topic Re: replace characters in a file in Operating System - HP-UX

replace characters in a file

Sajeesh O.K — Thu, 06 Jul 2006 04:49:47 GMT

Hi

Anybody know how to replace a charater string with a different string from a file having 1000 lines.The file is like this

aaaa:bbbb:cccc:dddd:eeee:ffff:gggg:

I want to replace cccc with xxxxx in all lines

Thanks
Sajeesh

Re: replace characters in a file

H.Merijn Brand (procura — Thu, 06 Jul 2006 04:54:39 GMT

# perl -pi -e's/cccc/xxxxx/' file

or more safe, if cccc can occur in a word like ddccccee

# perl -pi -e's/\bcccc\b/xxxxx/' file

Enjoy, Have FUN! H.Merijn

Re: replace characters in a file

Arunvijai_4 — Thu, 06 Jul 2006 04:59:15 GMT

HI,

You can use sed.

# sed 's/cccc/xxxxx/g' file

-Arun

Re: replace characters in a file

spex — Thu, 06 Jul 2006 06:11:53 GMT

Sajeesh,

'tr' also works:

tr 'cccc' 'xxxx'
PCS

Re: replace characters in a file

H.Merijn Brand (procura — Thu, 06 Jul 2006 06:14:22 GMT

no, tr works on a character by character basis.
what you wrote here is essentially the same as

tr c x new_file

and will change *every* occurance of 'c' to 'x' in file in every line. That's not what he wants

Enjoy, Have FUN! H.Merijn

Re: replace characters in a file

Jose Mosquera — Thu, 06 Jul 2006 06:43:09 GMT

Hi,

#cat file|sed 's/cccc/xxxxx/g' > newfile

Rgds.

Re: replace characters in a file

spex — Thu, 06 Jul 2006 06:48:03 GMT

Oops. Good catch, procura. My apologies.

Re: replace characters in a file

H.Merijn Brand (procura — Thu, 06 Jul 2006 06:49:04 GMT

Jose, that's the same answer that Arunvijai gave, and even worse, you promote this awful unneeded use of cat

# cat file | process

should be prevented wherever possible. It's just a useless extra load on the CPU, as processes are started that are not needed in a sane unix environment, where

# process < file

does exactly the same, but uses one less process. Cat can be useful if *more* input files create the input stream for process, as input redirection does not support that

# cat file1 file2 | process

Enjoy, Have (efficient) FUN! H.Merijn

Re: replace characters in a file

Peter Nikitka — Thu, 06 Jul 2006 07:54:50 GMT

Hi,

lets play 'sed' a little...
I think your field delimiter is ':' - this fact was not used in the sed-solutions so far.

from=cccc
to=xxxx
sed -e "s/^$from:/$to:/" -e "s/:$from/:$to/g" input >output

But as you see from the previous responses: there is more than one way ...

mfG Peter

Re: replace characters in a file

Jose Mosquera — Thu, 06 Jul 2006 07:58:32 GMT

Please excuse me, especially the partner Pharaoh, and omit this answer.

About the sharp comment of the Mr. Procura, there are not words...

Rgds.

Re: replace characters in a file

H.Merijn Brand (procura — Thu, 06 Jul 2006 08:16:13 GMT

Jose, sorry if I sounded harsh, and it is nothing personal, but somehow the useless use of cat always triggers this reaction with me.
Again, nothing personal, but yet another good oportunity to set things straight!

Enjoy, Have FUN! H.Merijn [ no points here please ]

Re: replace characters in a file

Sajeesh O.K — Fri, 07 Jul 2006 01:56:49 GMT

Hi All

Thanks for your answer. Sorry. I didn't make it clear.

aaa bbb ccc ddd eee fff represent strings.
and delemited with : just like passwd file.
I want to replace the 3rd field with "---"

Kindly provide the solution

Thanks
Sajeesh

Re: replace characters in a file

Junghun — Fri, 07 Jul 2006 02:25:13 GMT

for example the filename is 111 and turn it to file 222
#from=ccc
#to=---
#sed -e "s/$from/$to/" -e "s/$from/$to/g" 111 > 222

--Z.X.

Re: replace characters in a file

H.Merijn Brand (procura — Fri, 07 Jul 2006 02:28:31 GMT

> aaa bbb ccc ddd eee fff represent strings.
> and delemited with : just like passwd file.
> I want to replace the 3rd field with "---"

# perl -pi -aF: -e'$F[2]="---";$_=join":",@F' file

Enjoy, Have FUN! H.Merijn

Re: replace characters in a file

Sajeesh O.K — Fri, 07 Jul 2006 02:37:35 GMT

procura

Perfect. Thanks for the promt reply

Sajeesh

Re: replace characters in a file

Peter Nikitka — Fri, 07 Jul 2006 06:14:33 GMT

Hi,

to complete the thread, here a solution with awk:

awk -F: '$3 == "ccccc" {$3="-----"}
{print}' input >output

Or sed?

sed 's/^$[^:]*:[^:]*:$'$from'$.*$/\1'$to'\2/' input >output

mfG Peter