CODEWITHSUNDEEP
bashsed
Novice
Novice

Reputation: 1161

sed: How to remove entire line matching specific string?

I am having issues removing the entire line in the sample.txt

   cat sample.txt

   XYZ   -2.4   DNW
   ZYY   -2.4   138
   ZZZ   -3.4   200
   ZZZ   -2.4   DNW 

  sed '/DNW/d' sample.txt >> output.txt

  cat output.txt 

   XYZ   -2.4                 #removes the DNW, but not the entire line
   ZYY   -2.4   138
   ZZZ   -3.4   200
   ZZZ   -2.4

What I need is this: 

    cat output.txt 


   ZYY   -2.4   138      #Need the entire lines removed that matched the 3rd column string DNW
   ZZZ   -3.4   200

I am new to bash and was wondering whether there is an option to remove the entire line in the text file matching the search criteria?

Thank you!

p.s. I would be interested in possible solutions primarily using bash. However, I am starting to play with python as well and if there are solutions I'd be glad to learn those as well.

UPDATE

It turns out that my original sample.txt file was not formatted somehow correctly. The following fixed the issue, as it changed the rows into the comma delimited format (e.g. x, y, c = treated as a line). 

   cp sample.txt sample.csv
   sed '/DNW/d' sample.csv > output.txt #Please note any of the below user suggested answers/solutions work!

Cheers and thanks for all the help!

Upvotes: 6

Views: 11720

Answers (5)

Blaine McDonnell
Blaine McDonnell

Reputation: 11

Using the -i switch for sed will update the file based on pattern

sed -rie 's/DNW.+$//g' file

Upvotes: 1

Jotne
Jotne

Reputation: 41460

You can also use awk

awk '!/DNW/' file
   ZYY   -2.4   138
   ZZZ   -3.4   200

Upvotes: 0

jshort
jshort

Reputation: 1016

Your sed command works for me on OSX. Try to be super explicit and try this:

sed '/^.*DNW.*$/d' sample.txt >> output.txt

Upvotes: 2

abarnert
abarnert

Reputation: 366223

This is even easier with grep than sed:

grep -v DNW sample.txt >> output.txt

If you want to do it in Python, it's a lot more verbose, but not actually much harder:

with open('sample.txt') as fin, open('output.txt', 'a') as fout:
    for line in fin:
        if 'DNW' not in line:
            fout.write(fin)

Or, if you want it a little briefer (but probably harder for a novice to understand):

with open('sample.txt') as fin, open('output.txt', 'a') as fout:
    fout.writelines(line for line in fin if 'DNW' not in line)

Upvotes: 6

Jakub Kotowski
Jakub Kotowski

Reputation: 7581

You got it nearly right:

sed '/DNW/d' sample.txt >> output.txt

Upvotes: 9

Related Questions