awk - searching/replacing control characters

[Lemaitre]

People,

I am trying to delete a load of control characters from a file that has acquired a load of new text via conversion problems. I.e. the pipe delimitted file looks like :

aaaa|bbbb|cccc^M
dddd|eeee|ffff^M
gggg|hhhh|iiii^M

what I am trying to do is convert it to

aaaa|bbbb|cccc|eor
dddd|eeee|ffff|eor
gggg|hhhh|iiii|eor

I have tried the following :

cat filename | awk -F"|" '
{printf("%s|%s|%s|eor",$1,$2,$3)}'

but this gives :

aaaa|bbbb|cccc^M|eor
dddd|eeee|ffff^M|eor
gggg|hhhh|iiii^M|eor

With my limited knowledge of awk, could someone suggest a script that would do this, or better still, identify how awk (if it can that is), searches for this type of character (^M indicates a newline I think)

thanks to you all!
John

[RAC_1]

do this
cat test | awk -F"|" {printf("%s|%s|eor\n",$1,$2,$3)}'

Hope this helps.

[Chris Wilshaw]

As an alternative, using sed you could try

cat | sed s/.$/\|eor/g

This will replace any character followed by the end of line marker ($) with |eor

[Jean-Louis Phelix]

hi,

You should perhaps try to delete these ^m before reformatting using either sed ot dos2ux.

dos2ux filename | awk '{printf "%s|eor\", $0)}

or the most simple :

sed 's/^\(.*\)^M/&|eor/' filename

use ^V^M to avoid ^M interpretation.

Regards.

[u856100]

Hi,

I missed the \n in my previous response!

I have been using :

cat filename | awk -F"|" '
{printf("%s|%s|%s|eor\n",$1,$2,$3)}'


The awk script you mentioned, I think would miss out a column I need (cccc - eeee and iiii)

thanks
John

P.s. I don't know why, but I have just logged in as myself, and got logged in as some person called Lemaitre!

whats going on?????

[Jean-Louis Phelix]

Hi Chris,

Clever '.$' :^) , not to deal with ^M interpretation. I like it !

Jean-Louis.

[u856100]

people!

in addition to the fact that I have somehow logged in as another person (le-maitre).... I CANT assign points!!!

NOOOO!

[H.Merijn Brand (procura]

# perl -pi -e's/[\r\n]+$/|eor\n/' filename

[John Meissner]

did you ftp this file to your server? if you did ftp it... did you use bin or ascii mode? sometime i noted that using the wrong mode will add those characters to the end of each line. Try ftp'ing the file again and see if you get the same results.

If you didn't ftp the file then just follow the suggestions above.

[V. Nyga]

Hi,

only an additional info:
this ^M is the UX-interpretation of a DOS carrige return.
To eliminate this character you only need a dos2ux!
dos2ux file>file2
and the ^M will disappear!

Regards
Volkmar