Extract substring bash

Question

i have this string:

4017de9511f7    workflow-engine-ms    0.12%     4.609MiB / 7.574GiB  0.06%         29MB / 15.8MB     4.1kB / 0B          13

I have to extract the number 4.609, how can I do it in bash? thx!!

Not an exact dupe. The reference question and answers have nothing to do with further splitting the desired field further. — David C. Rankin, Jan 29 '19 at 08:56

score 1 · Answer 1 · answered Jan 29 '19 at 08:40

1

A simple way using awk and sprintf would be:

awk '{val=sprintf ("%g",$4); print val}' file
4.609

Using the "%g" conversion specifier, simply converts the 4.609 portion of the 4th field, ignoring the MiB characters.

answered Jan 29 '19 at 08:40

David C. Rankin

81,885
6
58
85

score 0 · Answer 2 · answered Jan 29 '19 at 08:18

0

If the format does not change, you can use sed -e 's# *# #g' | cut -d\ -f4 | cut -dM -f1 to extract the value.

The sed joins subsequent spaces into one, the first cut extract the 4.609MiB field, and the second one strips of the MiB.

answered Jan 29 '19 at 08:18

Rudi

19,366
3
55
77

score 0 · Answer 3 · answered Jan 29 '19 at 08:33

If by 'I have this string' you mean, you have it in some file and you need to grab first match of X.XXX Mib (or GiB, or kB), I've solved that like this:

echo "4017de9511f7    workflow-engine-ms    0.12%     4.609MiB / 7.574GiB  0.06%         29MB / 15.8MB     4.1kB / 0B          13" > test

to prepare test file, and

grep -Po "[0-9\.]+(?=(Mi|Gi|k))" test | head -1

to find matching string

Extract substring bash

3 Answers3