Looping through a file but ignoring the first two lines in Bash

Question

I am using the following command: diff -u file1 file2 > diff.txt and parsing the output. I have a loop that can iterate every line in the file. I am using a loop solution from https://stackoverflow.com/a/1521498/17386696.

while IFS="" read -r p || [ -n "$p" ] ; do
      # first character '+'
      if [[ ${p:0:1} = "+" ]] ; then
        # do something
      # first character '-'
      elif [[ ${p:0:1} = "-" ]] ; then
        # do something
      fi
    done < diff.txt

My current issue is that the first two lines of the file look like:

--- file1  2022-03-19 12:28:10.119916406 -0400
+++ file2  2022-03-19 12:28:11.171926970 -0400

I know I could create another conditional statement for the +++ and --- lines if all else fails. I was curious if there was a way to adjust the loop to start at line three to avoid the triple symbols.

try `done < <(tail +3 diff.txt)` or `done < <(tail -n +3 diff.txt)` — markp-fuso, Mar 19 '22 at 20:17
also `done < <(sed '1,2d' diff.txt)` and `done < <(awk 'NR>2' diff.txt)` — markp-fuso, Mar 19 '22 at 20:50

Cyrus · Accepted Answer · 2022-03-19T22:29:08.617

1

Given the small number of lines (2) to be ignored, I suggest the following:

{
  read; read
  while IFS="" read -r p || [ -n "$p" ] ; do
    ...
  done
} < diff.txt

read; read reads the first two lines from stdin (diff.txt).

edited Mar 19 '22 at 22:29

answered Mar 19 '22 at 22:12

Cyrus

84,225
14
89
153

It works! Wondering if you could explain what the curly braces are doing? I do not fully understand them. – JohnnAustin12108 Mar 19 '22 at 23:32
`{ ...; }` is a compound command. In it all commands can read from same stdin (`diff.txt`). – Cyrus Mar 20 '22 at 00:02

Looping through a file but ignoring the first two lines in Bash

1 Answers1