I had this problem where I needed to match a lot of ids (approx. 10.000) against a data file. It would take too long time to write a program that validated the ids precense. What to do?

The Unix shell to the rescue. I remembered grep have some of the functionality I wanted.

man grep

Gave me that

grep -f <file>

will read the match patterns from a file. Next problem is to get grep to match the pattern on a fixed word.

grep -f <file> -Fw <match file>

This got me some of the way. Only problem remaining is the match result wasn’t complete - treating the file as ascii helped.

grep -f <file> -Fwa <match file> | cut -d \; -f <index> > matches

Gave the result I hoped for. All of the matching ids in a file ( matches ) of its own.

Thank you grep