I need to use grep -P regular expression to find and replace two columns in a bash script

Question

If I have a list such as:

5       chr1
5       chr1
5       chr2
3   chr5

How do I use grep -P to switch the position of column 1 and 2?

If I use this line:

grep -P "(\w+)\t(\w+)" test.txt

to capture the text, how would I then replace it?

Answer 1

Better (easier) to do that in awk:

awk '{a=$1; $1=$2; $2=a}1' OFS='\t' file
chr1    5
chr1    5
chr2    5
chr5    3

Even simpler is: (thanks to @BrenoZan)

awk '{print $2"\t"$1}' file

Answer 2

You could do this

grep -oP "\w+" file | rev | paste - - | rev

but you wouldn't

Answer 3

您可以使用@@ anubhava提及之类的awk （ 我爱awk ）可以正常工作，但是如果您要更改文件内容，请尝试sed ...

 sed 's/\([^ ]\+\)\s*\(.*\)$/\2\t\1/' -i file

Answer 4

与纯壳。

while read -r a b X; do echo $b $a; done < test.txt