[SOLVED] How to delete a line from a file with sed following a pattern

Issue

I have a file with the following lines:

auth_service_oauth_secret_accfrsh
auth_service_oauth_secret_accfrshv12
auth_service_oauth_secret_accfrshv13

I would like to be able to remove only the first line, where my pattern will be only "accfrsh", but I tried with multiple sed commands and it's not working. All lines with the pattern "accfrsh" are removed or nothing change.

I tried the below sed commands (it doesn't matter with -e or -i) :

sed -e "/accfrsh/d" my_file (remove all lines)
sed -e 's/\<accfrsh\>//g' my_file (nothing changes)
sed -e "/'^'accfrsh'$'/d" my_file (nothing changes)

Solution

This might work for you (GNU sed):

sed '/_accfrsh\b/d' file

Delete lines which contain _accfrsh followed by a word boundary.

Compare it to:

sed '/_accfrsh\B/d' file

N.B. The \b is related to /< and /> which denote the start and end of a word boundary (see here).

Answered By - potong

Answer Checked By - Senaida (WPSolving Volunteer)