sorting and outputing only unique records based on first field

[horse123]

Hi I've got a file with about 20,000 records in the following lines (layout):

XXXXXXXXX||XXXXX|XXXXX

X means alpha-numeric chars

I want everything in XXXXXXXXX||, which is a kind of first(field),to be unique and output to a separate file.

I first did this with "sort file_name | uniq > new_file_name"
I believe this is an incorrect approach.

I then did this with "sort +0, -1, -u, -t "||" file_name > new_file_name"
I am still not convinced.
Is there any better approach? Thanks in advance

 

[PHV]

Try something like this:

sort -t "|" -k 1,1 file_name > new_file_name

Hope This Help
PH.

 

[horse123]

Hi, I've tried using the above command:
sort -t "|" -k 1,1 file_name > new_file_name

It results an output with same lines as the original file.

I wonder ...

but how to do with two pipes ||

 

[horse123]

I solved this problem by adding a -u , so

sort -t "|" -k 1,1 -u file_name > new_file_name

I got an output with the same lines as when I used the following:

"sort +0, -1, -u, -t "|" file_name > new_file_name"