Scripting in Unix - Need Help Starting to Write Script that track Logins

circuit_girl · 03-26-2009, 09:03 PM

I want to see who owns more files me or root? I do not want to include directories in my search. I want it to output two lines...root (# of files) myusername (# of Files). What command tells who owns a file? and where is a good place to start?

I am thinking find \ -name root ... but I do not know how to start from the root in the file tree and I do not know how to count them.

Thank you

circuit_girl · 03-26-2009, 09:37 PM

I want to see who owns more files me or root? I do not want to include directories in my search. I want it to output two lines...root (# of files) myusername (# of Files). What command tells who owns a file? and where is a good place to start?

I am trying unix command: find -user but I do not know how to count them and do both my files and root at the same time. I also want to do others can I check them If I am not the owner?

Thank you

Sergei Steshenko · 03-26-2009, 11:42 PM

Quote:

Originally Posted by circuit_girl

I want to see who owns more files me or root? I do not want to include directories in my search. I want it to output two lines...root (# of files) myusername (# of Files). What command tells who owns a file? and where is a good place to start?

I am trying unix command: find -user but I do not know how to count them and do both my files and root at the same time. I also want to do others can I check them If I am not the owner?

Thank you

Why ? I.e. what bad is going to happen if you first find files owned by root and then by a non-root user ?

circuit_girl · 03-27-2009, 12:48 AM

TB0ne
thank you so much it has lead me to many other things I wanted to do. Thank you
Circuit_Girl

circuit_girl · 03-27-2009, 01:00 AM

Code:

$root_str = `find / -user root | wc -l`;

I am trying to experiment I am doing this but I want to leave out directories in the count and block devices...I do not want the count to include /dev and /proc. I have been experimenting with -prune but can not get it to work.

TB0ne · 03-29-2009, 11:56 AM

Quote:

Originally Posted by circuit_girl

Code:

$root_str = `find / -user root | wc -l`;

I am trying to experiment I am doing this but I want to leave out directories in the count and block devices...I do not want the count to include /dev and /proc. I have been experimenting with -prune but can not get it to work.

Try this:

Code:

$root_str = `find / -user root -type f | grep -v "/dev" | grep -v "/proc" | wc -l`;

Very dirty, with the double-greps. Check the MAN pages on the find command, it will explain what the -type option does. Also pay particular attention to the -iregex flag. Regular expressions are VERY handy to know, and are useful not only in Perl, but alot of 'regular' Linux commands. The regex can be used to filter out the things you don't want to see....

Sergei Steshenko · 03-29-2009, 01:04 PM

Quote:

Originally Posted by TB0ne

Try this:

Code:

$root_str = `find / -user root -type f | grep -v "/dev" | grep -v "/proc" | wc -l`;

Very dirty, with the double-greps. Check the MAN pages on the find command, it will explain what the -type option does. Also pay particular attention to the -iregex flag. Regular expressions are VERY handy to know, and are useful not only in Perl, but alot of 'regular' Linux commands. The regex can be used to filter out the things you don't want to see....

Not precise enough - think of

Code:

foo/devwhatever
bar/procwhatever

.

TB0ne · 03-29-2009, 01:07 PM

Quote:

Originally Posted by Sergei Steshenko

Not precise enough - think of

Code:

foo/devwhatever
bar/procwhatever

.

Right...which is why I said "Very dirty", and to look up how to use REGEX'es, which is the best way to do it.

choogendyk · 03-29-2009, 02:24 PM

So,

... | grep -v '^/dev/' | ...

will only identify "/dev/" at the beginning of a line.

TB0ne · 03-29-2009, 05:24 PM

Quote:

Originally Posted by choogendyk

So,

... | grep -v '^/dev/' | ...

will only identify "/dev/" at the beginning of a line.

