bash script counting lines

With awk:

$ var=$(awk '$0=="Firmware state: Rebuild"{print --NR; exit}' file)
$ echo "$var"
14

The awk script prints the input record number minus 1 when the current line equals Firmware state: Rebuild and exits the script. The output of the command substitution $(...) is then assigned to variable var.

This should do what you want:

$ var=$(( $(grep -m1 -n 'Firmware state: Rebuild' file | cut -d: -f1) -1 ))
$ echo $var
14      

Explanation

• grep -m1 -n 'Firmware state: Rebuild': this will search for the first (-m 1) line matching Firmware state: Rebuild, and print that line including its line number (-n):

  $ grep -m1 -n 'Firmware state: Rebuild' file 
  15:Firmware state: Rebuild
  

• cut -d: -f1: the output of the grep above is then piped through this cut which takes : as the field delimiter and prints the first field, the line number. *var=$(( $(command) -1 )): The $(( )) construct lets us perform mathematical operations. For example:

  $ echo $((10-2))
  8
  

  The $(command) construct lets us use the output of a command as though it were a variable. So, $(( $(command) -1 )) will print the result of subtracting 1 from the output of command. Since command here is the grep | cut above, whose output is the line number (15), this will save the line number minus one as $var.

Note that the solution above will also match a line like Firmware state: Rebuild again or anything else that contains Firmware state: Rebuild as a substring. If you need to ensure it only matches when the entire line contains nothing but Firmware state: Rebuild, use this instead:

var=$(( $(grep -xm1 -n 'Firmware state: Rebuild' file | cut -d: -f1) -1 ))

var=$(grep -n "Firmware state: Rebuild" filename| awk -F ":" '{print $1-1}')

This is a variation of terdon's answer, using awk instead of cut and shell arithmetic to calculate the zero-based line number.

sed -n '/Rebuild$/!d;=' file

Or using antipattern:

sed -n '/,/d;=' file

15

Prints the line number with the given pattern. In this case, additional mathematical actions are required.

sed -n '/Rebuild$/Q;=' file

Or using antipattern:

sed -n '/,/!Q;=' file

1
2
3
4
5
6
7
8
9
10
11
12
13
14

Prints line numbers before match the pattern. Here it is possible to use the q command instead of the Q command.
It is necessary to separate the last value for example via the | tail -1 command. Or in a more sophisticated way:

: $(sed -n '/,/!q;=' file)
echo $_
14