How to select and print IP address from a set of strings?

It's possible, but not elegant:

echo 'This sentence contains an ip number 1.2.3.4 and port number 50, i want to print the IP address only.' \
| sed 's/.*\([0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\).*/\1/'

[0-9] matches any digit, \{1,3\} means it can be repeated 1 to 3 times. \. matches a dot. The whole IP is captured by the \(...\) parentheses, what comes before and after is matched by .*, i.e. anything repeated zero or more times. The whole matching string (i.e. the whole line) is then replaced by the contents of the first matching group.

You can make it more readable by introducing a variable:

n='[0-9]\{1,3\}'
... | sed "s/.*\($n\.$n\.$n\.$n\).*/\1/"

It prints the whole string if the IP is not found. It also doesn't check for invalid IPs like 256.512.999.666.

This is a grep solution:

echo "$sentence" | grep -oE '[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+'

• -o print only the matching part
• -E switches to extended regex
• the pattern matches every digit ([0-9]) one or more times (+) then a dot (\.) and again digits...

Here another solution with perl:

echo "$sentence" | perl -l -ne '/[0-9]+\.[0-9]+\.[0-9]+\.[0-9]+/ && print $&'

• -l specified the line terminator (newline)
• -n loops trough the input given by echo (could be multiple lines)
• -e code follows
• the regex inside the perl code is very much the same as in the grep solution above

Use this command of grep:

grep -Eo '[0-9.]+ ' file

Or even better:

grep -oP '\d+\.\d+\.\d+\.\d+' file

or

grep -Eo "([0-9]{1,3}[\.]){3}[0-9]{1,3}" file

I use grep:

echo 'This sentence contains an ip number 1.2.3.4 and port number 50, i want to print the IP address only.' | grep -oE '((1?[0-9]?[0-9]|2[0-4][0-9]|25[0-5])\.){3}((1?[0-9]?[0-9]|2[0-4][0-9]|25[0-5]))'

This will print only valid ip adresses, unlike the other answers

I would do like this in python 3 interpreter. It not only grabs the text which are in this 111.111.111.111 format but also it checks for valid or not.

>>> import re
>>> import ipaddress
>>> text = "This sentence contains an ip number 1.2.3.4 and 111.111.111.111 451.976.897.786 port number 50, i want to print the IP address only."
>>> foo = re.findall(r'(?<!\S)(?:\d{1,3}\.){3}\d{1,3}(?!\S)', text)
>>> foo
['1.2.3.4', '111.111.111.111', '451.976.897.786']
>>> for i in foo:
...     try:
...         ipaddress.ip_address(i)
...     except:
...         pass
... 
IPv4Address('1.2.3.4')
IPv4Address('111.111.111.111')

To get the python 3 interpreter, type python3 command on the terminal.

Extension to choroba's answer:

If you don't want new line to be printed and you only want to print IP's:

$ echo -e 'This sentence contains an ip number 1.2.3.4 and port number 50, \ni want to print the IP address only.\n One more IP is 1.24.53.3.' \
| sed -n 's/.*\([0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\.[0-9]\{1,3\}\).*/\1/1p'

Output:

1.2.3.4
1.24.53.3

Explanation:

-n flag is for quiet / silent.
 p print the replaced text

AWK combined with RegExp is very appropriate for processing parts of lines.

Basic idea of the one-liner bellow is to for-loop through a line, and check for presence of four digits and dots, repeated at maximum 4 times; at the same time we can check for a digit repeated 2 to 4 times for port number

awk '{for(i=1;i<=NF;i++) { if ($i~/[[:digit:]\.]{4}/) printf $i; if ( $i~/[[:digit:]]{2,4}/) printf ":"$i  }}'

Sample run

$ echo "This sentence contains an ip number 1.2.3.4 and port number 50, i want to print the IP address only." |  awk '{for(i=1;i<=NF;i++) { if ($i~/[[:digit:]\.]{4}/) printf $i; if ( $i~/[[:digit:]]{2,4}/) printf ":"$i  }}'

1.2.3.4:50,

Your sentence contains 50 and, together without separation, hence printed together, but with gsub(/[[:punct:]]/,"") that can be removed if desired.