replace a line with another

[watergrinder]

Hello,

Could a line be replaced by another line in file?

Ex:

11111
22222
33333
44444


3333 should be replaced by 5555. Similarily, 3 3 3 3 should be replaced .... i.e 3rd line should be replaced by 5555. Is there something that can use NR (line number).

Thanks in advance.

 

[PHV]

Something like this ?
awk 'NR==3{gsub(/3/,"5",$0)}1' /path/to/input

Hope This Help, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884

 

[watergrinder]

Thanks! Is there a way to replace line 3 with whatever we want?

 

[PHV]

Try something like this:
nawk -v new3="New third line" '{print (NR==3 ? new3 : $0)' /path/to/input

Hope This Help, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884

 

[watergrinder]

I am using HP-UX. Bourne shell. It says nawk not found. I tried putting it in a script using csh & ksh. It still says the same.

 

[PHV]

And what about this ?
awk -v new3="New third line" '{print (NR==3 ? new3 : $0)' /path/to/input

Hope This Help, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884

 

[watergrinder]

Is there a way to use shell variable in awk script?

#!/usr/bin/sh
file = r
Number=3
echo "`cat $filename| awk -v replace="replace" '{print NR==$Number  ? replace : $0)}'`" > $filename

The above $Number as NR, won't work. Any ideas?

Thanks in advance.

 

[PHV]

Try this:
Number=3
echo "`awk -v replace='replace' '{print (NR=='$Number' ? replace : $0)}' $filename`" >NewFile
Or this:
echo "`awk -v replace='replace' -v Number=3 '{print (NR==Number ? replace : $0)}' $filename`" >NewFile

Hope This Help, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884

 

[watergrinder]

Thanks!

This works with static text in replace' '
echo "`awk -v replace='replace' '{print (NR=='$Number' ? replace : $0)}' $CountFile`" > $CountFile

I want to replace that with $* ... shell parameters.

It gives an error with $*.

 

[PHV]

echo "`awk -v replace='replace' '{print (NR=='$Number' ? replace : $0)}' $CountFile`" > $CountFile
I don't think it's safe to have the same $CountFile file as stdin for awk and stdout for echo: it may be erased by the shell before awk have any chance to read it.
It gives an error with $*
Which error ? where ? can you please post the code ?

Hope This Help, PH.
Want to get great answers to your Tek-Tips questions? Have a look at FAQ219-2884