Question

I have a text file which goes like this :

      125
      126
      127    {
      566
      567
      568
      569       # blah blah
      570    {  #blah blah
      700
      701    {

The numbers are left aligned and the pattern is always the same in the sense increasing and a curly braces at the end .I need to catch just the starting number .The braces are always found and limited to the sequence end .The start of the file is as shown starting with ‘125’.

In short I need :

      125
      566
      700

What I have come up with :

      grep -A1 '{' | grep -v '{' | grep -oE '(^[0-9]+?)'

but this omits ‘125’ but I overcame by appending a newline at the head and inserting a { .

I hope to reduce this into a single regex.

Suggestions and better algorithms are welcome

Answer 1

awk 'BEGIN {p=1} p==1 {print $1;p=0} $0~/{/ {p=1}'

Output:
125
566
700

Given the file format above, you could use awk and a variable/flag to keep track on when you find an opening {