How to create dynamic substring with awk [closed] - bash

Question

Closed. This question needs details or clarity. It is not currently accepting answers.
Want to improve this question? Add details and clarify the problem by editing this post.
Closed 7 years ago.
Improve this question
Let say i have file like below.
ABC_DEF_G-1_P-249_8.CSV
I want to cut to be like this below.
ABC_DEF_G-1_P-249_
I use this awk command to do that like below.
ls -lrt | grep -i .CSV | tail -1 | awk -F ' ' '{print $8}' | cut -c 1-18
Question is, if the number 1, is growing, how to make the substring is dynamic
example like below...
ABC_DEF_G-1_P-249_
....
ABC_DEF_G-10_P-249_
ABC_DEF_G-11_P-249_
...
ABC_DEF_G-1000_P-249_

Answer 1

To display the file names of all .CSV without everything after the last underscore, you can do this:
for fname in *.CSV; do echo "${fname%_*}_"; done
This removes the last underscore and evertyhing that follows it (${fname%_*}), and then appends an underscore again. You can assign that, for example, to another variable.
For an example file list of
ABC_DEF_G-1_P-249_9.CSV
ABC_DEF_G-10_P-249_8.CSV
ABC_DEF_G-1000_P-249_4.CSV
ABC_DEF_G-11_P-249_7.CSV
ABC_DEF_G-11_P-249_7.txt
this results in
$ for fname in *.CSV; do echo "${fname%_*}_"; done
ABC_DEF_G-1_P-249_
ABC_DEF_G-10_P-249_
ABC_DEF_G-1000_P-249_
ABC_DEF_G-11_P-249_

Answer 2

You can do this with just ls and grep
ls -1rt | grep -oP ".*(?=_\d{1,}\.CSV)"
If you are concerned about the output of ls -1, as mentioned in the comments you can use find as well
find -type f -printf "%f\n" | grep -oP ".*(?=_\d{1,}\.CSV)"
Outputs:
ABC_DEF_G-1_P-249
ABC_DEF_G-1000_P-249_
This assumes you want everything except the _number.CSV, if it needs to be case insensitive then you can the -i flag to the grep. The \d{1,} allows for the number between _ and .CSV to grow from one to many digits. Also doing it this way you don't have to worry about if the number 1 in your example increases:
ABC_DEF_G-1_P-249

Answer 3

You should not be parsing ls. Perhaps you are looking for something like this:
base=$(printf "%s\n" * | grep -i .CSV | tail -1 | awk -F ' ' '{print $8}' | cut -c 1-18)
However, that's a useless use of grep you want to get rid of right there -- Awk does everything grep does, and everything tail does, too, and actually, everything cut does as well. The grep can also be avoided by using a better wildcard, though:
base=$(printf "%s\n" *.[Cc][Ss][Vv] | awk 'END { print substr($8, 1, 18) }')
In the shell itself, you can do much the same thing with no external processes at all. Proposing a suitable workaround would perhaps require a better understanding of what you are trying to accomplish, though.