Akame is training on an infinite 2D Cartesian plane. There is an infinitely long rigid and vertical string at each integer x-coordinate. She takes out her katana, Murasame, and makes infinitely long slashes on the plane. Each slash can be seen as a line with the equation , for some , , and . A slash will never be a vertical line.

Each string Akame slashes will immediately be broken into two smaller strings. For example, a string originally from to severed by the line will be broken into the two strings to and to . A string is considered disconnected if it is of a finite length. For example, the string from to is not of finite length, but the string from to is.

There are points on strings Akame wants to disconnect. A point is said to be disconnected if it is either slashed or the string it's on is disconnected. To measure her performance, Akame would like to know the earliest moment (that is, after which slash) each point is disconnected. The slashes are numbered from to .

#### Input Specification

The first line of input will have and , separated by a single space.

The next lines of input will each describe one of Akame's slashes, with , , and separated by spaces.

The next lines of input will each describe a point Akame wants to disconnect, with and .

#### Output Specification

There should be lines of output. The line should contain the smallest index of the slash that after it is performed, the point is disconnected. If the point is never disconnected, print `-1`

instead.

#### Constraints

#### Sample Input 1

```
3 5
0 1 3
0 1 5
0 1 7
1 2
1 3
1 4
1 5
1 6
```

#### Sample Output 1

```
-1
1
2
2
3
```

#### Sample Input 2

```
10 10
81 -36 72
64 -69 24
23 47 47
83 36 18
1 25 77
12 -81 -3
13 -90 -34
23 19 -34
19 23 78
50 -96 82
56 59
85 47
46 -23
1 13
-74 -69
23 99
-98 72
-31 -65
-78 76
9 -69
```

#### Sample Output 2

```
2
3
4
-1
2
-1
4
2
4
-1
```

## Comments

OK the issues were silently fixed

Seems like something is seriously screwed in input? assert(scanf(“%d %d”,&n,&m) == 2) gives RTE for all cases

A recent update broke several problems including this one. It will (hopefully) be fixed soon.

It seems that some of tests have N=0, can you please fix this problem(I submitted a program which exits on zero input and has an infinite loop on other inputs, there are a couple of tests where I get WA, so they have N=0)?