output of head command in a variable preserving newline characters in bash script

Asked Apr 13 '17 at 16:02

Active Apr 13 '17 at 16:02

Viewed 407 times

Let's say, I have a file abc.txt and there is a pattern some_pattern in the file and I want to print first column of that line.

cat abc.txt | grep some_pattern | awk '{print $1}'

I know in advance that the pattern is present in first 10 lines of the file. So, wanted to optimize using:

head -10 abc.txt | grep some_pattern | awk '{print $1}'

But what if now I had to search for another pattern some_pattern2 which is again in first 10 lines of the file? I wish I could store the output of head command in some variable like:

trimmed_file=`head -10 abc.txt`
echo $trimmed_file | grep some_pattern | awk '{print $1}'

But unfortunately, $trimmed_file has first 10 lines of file with no newline character due to which there is just a single line and it always print the first word of file.

asked Apr 13 '17 at 16:02

Vaibhav Nigam

1,334
12
21

3

Btw: `awk '/some_pattern/ {print $1}' abc.txt` – Cyrus Apr 13 '17 at 16:06
2

It is your _unquoted reference to_ `$trimmed_file` that makes the newlines "disappear" - simply use `"$trimmed_file"`. – mklement0 Apr 13 '17 at 16:07
Thanks @Cyrus. You made my day. :) – Vaibhav Nigam Apr 13 '17 at 16:15
`head -10 abc.txt | tee >(awk '/some_pattern/ {print $1}') | awk '/some_pattern2/ {print $1}'` (use as many `tee` as you have patterns to chain). This way you never store the output of head which, IMHO, was bad practice. – vdavid Apr 13 '17 at 16:17
@vdavid - getting error: cannot make pipe for process substitution: Function not implemented – Vaibhav Nigam Apr 13 '17 at 16:38
@lazyvab your Bash version must be too old. Sorry about that. – vdavid Apr 13 '17 at 17:10

output of head command in a variable preserving newline characters in bash script

0 Answers0