Regular expressions to match IP addresses

[SagaciousKJB]

Hello, I'm having some problems using grep to match IP addresses. At first I was attempting to learn how to do it on my own, but I gave up and Google'd some solutions. The problem is, none of them work either. The first one I tried is this beast:

\b(?:(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\.){3}(?:25[0-5]|2[0-4][0-9]|[01]?[0-9][0-9]?)\b

This didn't work because of the "(" characters, yet when I tried to escape them, it didn't work either.

grep -e [0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3}\.[0-9]{1,3} ./ips

This one is less accurate, but I figured it would work, no dice

grep: [0-9]1.[0-9]1.[0-9]1.[0-9]3: No such file or directory
grep: [0-9]1.[0-9]1.[0-9]1.[0-9]3: No such file or directory
grep: [0-9]1.[0-9]1.[0-9]3.[0-9]1: No such file or directory
grep: [0-9]1.[0-9]1.[0-9]3.[0-9]3: No such file or directory
grep: [0-9]1.[0-9]3.[0-9]1.[0-9]1: No such file or directory
grep: [0-9]1.[0-9]3.[0-9]1.[0-9]3: No such file or directory
grep: [0-9]1.[0-9]3.[0-9]3.[0-9]1: No such file or directory
grep: [0-9]1.[0-9]3.[0-9]3.[0-9]3: No such file or directory
grep: [0-9]3.[0-9]1.[0-9]1.[0-9]1: No such file or directory
grep: [0-9]3.[0-9]1.[0-9]1.[0-9]3: No such file or directory
grep: [0-9]3.[0-9]1.[0-9]3.[0-9]1: No such file or directory
grep: [0-9]3.[0-9]1.[0-9]3.[0-9]3: No such file or directory
grep: [0-9]3.[0-9]3.[0-9]1.[0-9]1: No such file or directory
grep: [0-9]3.[0-9]3.[0-9]1.[0-9]3: No such file or directory
grep: [0-9]3.[0-9]3.[0-9]3.[0-9]1: No such file or directory
grep: [0-9]3.[0-9]3.[0-9]3.[0-9]3: No such file or directory

I am not a genius when it comes to grep, so there is probably something very simple I am missing. Can anyone help please?

[anomie]

Try single-quoting (') your regexps.

[Cabhan]

First off, those appear to be different sorts of regular expressions. The first appears to be Perl syntax, while the second is a mix of POSIX and Extended POSIX.

Your problem with the second is that Bash is taking the '{'s and '}'s away. Try escaping them on the commandline.

For instance, using Perl regular expressions, this is the easiest way:

\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}

Using Extended POSIX regular expressions (enabled by using 'egrep' instead of 'grep'), you could do:

[:digit:]{1,3}\.[:digit:]{1,3}\.[:digit:]{1,3}\.[:digit:]{1,3}

Having said all of this, there are a few problems. For instance, both of these would accept 300.1.1.1, which is obviously not a valid regular IP address. Your best bet would be something like:

IP=127.0.0.1

valid=1
count=0

for field in $(echo "$IP" | cut -d. --output-delimiter=$'\012' -f1-); do
    count=$[count+1]
    if [ "$field" -eq $[field+0] -a "$field" -le 256 -a "$field" -ge 0 ]; then
        : yay
    else
        valid=0
    fi
done

if [ "$valid" -a "$count" -eq 4 ]; then
    echo "Valid IP address!"
else
    echo "Invalid IP address!"
fi

This uses cut to break the string into its '.'-separated parts, then employs the loop to ensure that each part is between 0 and 255 inclusive, and that there are exactly 4 parts.

Make sense?

[SagaciousKJB]

First off, those appear to be different sorts of regular expressions. The first appears to be Perl syntax, while the second is a mix of POSIX and Extended POSIX.

Your problem with the second is that Bash is taking the '{'s and '}'s away. Try escaping them on the commandline.

