Using awk to only grab first instance of a regex

Question

I have a bunch of .txt files in a directory which have information regarding dipole moment. This is how it looks like:

Dipole Moment: [D]
     X:     1.2808      Y:     0.2908      Z:     1.0187     Total:     1.6622
lorem ipsum
text 
that is 
not 
relevant 
Dipole Moment: [D]
     X:     1.2808      Y:     0.2908      Z:     1.0187     Total:     1.6622
more text

I want to extract the total dipole moment from these files. I am running the following script:

awk '/Dipole Moment: \[D\]/{found=1;next} found{print $NF;found=""}' *.txt > dipole_bma.txt

This script prints out 1.6622 twice. And every other Total dipole moment in each file twice. I see that it prints it out twice because the regex appears twice in the file.

My question is, how do I print total dipole moment only once from each file?

I’m voting to close this question because I answered this question for the OP earlier today at https://stackoverflow.com/a/65793414/1745001. — Ed Morton, Jan 19 '21 at 23:25

score 1 · Answer 1 · answered Jan 19 '21 at 18:25

1

Use nextfile statement:

awk '/Dipole Moment: \[D\]/{found=1;next} found{print $NF;nextfile;}' *.txt

answered Jan 19 '21 at 18:25

pLumo

22,565

Thanks for your response @pLumo. However, it doesn't seem to work... I am getting blank lines after the first output... – megamence Jan 19 '21 at 18:29
with your example it works for me ... do you have some non-unix line endings? – pLumo Jan 19 '21 at 18:30
Technically, the files I am using are .out files, so the final .txt becomes .out. I dont know how that could make a difference... – megamence Jan 19 '21 at 18:36
no, the extension does not make any difference – pLumo Jan 19 '21 at 18:39
@megamence sounds you don't have GNU awk that supports nextfile; instead use another second flag for that like awk 'FNR==1 {once=0; }; /Dipole Moment: \[D\]/ { found=1; next}; found && !once{ print $NF; found=0; once=1; }' file*, but however it still continue reading whole file while it only print once. – αғsнιη Jan 19 '21 at 18:59
that link says that nextfile is Posix since 2012 ... – pLumo Jan 19 '21 at 19:12
@pLumo please see https://unix.stackexchange.com/q/588906/72456 – αғsнιη Jan 19 '21 at 19:20

guest_7 · Answer 2 · 2021-01-20T03:31:42.040

Gnu sed can do as follows:

sed -ns '
  /^Dipole Moment: \[D]/!d
  $!N;/\n/s/.* //p;:n;n;bn
' ./*.txt

-s option (nonPosix) to treat files as separate streams.
-n will inhibit default print of the pattern space prior to fetching the next record.
After the dipole moment line we stick the next line to the patter space.
Remove till the last space (assuming no trailing whitespace) this prints the last field, i.e., the value of dipole moment.
then we skid to the end of current file. The whole process repeats for the next file.

Using awk to only grab first instance of a regex

2 Answers2