Where to close the file? 1

[AnotherAlan]

Hi All,

I've just returned from two weeks in Italy and the brain is fried, consequently can't get my head around this today.

I have a script that takes a file and splits it based upon the data between the /^imp_/ match in the below nawk statement.

This has been working great, but while I was away they have been putting more trades through the system, meaning that the nawk statement below breaks at file 21.

nawk '/^imp_/ {f="split."++d} f{print > f}' $FILENAME

I know I can put a close in here, but when I tried using

nawk '/^imp_/ {f="split."++d} f{print > f}; {close (f)}' $FILENAME

it is only printing the last line to the split.xx file and not the whole block of code I need.

I've also tried this;

nawk '/^imp_/ {f="split."++d} f{print >> f}; {close (f)}' $FILENAME

but this is putting the line breaks in the wrong place, as in the second imp_ to END block below;

$ cat -vet IMPORT.DAT
imp_blim$
keys,BRO3,FIN,,10280591,,$
block_detail,NL0000888691,"ADVANCED SHARES",ISIN,02-Sep-09,16:19,S,10000.00,10.11770000,07-Sep-09,EUR,,101177.000,P,202.350,,0.20,,N,O,1$
prorate,,0.000,0.000,0.000,202.350,,N$
END$
imp_blim$
keys,BRO3,FIN,,10280591,,$
block_detail,NL0000888691,"ADVANCED SHARES",ISIN,02-Sep-09,16:19,S,10000.00,10.11770000,07-Sep-09,EUR,,101177.000,$ - this line break should not be here
P,202.350,,0.20,,N,O,1$
prorate,,0.000,0.000,0.000,202.350,,N$
END$

BTW - The END lines are not present in the original file, I have added them for easier reading.
I hope this makes sense, if not then please refer to the opening sentence.

I'm no expert on awk/nawk and will continue to test, but
all help / pointers would be appreciated.

 

[moonring]

Maybe something like this,

nawk '/^imp_/{ [COLOR=red]close(f)[/color]; f="split." ++d }{ print > f }' filename

 

[AnotherAlan]

Beautiful, I forgot about working backwards...

Thanks very much, that was a big help.

Alan