AWK to search a specific sequence and if found search in the next line another sequence
As in life, when processing data it's much easier to make decisions based on what has happened in the past (data you have read) rather than what will happen in the future (data you are going to read) so instead of saying "if I have X and the thing after it is Y" write your requirements as "if I have Y and the thing before it was X" and the software to implement it usually becomes much more simple and obvious.
Is this what you're trying to do (using any awk in any shell on every Unix box):
$ cat tst.awk
($2 == 17) && (p1 ~ / 14 /) && (p2 ~ / 4b /) {
sub(/ 17 /," 12 ")
}
{ p2=p1; p1=$0" "; print }
$ awk -f tst.awk file
0000 09 06 07 04 00 00 01 00 1d 03 4b 2c a1 2a 02 01
0010 b7 09 01 47 30 22 a0 0a 80 08 33 04 03 92 22 14
0020 12 f0 a1 0b 80 00 81 00 84 01 00 86 00 85 00 83
0030 07 91 94 71 06 00 07 19
0000 09 06 07 04 00 00 01 00 2b 03 4b 27 a1 25 02 01
0010 00 09 01 66 30 1d a0 0a 80 08 33 04 03 92 22 14
0020 12 f0 a1 06 82 00 84 00 85 00 82 07 91 94 71 06
0030 00 07 19
If that's not all you need then edit your question to clarify your requirements and provide more truly comprehensive sample input/output including cases that the above doesn't work for.
I'm using sub(/ 17 /," 12 ") above instead of $2=12 to preserve white space between fields. It's safe to do that because the target field is $2, if it was any other field you couldn't do that as a field before the target one might also be 17. There are various sub()/match()/substr() ways to handle that of course.
Based on your shown samples, could you please try following, written and tested with GNU awk.
awk '
!NF{ found1=found2=0 }
/(^|[[:space:]])4b([[:space:]]|$)/{
found1=1
print
next
}
found1 && /(^|[[:space:]])14([[:space:]]|$)/{
found2=1
print
next
}
found1 && found2{
for(i=2;i<=NF;i++){
if($i==17){ $i=12 }
}
print
next
}
1
' Input_file
Explanation: Adding detailed explanation for above.
awk ' ##Starting awk program from here.
!NF{ found1=found2=0 }
/(^|[[:space:]])4b([[:space:]]|$)/{ ##Checking condition if line has 4b with spaces or coming in starting or ending of line.
found1=1 ##Then set found to 1 here.
print ##Printing the current line here.
next ##next will skip all further statements from here.
}
found1 && /(^|[[:space:]])14([[:space:]]|$)/{ ##Checking if found1 is SET AND if line has 14 with spaces or coming in starting or ending of line.
found2=1 ##Setting found2 to 1 here.
print ##Printing the current line here.
next ##next will skip all further statements from here.
}
found1 && found2{ ##Checking condition if found1 and found2 is SET then do following.
for(i=2;i<=NF;i++){ ##Traversing through all fields here starting from 2nd field.
if($i==17){ $i=12 } ##Checking condition if field value is 17 then make it 12.
}
print ##Printing current line.
next ##next will skip all further statements from here.
}
1 ##1 will print current line.
' Input_file ##Mentioning Input_file name here.
Your attempted awk command is pretty good, you just need to make sure to use -v RS= (empty RS) to make each paragraph a record.
Following should work for you in gnu-awk:
cat fmt.awk
{
ORS = RT # set ORS same RT variable populated using RS
}
{
r = t = p = ""
for ( i = 1; i <= NF; ++i ) {
# set r = 1 when we get 4b
if ( $i == "4b" )
r = 1
# set t = 1 when we get 14 when r==1
if ( r && $i == "14" )
t = 1
# when we get 4 digits save the position
if ($i ~ /^[0-9]{4}$/)
p = i+1
# replace 17 with 12 when we get 17 when t==1
if ( t && p == i && $i == "17" ) {
$0 = gensub("((\\S+\\s+){"i-1"})\\S+", "\\112", 1)
break
}
}
} 1
Run it as:
awk -v RS= -f fmrt.awk file
0000 09 06 07 04 00 00 01 00 1d 03 4b 2c a1 2a 02 01
0010 b7 09 01 47 30 22 a0 0a 80 08 33 04 03 92 22 14
0020 12 f0 a1 0b 80 00 81 00 84 01 00 86 00 85 00 83
0030 07 91 94 71 06 00 07 19
0000 09 06 07 04 00 00 01 00 2b 03 4b 27 a1 25 02 01
0010 00 09 01 66 30 1d a0 0a 80 08 33 04 03 92 22 14
0020 12 f0 a1 06 82 00 84 00 85 00 82 07 91 94 71 06
0030 00 07 19