Removing first occurrence of a string in a file with sed on mac

Question

From this I found the command and it works without any command substitution. But if I try to substitute a variable in the command, I get the following problems.

Note: I am using sed command(BSD OSX) / MacOS

To reproduce just create a file.txt with this:

SOME_TEXT
CHANGE_THIS
CHANGE_THIS
CHANGE_THIS

Add TMP variable in you terminal:

TMP=export PATH=/Users/Name/with pace/and_variables:$PATH

Now with this command:

sed -i '' "1h; 1 $ { x; s/CHANGE_THIS/$TMP/; p; }" file.txt

I get this error:

sed: 1: "1h; 1 $ { x; s/CHANGE_T ...": invalid command code $

In a privious answer I got this("," added to "1,$":

sed -i '' "1h; 1,$ { x; s/CHANGE_THIS/$TMP/; p; }" file.txt

expected output file.txt

SOME_TEXT
export PATH=/Users/Name/with pace/and_variables:$PATH
CHANGE_THIS
CHANGE_THIS

I get this:

SOME_TEXT
SOME_TEXT
SOME_TEXT
SOME_TEXT

Why use a variable at all, when you can just paste the text directly into the sed command? If you really want to do it like this, try `"$TMP"`. — max, Nov 14 '18 at 13:36

score 1 · Answer 1 · answered Nov 14 '18 at 13:41

1

Use:

sed -i '0,/.*CHANGE_THIS.*/s//my changed line/' file

answered Nov 14 '18 at 13:41

Incrivel Monstro Verde

888
6
11

6

MacOS `sed` does not support the GNU `sed` extension `0,` – tripleee Nov 14 '18 at 13:45

RavinderSingh13 · Accepted Answer · 2018-11-14T13:45:04.343

Could you please try following if you are ok with awk.

##Creating shell variable named TM here.
TMP="export PATH=/Users/Name/with pace/and_variables:$PATH"
awk -v tmp="$TMP" '/CHANGE_THIS/ && ++flag==1{$0=tmp} 1' Input_file

In case you want to save output into Input_file itself then append > temp_file && mv temp_file Input_file in above code too.

Adding explanation for above code too here.

awk -v tmp="$TMP" '              ##Creating an awk variable named tmp whose value is bash variable TMP value.
/CHANGE_THIS/ && ++flag==1{      ##Checking condition if a line is having string CHANGE_THIS and variable flag value is 1 then do following.
  $0=tmp                         ##Setting current line value to value of tmp awk variable here.
}                                ##Closing the block for condition here.
1                                ##By mentioning 1 printing edited/non-edited line here.
' Input_file                     ##Mentioning Input_file name here.

Thanks it works - just printing to console - I now do "awk ..." >> newfile.txt — Chris G., Nov 14 '18 at 13:49

Removing first occurrence of a string in a file with sed on mac

2 Answers2