sed usage

Hi

I have a file with data similar to the line below

1'~'Rowone#@#@#2'~'Rowtwo#@#@#3'~'Rowthree#@#@#4'~'Rowfour#@#@#5'~'Rowfive#@#@#


all the data is in one single stream, i mean to say it's one
single line of data.

when i try to vi the file i cant because the file itself is huge
and the data embedded in the file is in one single big line.

i need a script or command that can help me to create a file
with data as below from the file shown above

1|Rowone
2|Rowtwo
3|Rowthree
4|Rowfour
5|Rowfive


as you see in the source file the column separator is '~' and the row terminator is #@#@#
this is want to change to ,| as the column separator and
unix carriage return as the row terminator

any help would be appreciated

thanks
dsshed

one way:

sed -e "s#'~'#|#g;s/#@#@#/#/g" file | tr '#' '\n'

vgresh99

your reply is very apt and it works , and thank you for that
but here is the problem, the command you send doesnt
work on the file i am operating on.

but when i cat the file and copy half of its content and paste
it to a new file and run your command on this new file it works like a charm.

but when i run this against my original file it does nothing.
as i said i can't even open the file with vi as it comes up
with a message as below
Line too long

is there a way to get around this limit of size on the line
i am reading from my file, whic seems to be an extremely
lengthy line and sed command is not able to process it

any advise or help is very much appreciated

dsshed

what does 'wc -c file' return?

If on Solaris, try /usr/xpg4/bin/sed instead of plain old sed.

yes, that did it
xpg4 version of sed works fine

wc -c file output is 4143
i tried on that file and it worked very well

i need to see how to fine tune it to work on a file of size
of 1434194
i ran the command on this its taking some time to
process, which is abvious but needs to be tuned


one other way i am trying is this
cat CS.txt | awk 'BEGIN { RS="#@#@#"} {print $1}'

in those lines


thank you

unfortunately any of the Sun-supplied awk-s don't allow specifying RS longer than 1 character - gawk does though.

Here's something to try with /usr/xpg4/bin/awk:

dsshed.awk:


/usr/xpg4/bin/awk -f dsshed.awk CS.txt

(no UUOC, pls)