Limit one file with IDs of another file

cp167 · 09-14-2012, 03:29 AM

Hi,

I have two files:

File1:
1 abc def
3 ghi jkl
4 mno pqr
5 stu vwx

File2:
1
5

And I would like to restrict File1 with IDs of File2:

1 abc def
5 stu vwx

Do anyone have an idea and could please help me?

druuna · 09-14-2012, 03:47 AM

Try this:

Code:

awk 'BEGIN { while ( ( getline < "file2" ) > 0 ) xz[$1] = $0 } { if ( xz[$1] ) {   print } }' file1

Example:

Code:

$ cat file1
1 abc def
3 ghi jkl
4 mno pqr
5 stu vwx

$ cat file2
1
5

$ awk 'BEGIN { while ( ( getline < "file2" ) > 0 ) xz[$1] = $0 } { if ( xz[$1] ) {   print } }' file1
1 abc def
5 stu vwx

David the H. · 09-16-2012, 06:11 AM

The simplest solution is to just use the file as a pattern list:

Code:

grep -f file2.txt file1.txt

The only problem with this is that it will match numbers found anywhere in the line, not just at the front. But you can use bash's process substitution to pre-filter the list and turn the entries into regex patterns.

Code:

grep -f <( sed -n '/^$/! s/^/^/p' file2.txt ) file1.txt

The sed command adds a "^" to the start of each non-empty line, so that grep will treat is as "<start of line><number>".

Another option would be to loop through the file:

Code:

while read -r num; do
	if [[ $num ]]; then
		grep "^$num" file1.txt
	fi
done <file2.txt

But this does require a separate instance of grep for each line, and so is less efficient.