CODEWITHSUNDEEP
sed
chunwai
chunwai

Reputation: 63

How to remove comma in the last line of sed address range?

I have a file that contains a value on each line and it ends with a ,. I want to remove the , in the last line of selected address range. I am able to achieve the following with sed -n '1,3p' file | sed '$s/,$//', but is there a simpler way?

Example:

'12345',
'45322',
'90456',
'67895',
...
'34552',

Expected output:

'12345',
'45322',
'90456'

Upvotes: 1

Views: 59

Answers (2)

Cyrus
Cyrus

Reputation: 88766

With GNU sed:

sed -n '1,3{ 3s/,$//;p }' file

Output:

'12345',
'45322',
'90456'

Disadvantage: 3 must be entered twice.

Upvotes: 2

Shawn
Shawn

Reputation: 52529

Just run both commands in one sed invocation, the one to remove the comma first:

$ cat input.txt
foo,
bar,
baz,
quux,
$ sed -n '3s/,$//; 1,3p; 3q' input.txt
foo,
bar,
baz

Note using the same line number for the s/// as for the end of the range to print instead of $.

As an optimization, this also exits after printing the last line of the range instead of continuing to process the rest of the file (And do nothing with it).

Upvotes: 1

Related Questions