Scripting in solaris bourne shell

Question

I was wondering how would it be possible to match a string in a text file and then to choose a value present there.
For exapmle; there is a line in a text file that reads
ACT=ajbg110;PCN=55768;IUYT_ghTTY=88999yt;.......
I need to make a script that is to pick ACT value and PCN value. These values are ajbg110 and 88999yt respectively. Fields are termintaed by ';'
Limitations are :
1) file contains thousands of such lines.
2) each line contains many such fields, not three only. Even there is no fixed number.
3) The position of these fields keeps changing. i.e; ACT is coming first, but it may come in the middle of
the line.
4) value length may change too.
Is threre a possible strategy to get these values seperated.

Answer 1

using perl:
my @lineholder; #declare an array
my @splitholder; #another array
# Supposing you're reading this from the file /tmp/infile
my $infile = "/tmp/infile";
open(INFILE, "< $infile") or die "Unable to open $infile: $!"; #Open a File handle while reading from file
while(<INFILE>) {
push( @lineholder, split( ';', $_ ) ); #Split STDOUT at delim ";" and push into lineholder
for (@lineholder) {
next if ( $_ =~ m/ACT|PCN/ );
# print $_; #Print line if it matches regex above
@splitarray = split( '=', $_ );
print "ACT and PCN pair: @splitarray \n";
I haven't tested it, and since it's perl, I'm sure there's a better way of doing it than this.
HTH