Extract between lines starting with and

Shell Programming and Scripting
1

I have a combination of patterns to search.

file.txt contains below:

H2016-02-10
A74867712
I1556539758
Xjdflk534jfl
W0000055722327732
W0000056029009389
A74867865
I1556536434
W0000055822970840
W0000055722325916
A74868015
I1556541270
C0000055928920421
E

lines starting with A are start of a recordset till we reach the next line starting with A. so in this file, i have 3 recordsets.

now from each recordset, I want to print first 2 lines, if there is no line starting with X in the recordset.

for example: from this file, i need to print:

A74867865
I1556536434

but i dont need to print:

A74868015
I1556541270

because there is no line starting with W in this recordset.

2

Do not expect us to read your mind as your post is not making any sense to me so state what you need clearly...

3

my requirement is:

file.txt can have lines starting with either A,I,X,W,C,P, ignoring header line starting with H at the beginning of the file
and footer line starting with E at the end of file.

lines starting with A are start of a recordset till we reach the next line starting with A.

in the example i have given, tehre are 3 recordsets.

A74867712
I1556539758
Xjdflk534jfl
W0000055722327732
W0000056029009389

A74867865
I1556536434
W0000055822970840
W0000055722325916

A74868015
I1556541270
C0000055928920421

I want to print first 2 lines of those recordsets, for which there is no X line, whenever there is a W line.

so in the above 3 recordsets i need to print only the 2nd one:
i.e.

A74867865
I1556536434
W0000055822970840
W0000055722325916

i am trying with awk.

4

Please use code tags as required by forum rules!

How about

awk '
(/^A/ || /^E/) && 
 W && !X        {print A
                 print B
                }
/^A/            {W = X = 0
                 A = $0
                 getline B
                }
/^W/            {W = 1
                }
/^X/            {X = 1
                }
' file
A74867865
I1556536434
5

Another way, first putting empty lines between records:

awk '/^[AHE]/{print x}1' file | awk '/\nW/ && !/\nX/{print $1 ORS $2}' FS='\n' RS=
6

@RudiC:
Thanks for that it works perfectly.