How to extract certain portion of a file that starts with a pattern?

typecasket · 02-22-2017, 11:59 PM

Hi, seeing sample INPUT.TXT below how can I extract all portions that has "HDRHEADER0001" occurrences using sed? I tried using sed command below but requires trailing pattern.
sed -n "/^HEADER0001/,/^,<TRAILER>$/p" INPUT.TXT>OUTPUT.TXT

HDRHEADER0001 X004010850P
BEG00SAD202659801032017021699CANE
HDRHEADER0002 X004010850P
BEG00SAD202611701012017021499CANW
DTM01020170214
N1ST 92 0642397236
N315829 RUE BELLERIVE
N4MONTREAL QCH1A5A6 CANADA
HDRHEADER0003 X004010850P
BEG00SAP521006901012017021399CANOUT B16885
DTM01020170213
N1STCEGEP SAINT LAURENT 92 0642385892
HDRHEADER0001 X004010850P
BEG00SAD202659801032017021699CANE
HDRHEADER0003 X004010850P
BEG00SAP521006901012017021399CANOUT B16885
DTM01020170213

Expected output:

HDRHEADER0001 X004010850P
BEG00SAD202659801032017021699CANE
HDRHEADER0001 X004010850P
BEG00SAD202659801032017021699CANE

rknichols · 02-23-2017, 10:06 AM

First, there's a typo in your post since '^HEADER0001' doesn't match the actual lines that begin "HDRHEADER0001".

What defines the "section" that you want? If for "HDRHEADER0001" it's always just the next line, that's easy:

Code:

sed -n "/^HDRHEADER0001/,+1p" INPUT.TXT>OUTPUT.TXT

It it's everything up to but not including the next "HDRHEADER" line, that's going to be quite a bit more complicated. Yes, sed can do it (its language is Turing-complete), but it might not be a suitable tool for the job. That task would be simple in awk or perl, but somewhat convoluted in sed.

r3sistance · 02-23-2017, 12:41 PM

If you are just after the following line, you could use grep...

grep -A 1 "^HDRHEADER0001" INPUT > OUTPUT.TXT

should work.

typecasket · 02-23-2017, 08:36 PM

Quote:

Originally Posted by rknichols

First, there's a typo in your post since '^HEADER0001' doesn't match the actual lines that begin "HDRHEADER0001".

What defines the "section" that you want? If for "HDRHEADER0001" it's always just the next line, that's easy:

Code:

sed -n "/^HDRHEADER0001/,+1p" INPUT.TXT>OUTPUT.TXT

It it's everything up to but not including the next "HDRHEADER" line, that's going to be quite a bit more complicated. Yes, sed can do it (its language is Turing-complete), but it might not be a suitable tool for the job. That task would be simple in awk or perl, but somewhat convoluted in sed.

Sorry for the typo, yes it can be more than one line following "HDRHEADER0001".