Yep...but you spoiled the hours of fun "regex" learning, though.

circuit_girl · 04-01-2009, 09:55 PM

I am working on a new script that compares two files. The two files are list of users from two different groups. The scripts should look for common userids and print out one per line the userids common in the two files as well as the user numbers for those id's in each list.

The files will be in this format below...

file1:

Code:

jdoe:x:102:1001:John Doe:/home/jdoe:/usr/bin/tcsh
apache:x:103:60::/usr/local/apache2:
cmadeup:x:109:1001:Chris Madeup:/home/cmadeup:/usr/bin/bash

file2:

Code:

jdoe:x:107:1001:John Doe:/home/jdoe:/usr/bin/tcsh
apache:x:210:60::/usr/local/apache2:
asomthing:x:115:1001:Amy Somthing:/home/asomething:/usr/bin/bash

The output should find user names that are similar between the two files and print them in an output to the screen. Output two other columns (machine number of the first file) (machine number of second file). So the script would find jdoe and apache similar and print the 3rd field in each file.
Something like....

Code:

Userid Machine1 Machine2
jdoe   102      107
apache 103      210

Any help I would greatly appreciate..Super stumped.
Thank you

ghostdog74 · 04-01-2009, 10:18 PM

you have some knowledge in Perl, why don't you try doing that with Perl. then if you have problems still, post here.

TB0ne · 04-01-2009, 10:22 PM

Quote:

Originally Posted by circuit_girl

I am working on a new script that compares two files. The two files are list of users from two different groups. The scripts should look for common userids and print out one per line the userids common in the two files as well as the user numbers for those id's in each list.

Any help I would greatly appreciate..Super stumped.
Thank you

Sticking with Perl, you need to check out the Array::Compare and List::MoreUtils:

airwise modules from CPAN. Map will also work, and will probably prove to be much faster as well, although with smaller files, the different will be negligible.

Some examples:

Code:

use List::MoreUtils q=pairwise=;
my @a = ( 1, 2, 4, 5, 3 );
my @b = ( 1, 2, 4, 2, 3 );
my @c = pairwise { $a == $b ? 1 : 0 } @a, @b;
print "@c";

This assumes the input is two arrays, "1,2,4,5,3" and "1,2,4,2,3". If they match, print a 1. If they don't, print a 0. Output will be "1 1 1 0 1" Define your first file as array @a, second as @b. A quick bit of research will tell you how to get a file into an array..don't want to spoil all the fun.

And if you look on CPAN for that module, you'll fine very complete notes on it, and how it's chewing things up on the @c line.

The built-in map function works too:

Code:

my @c = map { $a[$_] == $b[$_] ? 1 : 0 } 0 .. $#a;

If you use map, you can omit the "use List..." line. From the notes on map:

Quote:

Evaluates the BLOCK or EXPR for each element of LIST (locally setting $_ to each element) and returns the list value composed of the results of each such evaluation. Evaluates BLOCK or EXPR in a list context, so each element of LIST may produce zero, one, or more elements in the returned value

Enjoy.

circuit_girl · 04-02-2009, 09:47 PM

I do not understand the examples listed. I am trying to find the users that are common in two files (these files are the password files) with a machine number that is the third field on a specific line. I only want to print the users that are in both list.

TB0ne · 04-03-2009, 09:36 AM

Quote:

Originally Posted by circuit_girl

I do not understand the examples listed. I am trying to find the users that are common in two files (these files are the password files) with a machine number that is the third field on a specific line. I only want to print the users that are in both list.

Right. The example above takes two arrays, and compares them, which is what you want to do...print a 1 if the array elements are the same, a 0 if not. That's easily modified to print whatever, if whatever element of the array, matches an element of the other array. Check out the modules I mentioned...they're worth playing with.

However, you COULD do this with shell commands too:

Code:

$ sort -k 1,1n file1 > file1.sorted
$ sort -k 1,1n file2 > file2.sorted
$ join -j 1 file1.sorted file2.sorted > same.names

Check the man pages for sort and join.