Extracting multiple instances of a regex pattern from file using KornShell

Question

I have a file that may have multiple instances of a string pattern (starting with X ending with Y). I want to extract each instance that matching this pattern. These may be on a single line or new line may or may not be at the beginning or end of a line. I have tried using grep with a regex pattern of (X.*Y) but the version of KornShell (ksh) that is running the ksh does not recognize the -c option so having a tough time extracting what I need. I have also read that awk can do this but not finding good documentation out there for awk.

Any help would be gladly appreciated.

Answer 1

If the patterns are contained on a single line, and only one instance on each line, you could try

awk -f ext.awk file

where file is your input file and ext.awk is

BEGIN {
    pat="X.*Y"
}
$0 ~ pat {
    match ($0,"(" pat ")",a)
    print a[1]
}

Update

If more than one pattern can be found on a line, or patterns extend over multiple lines, you can replace ext.awk with:

{
    line=line $0 RS
}
END {
    while(match(line,/(X[^Y]*Y)/,a)) {
        print a[1]
        line=substr(line,RSTART+RLENGTH)
    }
}