[SOLVED] remove duplicated stringd in a line...

masavini · 05-15-2012, 04:58 PM

hi,
i have this code:

Code:

$ echo "bubu gaga bubu dede" | awk '{ while(++i<=NF) printf (!a[$i]++) ? $i FS : ""; i=split("",a); print "" }'
bubu gaga dede

i need to modify the awk command so that "-" is considered a field separator (like " ").

i.e.:

Code:

$ echo "bubu gaga bubu-dede" | awk ....
bubu gaga dede

can you help me?
many thanks...

colucix · 05-15-2012, 05:24 PM

Code:

awk 'BEGIN{FS = "[ -]"}{ while(++i<=NF) printf (!a[$i]++) ? $i " " : ""; i=split("",a); print "" }'

masavini · 05-15-2012, 05:30 PM

many thanks colucix! it works great...
e anche il fatto che siamo entrambi di bologna credo sia una discreta coincidenza!

colucix · 05-15-2012, 05:38 PM

He he! Bologna forever!

masavini · 05-15-2012, 05:43 PM

btw,
if you don't mind, would you take a look at this thread, as well?
http://www.linuxquestions.org/questi...ations-945160/

if you could help me, you'd surely increase your bonus to get a new battery for your laptop: removed

colucix · 05-15-2012, 06:52 PM

Quote:

Originally Posted by masavini

btw,
if you don't mind, would you take a look at this thread, as well?
http://www.linuxquestions.org/questi...ations-945160/

if you could help me, you'd surely increase your bonus to get a new battery for your laptop:

Are you trying to corrupt a LQ mod?!

Jokes apart, please remove the commercial link (it could denote advertising). Thank you. Ciao.

masavini · 05-15-2012, 07:00 PM

ops...

i guess corruption is when you try to let someone do something he shouldn't... but i can't see why you shouldn't help me with that problem...

colucix · 05-16-2012, 05:56 AM

Quote:

Originally Posted by masavini

ops...

i guess corruption is when you try to let someone do something he shouldn't... but i can't see why you shouldn't help me with that problem...

I was joking! Actually I tried to help, but it was late in the night and I wasn't able to write down a (working) code. However, Nominal Animal solved the issue in my behalf!

grail · 05-16-2012, 07:01 AM

Here is a slight revision if you like:

Code:

echo "bubu gaga bubu-dede" | awk '{ORS=RT}!_[$0]++' RS="[ -\n]"

masavini · 05-16-2012, 07:32 AM

mmm...

Code:

$ echo "bubu gaga bubu-dede" | awk '{ORS=RT}!_[$0]++' RS="[ -\n]"
awk: cmd. line:1: fatal: Invalid range end: /[ -
]/

colucix · 05-16-2012, 07:52 AM

It depends on the shell you're using. It works in bash. Anyway, since the problem happens outside the awk one-liner, put the RS definition inside it and the trick is done:

Code:

echo "bubu gaga bubu-dede" | awk 'BEGIN{RS="[ -\n]"}{ORS=RT}!_[$0]++'

or eventually try single quotes in place of double quotes:

Code:

echo "bubu gaga bubu-dede" | awk '{ORS=RT}!_[$0]++' RS='[ -\n]'

masavini · 05-16-2012, 08:04 AM

mmm...

Code:

$ echo "bubu gaga bubu-dede" | awk '{ORS=RT}!_[$0]++' RS='[ -\n]'
awk: cmd. line:1: fatal: Invalid range end: /[ -
]/
$ echo "bubu gaga bubu-dede" | awk 'BEGIN{RS="[ -\n]"}{ORS=RT}!_[$0]++'
awk: fatal: Invalid range end: /[ -
]/

colucix · 05-16-2012, 08:23 AM

Which version of awk are you running?

Code:

awk --version

masavini · 05-16-2012, 08:25 AM

Code:

$ awk --version
GNU Awk 3.1.7

colucix · 05-16-2012, 08:36 AM

The same for me. At this point it might depend on the locale settings and the ambiguous interpretation of the dash - inside the character list. Your best bet is to move it at the beginning of the list:

Code:

echo "bubu gaga bubu-dede" | awk '{ORS=RT}!_[$0]++' RS='[-\n ]'