Delete last blank line from a file and save the file

Question

How can I delete the last line of a file without reading the entire file or rewriting it in any temp file? I tried to use sed but it reads the entire file into memory which is not feasible.

I want to remove the blank line from the end and save the change to the same file.

Since the file is very big and reading through the complete file would be slow, I am looking for a better way.

score 1 · Answer 1 · answered Apr 06 '17 at 08:55

1

simple sed command to delete last line:

sed '$d' <file>

here in sed $ is the last line.

You can try awk command:

awk 'NR > 1{print t} {t = $0}END{if (NF) print }' file

Using cat:

cat file.txt | head -n -1 > new_file.txt

answered Apr 06 '17 at 08:55

danglingpointer

4,708
3
24
42

David C. Rankin · Answer 2 · 2017-04-06T08:15:34.413

0

The easiest way I can think of is:

sed -i '${/^$/d}' filename

edited to delete only blank end line.

Your only option not using sed that won't read the entire file is to stat the file to get the size and then use dd to skip to the end before you start reading. However, telling sed to only operate on the last line, does essentially that.

edited Apr 06 '17 at 08:15

answered Apr 06 '17 at 08:07

David C. Rankin

81,885
6
58
85

sed -i -u ... -u mean *unbuffered*, so lower memory use – NeronLeVelu Apr 06 '17 at 08:59
In this case since you are jumping to the end, you would certainly want to use a buffered read to speed the process, especially for large files. – David C. Rankin Apr 07 '17 at 05:37
buffered speed the process of writing but the problem is memory here, not process speed says OP – NeronLeVelu Apr 07 '17 at 08:27
1

I get `sed: 1: "${/^$/d}": extra characters at the end of d command`. – Iulian Onofrei Mar 13 '18 at 14:29
2

It is necessary to put a semicolon at the end of 'd', otherwise you will get @IulianOnofrei error – Aug 23 '18 at 09:31
For me, only [`sed -i '.txt' '${/^$/d;}' test.txt`](https://www.mkyong.com/mac/sed-command-hits-undefined-label-error-on-mac-os-x/) worked, because after adding the semicolon it failed with `sed: 1: "test.txt": undefined label 'est.txt'`. – Iulian Onofrei Aug 23 '18 at 09:36
[`sed -i '' '${/^$/d;}' test.txt`](https://mpdaugherty.wordpress.com/2010/05/27/difference-with-sed-in-place-editing-on-mac-os-x-vs-linux/) also works. – Iulian Onofrei Aug 23 '18 at 09:39
Yes, that would work too. Good comment. – David C. Rankin Aug 23 '18 at 09:47

score 0 · Answer 3 · edited May 23 '17 at 11:46

0

Take a look at

Remove the last line from a file in Bash

Edit: I tested

dd if=/dev/null of=<filename> bs=1 seek=$(echo $(stat --format=%s <filename> ) - $( tail -n1 <filename> | wc -c) | bc )

and it does what you want

edited May 23 '17 at 11:46

Community

1
1

answered Apr 06 '17 at 08:14

rinn2883

366
1
14

Delete last blank line from a file and save the file

3 Answers3