今天运营出了点问题，需要对特定时间段充值数做一个处理，文件格式有特定要求，要符合erlang的格式
{roleID,gold}.
mysql导出所有数据
结果如下【取部分数据看】：
===================================
kuwo 4 50004106230500 100
kuwo 4 50004106230900 10
agent server role_id ****
uu178 5 380005100002500 300
agent server role_id ****
uu178 6 380006100000600 200
uu178 6 380006100110500 20
agent server role_id ****
yy365 1 860001100008000 2000
yy365 1 860001100008100 1900
。
。
。
===================================
===================================
首先去掉表头信息：
cat log.txt | grep -v agent
kuwo 4 50004104005100 50
kuwo 4 50004104284100 1000
kuwo 4 50004105824300 1000
kuwo 4 50004106230500 100
kuwo 4 50004106230900 10
uu178 5 380005100002500 300
uu178 6 380006100000600 200
uu178 6 380006100110500 20
yy365 1 860001100008000 2000
yy365 1 860001100008100 1900
==============================
去掉第一列每二列
cat log.txt | grep -v agent | awk '{print $3,$4}'
50004104005100 50
50004104284100 1000
50004105824300 1000
50004106230500 100
50004106230900 10
380005100002500 300
380006100000600 200
380006100110500 20
860001100008000 2000
860001100008100 1900
拼凑格式
cat log.txt | grep -v agent | awk '{print $3,$4}'| awk '{print "{"$1",",$2"}."}'
{50004104005100, 50}.
{50004104284100, 1000}.
{50004105824300, 1000}.
{50004106230500, 100}.
{50004106230900, 10}.
{380005100002500, 300}.
{380006100000600, 200}.
{380006100110500, 20}.
{860001100008000, 2000}.
{860001100008100, 1900}.
====================================
大功告成，比用脚本去逐行生成文件快多了
最后，整个过程可整理成一条管道命令
cat log.txt | grep -v agent | awk '{print "{"$3",",$4"}."}' > newfile.txt