My script is:

#!/bin/bash

awk 'BEGIN { substr($0,1,2) == "HI"}' filename >> output_file

I think $0 represents each line in "filename" and i want to match 2 charcter begining position 1.

I expect it to return all matching records into output_file...but it returns me a 0 for each line in "filename".

This works perfectly from command line....

where am i wrong....

Do you want to print all lines matching the condition? You have to simply use an expression, not an action. That is, don't put the brackets (nor the BEGIN section) and for each line matching the condition, the default action (print $0) will be executed.
The same can be accomplished using grep

Can i search for multiple keywords in --

1. ) egrep '^HI|^AZ' filename >> output_file:
Is the above a valid command to return all lines in the file that begin with an "HI" or "AZ".


2. ) In AWK :: awk 'substr($0,1,2) == "HI"' filename >> output_file
Can i look for more than 1 keyword in the same command i.e in just one line... (say AZ in addition to HI)


3. ) Can i inverse the AWK...(like we run grep and egrep with the -v switch)
say 2.) above returns all lines starting with "HI"....
How do i get all lines NOT starting with (HI and AZ)

if you want to use awk, there's no need to use egrep
inverse