求一个分行统计的awk

Question

求一个分行统计的awk，多谢多谢！

希望结果：
=========DNS:202.106.0.20
4 202.108.29.211

=========DNS:202.106.196.115
1 202.108.29.211
2 123.129.242.67

测试数据：
=========DNS:202.106.0.20
202.108.29.211
=========DNS:202.106.196.115
202.108.29.211
=========DNS:202.106.0.20
202.108.29.211
=========DNS:202.106.196.115
123.129.242.67
=========DNS:202.106.0.20
202.108.29.211
=========DNS:202.106.196.115
123.129.242.67
=========DNS:202.106.0.20
202.108.29.211

Answer 1

perl -ne 'if($.%2){$a=$_}else{$a{$a}{$_}++}END{for(keys %a){print;for $k (keys %{$a{$_}}){print "$a{$_}{$k} $k"}}}' file

Answer 2

本帖最后由 yinyuemi 于 2011-04-16 02:41 编辑

回复 1# nugget

try:

1. awk -vRS="=========" -vFS="\n"  '
2. NR>1{
3.         ++a[$1"|"$2]
4.         }
5. END{
6.         for (i in a) {
7.                 split(i,b,"|")
8.                 c[b[1]]=c[b[1]] "\n"a[i]" "b[2]}
9.         for(j in c) print RS j,c[j]"\n"
10. }'
11. =========DNS:202.106.0.20
12. 4 202.108.29.211
14. =========DNS:202.106.196.115
15. 2 123.129.242.67
16. 1 202.108.29.211

复制代码

Answer 3

感谢prel不太懂，学习awk的方法，多谢！

Answer 4

有大牛能解释下yinyuemi的awk过程么，c[b[1]]=c[b[1]] "\n"a[i]" "b[2]}这怎么理解