Sorting 按两行分组对文本文件排序
我有一个具有以下结构的文件：Sorting 按两行分组对文本文件排序,sorting,awk,sed,Sorting,Awk,Sed,我有一个具有以下结构的文件： A 35 74 dsadasd/1 0 + A 95 74 dsadasd/2 0 - B 75 159 dsadasd/2 0 + B 78 852 dsadasd/1 0 - C 12 789 dsadasd/1 0 + C 91 546 dsadasd/2 0 - A 87 52 dsadasd/2 0 + A 52 15 dsadasd/1 0
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 75 159 dsadasd/2 0 +
B 78 852 dsadasd/1 0 -
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 87 52 dsadasd/2 0 +
A 52 15 dsadasd/1 0 -
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$ awk '{
k=NR%2; a[k]=$4; b[k]=$0 # store compare value and
} # record for 0 and 1
!(NR%2) { # on even we compare
print b[(a[0]>a[1])] ORS b[(a[0]<=a[1])] # and print the smaller first
}' file
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$awk'{
k=NR%2；a[k]=4；b[k]=0#存储比较值和
}#记录0和1
!（NR%2）{即使我们比较
在awk中打印b[（a[0]>a[1]）]或b[（a[0]：
$ awk '{
k=NR%2; a[k]=$4; b[k]=$0 # store compare value and
} # record for 0 and 1
!(NR%2) { # on even we compare
print b[(a[0]>a[1])] ORS b[(a[0]<=a[1])] # and print the smaller first
}' file
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$awk'{
k=NR%2；a[k]=4；b[k]=0#存储比较值和
}#记录0和1
！（NR%2）{即使我们比较
打印b[（a[0]>a[1]）]或b[（a[0]这应该与awk
一起使用：
awk '{if(p==""){p=$0;p4=$4}
else{
if(p4>$4){print $0"\n"p}
else{print p"\n"$0};p=p4=""
}}' file
这应该适用于awk
：
awk '{if(p==""){p=$0;p4=$4}
else{
if(p4>$4){print $0"\n"p}
else{print p"\n"$0};p=p4=""
}}' file
应该有一个更简单的方法，但这是可行的
$ awk '{c+=p!=$1; p=$1; print c "\t" $0}' file | sort -k1,1 -k5 | cut -f2-
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
基于第一个字段组创建一个组id，先按它排序，然后按另一个键字段排序；删除虚拟组id。应该有一种更简单的方法，但这是可行的
$ awk '{c+=p!=$1; p=$1; print c "\t" $0}' file | sort -k1,1 -k5 | cut -f2-
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
基于第一个字段组创建一个组id，先按它排序，然后按另一个键字段排序；删除虚拟组id。尝试Perl。请注意，这会保留输入中的空格
perl -0777 -ne ' while( /(.+?)\n(.+?)\n/gms ) { $a=$1;$b=$2; (split(/\s+/,$a))[3] gt (split(/\s+/,$b))[3] ? print "$b\n$a\n" : print "$a\n$b\n" }'
有投入
$ cat angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 75 159 dsadasd/2 0 +
B 78 852 dsadasd/1 0 -
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 87 52 dsadasd/2 0 +
A 52 15 dsadasd/1 0 -
$ perl -0777 -ne ' while( /(.+?)\n(.+?)\n/gms ) { $a=$1;$b=$2; (split(/\s+/,$a))[3] gt (split(/\s+/,$b))[3] ? print "$b\n$a\n" : print "$a\n$b\n" }' angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$
请注意，这会保留输入中的空格
perl -0777 -ne ' while( /(.+?)\n(.+?)\n/gms ) { $a=$1;$b=$2; (split(/\s+/,$a))[3] gt (split(/\s+/,$b))[3] ? print "$b\n$a\n" : print "$a\n$b\n" }'
有投入
$ cat angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 75 159 dsadasd/2 0 +
B 78 852 dsadasd/1 0 -
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 87 52 dsadasd/2 0 +
A 52 15 dsadasd/1 0 -
$ perl -0777 -ne ' while( /(.+?)\n(.+?)\n/gms ) { $a=$1;$b=$2; (split(/\s+/,$a))[3] gt (split(/\s+/,$b))[3] ? print "$b\n$a\n" : print "$a\n$b\n" }' angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$
awk+排序
$ awk ' { $(NF+1)=int((NR+1)/2) } 1 ' angel.txt | sort -k7,7 -k4,4 | awk ' {$NF=""}1 '
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$ cat angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 75 159 dsadasd/2 0 +
B 78 852 dsadasd/1 0 -
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 87 52 dsadasd/2 0 +
A 52 15 dsadasd/1 0 -
$
awk+排序
$ awk ' { $(NF+1)=int((NR+1)/2) } 1 ' angel.txt | sort -k7,7 -k4,4 | awk ' {$NF=""}1 '
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 78 852 dsadasd/1 0 -
B 75 159 dsadasd/2 0 +
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 52 15 dsadasd/1 0 -
A 87 52 dsadasd/2 0 +
$ cat angel.txt
A 35 74 dsadasd/1 0 +
A 95 74 dsadasd/2 0 -
B 75 159 dsadasd/2 0 +
B 78 852 dsadasd/1 0 -
C 12 789 dsadasd/1 0 +
C 91 546 dsadasd/2 0 -
A 87 52 dsadasd/2 0 +
A 52 15 dsadasd/1 0 -
$
差不多，但../1和../2的顺序颠倒了。u.uOk.几乎改变了，但../1和../2的顺序颠倒了。u.uOk.改变了，效果很好，唯一的细节是我不确定为什么，但我用命令u在每一行后面多了一行。u.uWorked nice，唯一的细节是我不确定为什么，但我用命令在每一行后面多了一行u.u工作得很好，还感谢您扩展了命令的解释！工作得很好，还感谢您扩展了命令的解释！非常好！感谢Perl示例。非常好！感谢Perl示例。