My first programming instructor taught me to code correctly the first time, by verifying all code on paper before we even touched a computer.
Many of the modern code analysis tools flag long methods for being too long, by exceeding some arbitrary length.
These programs purpose splitting a "long" method into several single-use methods.
Designing, writing and testing with these many smaller methods and classes is easier, no question.
However, newcomer(outside) readability degrades with each split.
When I was taught to split the design into parts, I was also taught to recombine single-use methods and in-line multi-use methods as much as possible.
My question is: Has the recombine step been left out of modern CS education?
Aucun commentaire:
Enregistrer un commentaire