grep

[malladisk]

Hi, How can I force grep to return unique results? I mean, if the output has duplicate lines, how do I get grep to print only one line for all these duplicate lines?
Thanks,
Sashi

 

[mrmac228]

You could pipe the output of grep to uniq....

"your grep command"| uniq

I love deadlines. I like the whooshing sound they make as they fly by - Douglas Adams

 

[stefanwagner]

... but I'm quit sure, you have to sort the output before, to make the work not too complicated for uniq

"ygc" | sort | uniq

 

[CaKiwi]

Or pipe it to an awk program

grep ... | awk '{if (!a[$0]){print;a[$0]=1}}'

CaKiwi

 

[stefanwagner]

That will not match

foo
bar
foo

will it?

 

[CaKiwi]

stefanwagner,

If you are talking about my solution, I believe it will match your example. Did you try it?

CaKiwi

 

[aigles]

Instead of 'grep ere | awk' you can do :
[tt]
awk '/ere/ {if (!a[$0]){print;a[$0]=1}}'
[/tt]

Jean Pierre.

 

[stefanwagner]

CaKiWi: No - I didn't test.
Now I tested - and - surprise: it works.

But of course my solution is much more elegant and shorter - isn't it?

 

[CaKiwi]

stefanwagner,

If sorting the data is acceptable, your solution is good, although

grep ... | sort -u

is slightly simpler. If the data must be kept in its original order, then a solution like mine is necessary. Of course, aigles' solution is better.



CaKiwi

 

[stefanwagner]

yes. and maladisk vanished silently...