**Evaluating solutions to intractable problems**

There are many relatively easy and quick-to-solve problems. But there are also many that are just plain hard. In such a scenario, when a problem is unsolvable exactly in practice, we are usually happy enough to have any solution which gives results close enough to the optimum.

This type of problem makes for a great test: giving a programmer an impossible task and seeing how close they can get to solving the unsolvable. It takes lots of creativity and intuition.

**Sample challenge**

Imagine the following real life setting of the following: the so-called Dominating Set Problem. Let's say there are some cities connected by roads. There is only at most one road between any two cities (so some cities are not connected directly). A company which sells smartphones would like to take over as much of the market as possible. To do that they would have to own a store in some cities, so that for every city there either is a store in this city or in a neighboring one. Of course, the company wants to spend as little money as possible so the number of cities in which they will build a store should be minimal. Having a map with cities and roads, the programmer's task is to determine in which cities the company should build a store.

All known exact solutions to this problem will take more or less forever to calculate the answer for a large number of cities. One way to approach the problem would be to plow through (almost) all possible sets of cities and check whether they constitute a valid solution - but this would take very long. The programmer may, however, skip most of them, still getting a good result and cutting calculation time dramatically.

##### Input

The first line contains the number of cities **n** (n < 20000). After that there comes a
description of the road connections of each city starting from city number 1 to city number **n**. Each city is described
in two lines. In the first one, there is only the number of the city's neighbours. In the second one, there is a list of its neighbours.

##### Output

The first line should contain one number - the number of cities in which the company should build stores. The second line should contain a list of those cities.

##### Scoring

For each test, the programmer receives as many points as the number of cities in which stores needed to be built, provided that the solution fulfills the requirements. Thus, the smaller the score is, the better the program.

##### Example

**input**
7
2
2 5
3
3 5 6
1
6
1
5
0
1
7
**output**
3
2 4 7
**score**
3

Illustration of the input

**Comment to the example**

As you might have noticed, the output above is not the optimal solution.
Two cities are enough to fulfill the requirements: city 5 and city 6.