$:

• Node $1$ has color $1$.
• Node $2$ has color $2$ and exactly one neighbor with color $c_2 - 1 = 1$.
• Node $3$ has color $2$ and exactly one neighbor with color $c_3 - 1 = 1$.
• The value of this coloring is equal to $w_{c_1} + w_{c_2} + w_{c_3} = w_1 + w_2 + w_2 = 1 + 2 + 2 = 5$.

Example 2

warb.in

5 4
1 2 3 4
2 3 1 1 1
1 2
1 3
2 4
2 5

warb.out

8

Explanation

For the second example, an optimal coloring is $c = [2, 1, 1, 2, 2]$:

• Node $1$ has color $2$ and exactly two neighbors with color $c_1 - 1 = 1$.
• Node $2$ has color $1$.
• Node $3$ has color $1$.
• Node $4$ has color $2$ and exactly one neighbor with color $c_4 - 1 = 1$.
• Node $5$ has color $2$ and exactly one neighbor with color $c_5 - 1 = 1$.
• The value of this coloring is equal to $w_{c_1} + w_{c_2} + w_{c_3} + w_{c_4} + w_{c_5} = 2 + 1 + 1 + 2 + 2 = 8$.

Example 3

warb.in

7 4
1 9 8 7
2 2 0 0 1 1 1
1 2
1 3
2 4
2 5
3 6
3 7

warb.out

46

Explanation

For the third example, an optimal coloring is $c = [2, 1, 1, 3, 2, 2, 2]$.

Example 4

warb.in

6 5
7 9 50 30 80
0 1 0 0 1 0
1 2
2 3
3 4
3 5
5 6

warb.out

380

Explanation

For the fourth example, an optimal coloring is $c = [4, 5, 5, 5, 5, 4]$.

Example 5

warb.in

7 7
458 2960 8526 2565 6679 9621 8814
1 0 0 0 2 1 1
3 4
1 2
1 4
5 7
4 5
4 6

warb.out

49102

Explanation

For the fifth example, an optimal coloring is $c=[7,7,6,6,1,7,2]$.

Example 6

warb.in

7 7
3564 9408 8285 6312 7323 3223 4441
0 0 0 1 1 1 2
3 2
5 3
4 1
7 3
4 7
6 7

warb.out

54831

Explanation

For the sixth example, an optimal coloring is $c=[2,2,4,2,5,2,1]$.