Awk script

sleeper0110 · 06-05-2009, 10:55 AM

I'm writing a script to convert \r\n to \n in awk.

I have the following:
awk ' { sub(/\r/, "") ; print } '

Which works perfectly.

I started out with:
awk ' { sub(/\r\n/, "\n") ; print } '

Which doesn't work at all. I don't understand why the second version doesn't work. To me it reads "Replace \r\n with \n.

Can someone explaine why the second line I posted doesn't work but the first one does?

Thanks.

ghostdog74 · 06-05-2009, 11:11 AM

just use dos2unix

colucix · 06-05-2009, 11:19 AM

Quote:

Originally Posted by sleeper0110

Can someone explaine why the second line I posted doesn't work but the first one does?

Because \n is the Record Separator in awk and it is not taken in account in $0. Hence the pattern \r\n is not found, whereas \r is the last character in the string. Anyway, I second ghostdog: don't reinvent the wheel and try dos2unix, instead.

sleeper0110 · 06-05-2009, 11:29 AM

colucix: Thanks. Good explanation. Your suggestion of dos2unix doesn't serve my purposes in this case.

ghostdog74: That doesn't answer the question that was asked. And yes, I know what dos2unix is.

ghostdog74 · 06-05-2009, 11:46 AM

if you still want to use awk, another way (not tested)

Code:

awk 'BEGIN{RS="\r\n"; ORS="\n"} {print}' file

H_TeXMeX_H · 06-05-2009, 02:52 PM

You could also use 'tr' ... why does it have to be awk ?

ghostdog74 · 06-05-2009, 06:50 PM

Quote:

Originally Posted by H_TeXMeX_H

why does it have to be awk ?

because one needs to only learn 1 good tool to do the job of many combined?