Description

There are $m$ events in total, and the events are divided into the following $3$ types.

1. Xiao Z receives $a$ yuan of lucky money.
2. Xiao Z spends $a$ yuan of lucky money to buy skins.
3. Xiao Z seals $a$ yuan of his money. The seal will be lifted $1$ second before event $b$ happens. It is guaranteed that each time, the money Xiao Z currently has is greater than or equal to the amount being sealed.

If Xiao Z does not have enough money to spend on some event, Xiao Z will feel unhappy. At the same time, when he does not have enough money, Xiao Z will not spend the money.

Please tell Xiao Z in how many events he will not have enough money to spend.

Input Format

The first line contains an integer $m$, which indicates the total number of events.

The next $m$ lines each start with an integer $t$, indicating the type of the event.

If $t=1$ or $t=2$, then it is followed by an integer $a$.

If $t=3$, then it is followed by two integers $a,b$.

Output Format

Output one line with one integer, indicating the number of events in which there is not enough money to spend.

3
1 10
2 20
2 10

1

5
1 10
3 5 5
2 10
1 10
2 20

1

Hint

[Sample 1 Explanation]

Day 1: Income $10$ yuan, balance $10$ yuan.

Day 2: Not enough to spend $20$ yuan, balance $10$ yuan.

Day 3: Spend $10$ yuan, balance $0$ yuan.

Total: $1$ day.

[Sample 2 Explanation]

Day 1: Income $10$ yuan, balance $10$ yuan.

Day 2: Seal $5$ yuan, balance $5$ yuan.

Day 3: Not enough to spend $10$ yuan, balance $5$ yuan.

Day 4: Income $10$ yuan, balance $15$ yuan.

Day 5: The seal is lifted, spend $20$ yuan, balance $0$ yuan.

Total: $1$ day.

[Constraints]

For $50\%$ of the testdata, $m,a \leq 5\times 10^3$.

For $70\%$ of the testdata, $t \ne 3$.

For $100\%$ of the testdata, $1 \leq b \le m \leq 10^6$, $1 \le t \le 3$, $1 \leq a \leq 10^7$.

Translated by ChatGPT 5