Shell Programming and Scripting

I am studying awk, and have a test on variable "RS "

$ cat > /tmp/test
1
2
3
here
a
b
c

and here is my command

$ awk 'BEGIN { FS="\n" ; RS="here"} {print $1" " $2" " $3}' /tmp/test
1 2 3
ere a b

you can see, awk did not treat "here" as record separator, instead, it use "h" as separator, why this happen? I think it should use "here" not "h", am I right?

how can I use "here" as separator? thank you!

In AWK RS should be a single character.
In GAWK RS can contain regexp.

Code:

sed "s/here//" tmp | awk 'BEGIN { FS="\n" ; RS=""} {print $1" " $2" " $3}'

RS doesn't do that it only pays attention to the first character of RS. The behavior you see is correct.

thank you anbu23, learn something from you,

but, looks like gawk also does not work fine with "here"

$gawk 'BEGIN { FS="\n" ; RS="here"} {print $1" " $2" " $3}' /tmp/test
1 2 3
a b

and with /^here/;

$ gawk 'BEGIN { FS="\n" ; RS="/^here/"} {print $1" " $2" " $3}' /tmp/test
1 2 3

Gawk works fine...

Code:

$ gawk 'BEGIN {FS="\n" ; RS="here\n"} {printf "Row:%d Data:[%s %s %s]\n", NR, $1, $2, $3}' file
Row:1 Data:[1 2 3]
Row:2 Data:[a b c]

AH! thank you! Ygor!