Awk Variable Substition

[gstonian]

Hey,

Just want to strip everything prior to the last backslash for a paramter in a file using awk.

i.e.

cat file | awk ' BEGIN { FS="\n" ; RS="" } { print $5 } '
Test\SQL Servers\Server1
Edinburgh\SQL Servers\Server2
London\SQL Servers\Server3
Test Servers\Server4

to be
Server1
Server2
Server3

Outside of awk I know i can do ${variable##*\\}
but not sure how to get it to work alongside awk and everything I see appears to work on the first occurence i.e.

awk ' BEGIN { FS="\n" ; RS="" } { print substr($5,(index($5, "\\")+1),20) } '
works on the first occurence of "\" but doesn't handle any further occurences

I'm sure I'm missing something simple here so please let me know ;-)

Cheers

[john korterman]

Hi,

you just want the last part?

awk -F\\ '{print $NF }' infile

regards,
John K.

[James R. Ferguson]

Hi:

In Perl, dimply:

# perl -nle 'print $1 if m/.*\\(.*)/' filename

Regards!

...JRF...

[gstonian]

Thanks but it's not the last part of the file, just simply the last part of the parameter ($5) I want to print. There are other fields after this one $6,$7 etc. I just simplified the issue down .

Thanks for the perl solution but would like to use awk if possible

Cheers

[James R. Ferguson]

Hi (again):

OK, then in 'awk':

# awk '{sub(/.+\\/, "",$0);print}' filename

Regards!

...JRF...

[Tom Danzig]

Sounds like basename would be the simplest in your situation:

# basename London\SQL Servers\Server3
LondonSQL

Easiest awk would be:
awk -F'\' '{print $NF}' filename

[Tom Danzig]

Please excuse my ridiculous 'basename' suggestion above.

[gstonian]

NO problem,

I am trying to do something with other parameters and formatting the input so ideally would like something to be able to do something all in one line

i.e.

| awk ' BEGIN { FS="\n" ; RS="" } { print $1 ", " $2 ", " $3 ", " ", " $6 } '

[Peter Nikitka]

Hi,

just split the filed in request using '\' as seperator:

awk 'BEGIN { FS="\n" ; RS="" }
{n=split($5,a,"\\"); if(n>1) print a[n]}' file

mfG Peter

[Peter Nikitka · Accepted solution]

Hi,

sorry for the follow-up, but I missed your last post.

To include my solution into in your code, use
the splitted result directly. If you set ORS, then there is no need to diddle with output formatting:

| awk ' BEGIN { FS="\n" ; RS=""; ORS=" " }
{ n=split($5,a,"\\"); print $1,$2,$3,a[5],$6}'

mfG Peter

[gstonian]

Thanks for all your answers. I ended up taking your answers and ended up with using awk twice to get what I wanted, which was.


cat file.out | awk -F\\ '{print $NF } ' | awk ' BEGIN { FS="\n" ; RS="" } { print $1", "$2", "$3", "$4", "$5" "$7" "$8} '

Hope someone else finds a use for this !

[gstonian]

Thanks