Bash creating a string, if string already exists replace it

Question

I would like to achieve this procedure. I have this command: sed -i -e 's/few/asd/g' /usr/local/sbin/.myappenv This command works if in the file .myappenv there is a few string text. But this command fails to simply create the asd text wheter or not the few is found.

So, if there is the few text, it should replace it, if few text is missing, just create it on the first line.

EDIT_:

It should create the string on the first line of the document .myappenv, it should only be created if there is no coincidence.

The file .myappenv should contain =>

asd

If the file is already populated with few just replace it =>

asd

`just create it.` where? On the beginning of the line? On the end? At the 5th character? — KamilCuk, Jul 17 '21 at 17:02
Please add sample input (no descriptions, no images, no links) and your desired output for that sample input to your question (no comment). — Cyrus, Jul 17 '21 at 17:02
Sorry for not specifying that, it should create it on the first line @KamilCuk — pullhyphoonprotocol, Jul 17 '21 at 17:06
I edited the question to explain the problem I am facing @Cyrus — pullhyphoonprotocol, Jul 17 '21 at 17:08
I still don't understand this. There seem to be two scenarios: the file contains `few`, or it does not. Can you show input and desired output for the two scenarios? — Benjamin W., Jul 17 '21 at 17:17
Hi @BenjaminW. the file is a set of env variables, the ``few`` could be found or not, if it is found I should replace it with ``asd`` if on that file, ``few`` is NOT found i just need to add ``asd`` sorry if the explanation is not clear. I am trying my best. — pullhyphoonprotocol, Jul 17 '21 at 17:23
When the original file already has a line with `asd`, the requirements suggest that you want to add an additional line with `asd`. Is that correct? — Walter A, Jul 17 '21 at 21:41

score 3 · Accepted Answer · answered Jul 17 '21 at 17:26

if there is the few text, it should replace it, if few text is missing, just create it on the first line.

That's an if.

if <file has text>; then <replace text>; else <add text to first line>; fi

or in bash:

file=/usr/local/sbin/.myappenv
if grep -q few "$file"; then
     sed 's/few/asd/' "$file"
else
     { 
        echo asd
        cat "$file"
     } > "$file".tmp
     mv "$file".tmp "$file"
fi

How to test if string exists in file with Bash? https://unix.stackexchange.com/questions/99350/how-to-insert-text-before-the-first-line-of-a-file and https://mywiki.wooledge.org/BashGuide/TestsAndConditionals . You might interest yourself in some automation methods, like ansible lineinfile or chezmoi depending on the goal.

score 0 · Answer 2 · answered Jul 17 '21 at 17:59

0

Using a single awk you can use this:

awk 'FNR == NR {if (/few/) {n = 1; nextfile}; next}
FNR == 1 && !n {print "asd"} n {gsub(/few/, "asd")} 1' file file

answered Jul 17 '21 at 17:59

anubhava

761,203
64
569
643

score 0 · Answer 3 · answered Jul 18 '21 at 10:36

This might work for you (GNU sed):

sed 'H;$!d;x;/few/!s/^/asd/;t;s//asd/;s/.//' file

Make a copy of the file in the hold space (first line will have a prepended newline).

At the end of the file, if few is not matched, insert the literal asd.

Otherwise, replace the first occurrence of few with asd and remove the introduced newline at the start of the file.

To replace all occurrences of few with asd use:

sed 'H;$!d;x;/few/!s/^/asd/;t;s//asd/g;s/.//' file

Alternative:

grep -q few file && sed 's/few/asd/' file || (echo asd; cat file)

Bash creating a string, if string already exists replace it

3 Answers3