Operations on awk fields

Question

I have a file that contains full paths:

/home/usr/file_name_delimited_by_underscore_123.txt  
/home/usr/another_example_456.rar

I would like to print the file name from the path without the extension and to print next to it the string after the last _.

Output:

file_name_delimited_by_underscore_123 123
another_example_456 456

I figured a way to get the desired output using piped awk commands:

cat file | awk -F[/.] '{print $(NF-1)}' | awk -F_ '{print $0" "$NF}'

Is there a way to achieve this without piping?

My question boils down to is it possible to perform actions on the fields parsed by awk?

Thanks for your help.

Answer 1

Yes, you can perform any operations you like on the fields. For example:

$ cat file | awk -F[/.] '{n = split($(NF-1),a,/_/); print $(NF-1)" "a[n]}'
file_name_delimited_by_underscore_123 123
another_example_456 456

Of course, you don't need cat here; you could have awk read the file directly - and since the default output field separator OFS is a space, it would be more idiomatic to write the results as separate output fields instead of a string concatenation:

awk -F[/.] '{n = split($(NF-1),a,/_/); print $(NF-1), a[n]}' file

Answer 2

With any sed:

$ sed 's:.*/\(.*_\(.*\)\)\..*:\1 \2:' file
file_name_delimited_by_underscore_123 123
another_example_456 456

Answer 3

You can use sed:

$ sed -e 's;^.*/\(.*_\)\([0-9]*\)\.[^\.]*$;\1_\2 \2;' file
file_name_delimited_by_underscore_123 123
another_example_456 456

---

^.*/ deletes the path.

\(.*_\) captures the name until the last underscore.

\.[^\.]* removes the extension.

\1\2 \2 replace by the captured groups.