Hi,

I have file which contain data like -
1311629642,abc
1311629642,def
1311629906,ghi
1311630077,jkl
1311631671,mno
1311632245,pqr

I need o/p as -
2 1311629642 abc
def
1 1311629906,ghi
1 1311630077,jkl
1 1311631671,mno
1 1311632245,pqr

I already tried with - cat filename | cut -d "," -f1,2 | sort | uniq -c
but no luck.... could someone let me know the correct command ?

Thanks.

One of the correct methods would be to use the language awk.

Another way, if you want to eliminate the whole line, would be to take advantage of sort utility's full capabilities.

See "man sort" and "man uniq".

However, if you need to retain the line with "def" or someting else more fancy then awk or perl is needed.

tried with sort -t, -k1,1 /path/to/some/filename | uniq -c -w 10 command but gives o/p as -
2 1311629642 abc
1 1311629906 ghi
1 1311630077 jkl
1 1311631671 mno
1 1311632245 pqr

def missing in this... it should be print as

2 1311629642 abc
def

use Perl.

Output:

1 members found this post helpful.

thanks for this

but will it be possible by any shell script / linux command instead of perl

perl actually is a "linux" "command" You will find it on all systems, just like awk. You will even find it on the various BSD operating systems, such as OS X.

Sorry, but I am not familier to perl and need this in shell scripts / any linux command only

Now, I tried by creating 2 files such as -
cat file name | cut -d "," -f1 | uniq -c > test1
cat file name | cut -d "," -f2 > test2

So now, I have 2 files with o/p as -
test1
2 1311629642
1 1311629906
1 1311630077
1 1311631671
1 1311632245

test2
abc
def
ghi
jkl
mno
pqr

now just want to paste the 2 files (test1 and test2) in such way that it should print last o/p as -
2 1311629642 abc def
1 1311629906 ghi
1 1311630077 jkl
1 1311631671 mno
1 1311632245 pqr

If there were a common field between the two files you could have used join. But going over the data as you have originally presented it you can use Linux's awk or perl commands.

better to leave everything... can you help me to get desired o/p using awk / sort / any shell script ?

Well, if you aren't going to use Linux's perl command, I'd still recommend that you put it on your todo list for the future. It is going to come up again and again in similar circumstances.

Back to awk, my approach would be to set the input Field Separator to a comma and then iterate through the data file. As each line is read, make an array element with the first field as a key and increment that element. Also make an element in a second array with the first field as the key, too, and append field two to its value using a comma or other separator. Then the output occurs in the END {} block by looping through the keys in the first array and printing the relevant key and element value. Then use split() to divide the corresponding element in the second array and loop through it, printing each element.

The printf() function is useful in all that.

It may not be the best approach, but it is simple to implement.

Add it in a piece at a time. We can walk you through the steps in awk or perl, but do show us what you have started and which guides you are working from.

Is that original input file sorted ?

You don't seem to get the fact that Perl is every bit a standard UNIX/Linux command as awk/sed/etc. Plus, it's much faster, more efficient and more powerful.

I cleaned up the code a bit to add

and

Plus some other stuff.

This is always good Perl programming practice:

That helps catch a lot of bugs, especially in very large programs.