...and when to cut and paste code...

The idea of cut and paste code makes me shudder as it tends to create bugs very easily (speaking from experience), the checking and correction of which may well consume any time saved by the cut and paste operation. Fortunately I think it can often be avoided:

• if you are pasting code and not changing it (i.e. it is a duplicate) then that certainly needs to be a subroutine (or sometimes just a loop depending on the complexity). Remember that if the source and destination of the cut and paste are identical now, they should probably always be identical, and the only way to ensure that is to use a single subroutine.
• if you are pasting code and changing the copy slightly, it may still be better as a subroutine if the code is fundamentally the same.
• as others have said, if you have a hard time understanding code with a lot of subroutine calls, then you probably need better names. After all, the built-in Perl functions are also aggregating code and you're far better off calling split than pasting in code to do the splitting each time! In fact, many people will often start a project with some vague pseudocode like

  for each data file {
    check for bad data and skip file if found
    repair damaged data
    calculate statistics
    add to database
    add to report
  }
  [download]

  and something like that could easily be made into the program's main loop almost word-for-word, making the code self-documenting. </code>

Finally, here's another good Monastery thread on this topic: When do you function? - I'm certain there are more.

--
I'd like to be able to assign to an luser