Hi friends,

I need sample program either in bash or python

Question:You have a hierarchy of directories and somefiles in some of those directories.

e.g /u01/testing/dir1/file1.txt,file2.txt....
/u01/testing/dir2/file.txt ,file1.txt ....
/u01/testing/dir2/file2.txt .......
some of these files contain IP addresses inside the text. An IP address is a string of form x.x.x.x where x is a number from 0 to 255(inclusive).
example file.txt that looks like below.

file1.txt
---------
hello mike 127.0.0.1
this is some example 128.99.107.55
file with some correct and incorrect 128.128.4.11 ip 0.11.1115.78 addresses

file1.txt
---------
hello rocky 127.65.64.1 127.0.64.1 127.0.0.1
example 128.57.107.76 128.57.907.70
file with some correct and incorrect 67.128.4.11 ip 7.7.7.8 addresses

file5.txt
---------
hello rocky 127.69.64.1 127.0.68.1 127.0.1.1
example 128.57.10.76 128.57.9079.70
file with some correct and incorrect 67.128.4.11 ip 7.78.7.8 addresses

task is to find all distinct IP addresses from all the files in the /root/u01/testing/ directory and print them in lexicographical order.

output should be contain only valid ip addresses
-----------------

Good for you.
What have you tried ?. What didn't work as you expected ?. You attempt to solve your problem, we'll help where we can with problems you run in to.

1 members found this post helpful.

From the LQ Rules:
That said,
Lexicographical order, really? What for? If I were your instructor, I would require them to be printed in numerical order.

Now think ... maybe find the file names, extracting only the names ... then "pipe" that into egrep or awk to extract the IP-address-like portions ... then "pipe" that into sort.

Unix/Linux is specifically designed to make it easy to construct solutions like that, even on the command line, where several child processes work together by "piping" their STDOUT into the next's STDIN.

Nowadays, -R is supported by both GNU and *BSD grep, so find is probably unneeded. But the assignment strikes me as a completely contrived task.

Let's say those were some log files. In that case, it'd be reasonable to assume that they contained only valid IP addresses. Then you wouldn't have to validate them. Just matching against (\d{1,3}\.){3}\d{1,3} would be enough. And I definitely would like them to be printed in numerical order, so that spotting contiguous ranges would be easier.

Or those were some configuration files that might contain errors. In that case, a more hands-on task would be to match only against invalid IP addresses, i.e. series of four numbers separated by dots where at least one of the numbers were greater than 255. Of course, this wouldn't catch all the errors, but at least, the most obvious typos. But there's no point in sorting the results then. Rather, I'd like each invalid IP to be accompanied by the file name and the line number it was found at. And perhaps, some lines of surrounding context, too.

And possibly uniq.

(Yes, I know sort can remove duplicates).