让文本飞 linux shell脚本数据挖掘实践
参考书:Linux Shell 脚本攻略
对文件中的以下网址排序,统计出出现次数排名top3的网址。
文件名:
website
http://www.google.com
http://www.baidu.com
http://www.sina.com
http://www.bjtu.edu.cn
http://www.codeproject.com
http://www.csdn.com
http://www.sohu.com
http://www.yahoo.com
http://mail.163.com
http://www.bjtu.edu.cn
http://www.codeproject.com
http://www.csdn.com
http://www.sohu.com
http://www.yahoo.com
http://mail.163.com
http://www.codeproject.com
http://www.csdn.com
http://www.sohu.com
http://www.yahoo.com
http://mail.163.com
http://www.qq.com
http://www.hao123.com
http://www.163.com
http://youku.com
http://taobao/com
http://www.bjtu.edu.cn
http://www.codeproject.com
http://www.csdn.com
http://www.sohu.com
http://www.yahoo.com
http://mail.163.com
http://www.codeproject.com
http://www.csdn.com
http://www.sohu.com
http://www.yahoo.com
http://mail.163.com
http://www.qq.com
http://www.hao123.com
http://www.163.com
http://youku.com
http://taobao/com
shell脚本如下:
#!/bin/bash
#文件名:websort
#用途:计算文件中网址出现的次数
foo()
{
if [ $# -ne 1 ];
then
echo "Usage:$0 filename";#$0脚本名字
exit -1
fi
filename=$1#函数第一个参数
egrep -o "http://[a-zA-Z0-9.]+\.[a-zA-Z]{2,3}" $filename | \
#网站出现的次数统计,用到关联数组
awk '{ count[$0]++ }#$0代表每行,只需遍历一次
END{ printf("%-30s%s\n","wensite","count");
for(ind in count)
{ printf("%-30s%d\n",ind,count[ind]); }
}'
}
#输出出现次数排名前三位的网址
foo website | sort -nrk 2 | head -n 3 >websorted2.txt;
键入websort
输出
http://www.yahoo.com 5
http://www.sohu.com 5
http://www.csdn.com 5
转载请注明出处 http://write.blog.csdn.net/postlist