CF549A.Face Detection

The developers of Looksery have to write an efficient algorithm that detects faces on a picture. Unfortunately, they are currently busy preparing a contest for you, so you will have to do it for them.

In this problem an image is a rectangular table that consists of lowercase Latin letters. A face on the image is a 2 × 2 square, such that from the four letters of this square you can make word "face".

You need to write a program that determines the number of faces on the image. The squares that correspond to the faces can overlap.

Looksery 公司的开发者需要编写一个高效的算法，用于在图片中检测人脸。不幸的是，他们目前正在忙于为你准备一场竞赛，因此这项任务将由你来完成。

本题中，一张图片是一个由小写拉丁字母组成的矩形表格。图片中的一张“人脸”定义为一个 $2 \times 2$ 的正方形区域，且该正方形内四个字母可以重新排列组成单词 "face"。

你需要编写一个程序，计算图片中“人脸”的数量。对应于不同“人脸”的正方形区域可以相互重叠。

The first line contains two space-separated integers, n and m (1 ≤ n, m ≤ 50) — the height and the width of the image, respectively.

Next n lines define the image. Each line contains m lowercase Latin letters.

第一行包含两个以空格分隔的整数 $n$ 和 $m$（$1 ≤ n, m ≤ 50$），分别表示图像的高度和宽度。

接下来的 $n$ 行描述该图像。每行包含 $m$ 个小写拉丁字母。

In the single line print the number of faces on the image.

在单行中输出图像中的人脸数量。

In the first sample the image contains a single face, located in a square with the upper left corner at the second line and the second column:

In the second sample the image also contains exactly one face, its upper left corner is at the second row and the first column.

In the third sample two faces are shown:

In the fourth sample the image has no faces on it.

在第一个样例中，图像包含一张人脸，位于一个正方形区域内，该正方形的左上角位于第2行、第2列：

在第二个样例中，图像同样恰好包含一张人脸，其左上角位于第2行、第1列。

在第三个样例中，图像显示了两张人脸：

在第四个样例中，图像中没有人脸。