Question

In the header of my bash script, I have to ask for the user name and his IP address.

I need to validate entries in this way:

• The user name have to be stored in a name variable, having his space replaced by underscores. For sample if user hit "Albert Smith", I have to store "Albert_Smith" into my variable name.
• For the ip address, I need to know if the ip address is consistent with the format.
  If the format is not good, I must report the error and prompt the user to enter a good ip address to continue the operation.

I know that i need to use “read”, but i don’t know to realize my short script (replace and verify).
Thank in advance 🙂

Answer 1

I suppose your entries will be line based (ie. the name will be the first line and the IP the second). You can do this:

#!/bin/bash

read name    
while [[ ! "$name" =~ '[A-Za-z ]' ]]; do
    read -p "Wrong name format. Re-enter: " name
done

name="${name// /_}"

read ip
while [[ ! "$ip" =~ '^((25[0-5]|2[0-4][0-9]|[01][0-9][0-9]|[0-9]{1,2})[.]){3}(25[0-5]|2[0-4][0-9]|[01][0-9][0-9]|[0-9]{1,2})$' ]]; do
    read -p "Not an IP. Re-enter: " ip
done

The IP regex is rather complex, but it can be decomposed as:

1. After the start of the string, there must be three groups of “octet” followed by periods, then followed by another “octet” and the end of the string
2. Octet is either
   1. 250-255,
   2. 200-249
   3. 100-199 (including 000-099, with the explicit 0s before the number)
   4. One or two digits (ie. 0-99)

The loop forms will negate ! the result of the grep, so it will continue looping while the grep command fails. We use printf in order to pass the variable into grep. The -E flag on the second grep allows the use of extended regular expressions, which include the | OR operator, () grouping and {} repetitions.

Hope this helps a little =)