For instance, using Perl regular expressions, this is the easiest way:

\d{1,3}\.\d{1,3}\.\d{1,3}\.\d{1,3}

Using Extended POSIX regular expressions (enabled by using 'egrep' instead of 'grep'), you could do:

[:digit:]{1,3}\.[:digit:]{1,3}\.[:digit:]{1,3}\.[:digit:]{1,3}

Having said all of this, there are a few problems. For instance, both of these would accept 300.1.1.1, which is obviously not a valid regular IP address. Your best bet would be something like:

IP=127.0.0.1

valid=1
count=0

for field in $(echo "$IP" | cut -d. --output-delimiter=$'\012' -f1-); do
    count=$[count+1]
    if [ "$field" -eq $[field+0] -a "$field" -le 256 -a "$field" -ge 0 ]; then
        : yay
    else
        valid=0
    fi
done

if [ "$valid" -a "$count" -eq 4 ]; then
    echo "Valid IP address!"
else
    echo "Invalid IP address!"
fi

This uses cut to break the string into its '.'-separated parts, then employs the loop to ensure that each part is between 0 and 255 inclusive, and that there are exactly 4 parts.

Make sense?

Well, I don't think I quite understand what that scripts function is, but matching invalid addresses like 300.1.1.1 or the like is okay for my purposes. I simply need to make a script that grabs each individual IP address from a log file. I didn't realize egrep was needed for the [::] type regular expressions, that's very useful information. however I still get an output like this

grep: [:digit:]1.[:digit:]1.[:digit:]1.[:digit:]3: No such file or directory
grep: [:digit:]1.[:digit:]1.[:digit:]3.[:digit:]1: No such file or directory
grep: [:digit:]1.[:digit:]1.[:digit:]3.[:digit:]3: No such file or directory
grep: [:digit:]1.[:digit:]3.[:digit:]1.[:digit:]1: No such file or directory
grep: [:digit:]1.[:digit:]3.[:digit:]1.[:digit:]3: No such file or directory
grep: [:digit:]1.[:digit:]3.[:digit:]3.[:digit:]1: No such file or directory
grep: [:digit:]1.[:digit:]3.[:digit:]3.[:digit:]3: No such file or directory
grep: [:digit:]3.[:digit:]1.[:digit:]1.[:digit:]1: No such file or directory
grep: [:digit:]3.[:digit:]1.[:digit:]1.[:digit:]3: No such file or directory
grep: [:digit:]3.[:digit:]1.[:digit:]3.[:digit:]1: No such file or directory
grep: [:digit:]3.[:digit:]1.[:digit:]3.[:digit:]3: No such file or directory
grep: [:digit:]3.[:digit:]3.[:digit:]1.[:digit:]1: No such file or directory
grep: [:digit:]3.[:digit:]3.[:digit:]1.[:digit:]3: No such file or directory
grep: [:digit:]3.[:digit:]3.[:digit:]3.[:digit:]1: No such file or directory
grep: [:digit:]3.[:digit:]3.[:digit:]3.[:digit:]3: No such file or directory

Escaping the "{}" characters produced no errors however it did not match any of the IPs in the file. I'm running it on a test file with only two example IPs, the conent of which follows

68.112.34.23
235.37.123.244

[drl]

Hi.

Adding brackets around the digit expressions like this:

#!/bin/sh -

# @(#) s1       Demonstrate egrep.

cat >data1 <<'EOF'
garbage with no digits
68.112.34.23
235.37.123.244
frog
EOF

echo
echo " Input file:"
cat data1

echo
echo " Results from grep, expecting 2 lines:"
egrep '[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}' data1

exit 0

produces:

% ./s1

 Input file:
garbage with no digits
68.112.34.23
235.37.123.244
frog

 Results from grep, expecting 2 lines:
68.112.34.23
235.37.123.244

See man egrep for details ... cheers, drl

[SagaciousKJB]

Hi.

Adding brackets around the digit expressions like this:

#!/bin/sh -

# @(#) s1       Demonstrate egrep.

cat >data1 <<'EOF'
garbage with no digits
68.112.34.23
235.37.123.244
frog
EOF

echo
echo " Input file:"
cat data1

echo
echo " Results from grep, expecting 2 lines:"
egrep '[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}\.[[:digit:]]{1,3}' data1

exit 0

produces:

% ./s1

 Input file:
garbage with no digits
68.112.34.23
235.37.123.244
frog

 Results from grep, expecting 2 lines:
68.112.34.23
235.37.123.244

See man egrep for details ... cheers, drl

Youre modifcation worked prefectly, as you already know. Are there any more indepth guides in the subtler bits of using grep like this? On my machine, "man egrep" opens up the grep man page, which I've read quite a bit.

Anyway, I think I'll search for that myself, thanks for the help.