I have heard the “first rule of the macro club” to be “don’t write macros”.

I have heard the “first rule of the macro club” to be “don’t write macros”. The idea is before writing a macro, you should try writing it as a function instead. If that is possible, that is usually better, because functions, unlike macros, can be passed around as first-class values, and I think they are easier to debug. You should break that rule only when the behavior can’t be written as a function, such as these cases: • The call needs to avoid evaluating its arguments. For example, the `if` built into the language is sometimes defined as a macro. (if (= (+ 2 2) 4) (print "math works") (print "math is broken")) If `if` were a function, it would first print both statements, and then return the return value of `print` in whichever branch. By making `if` a macro, it can avoid evaluating the branch that is inapplicable. • The call relies on information about the environment only available at compile-time. Perhaps the macro reads a configuration file on the developer’s computer to decide how to set something up. • You have profiled the program and determined that it is better to run the function at compile-time. For example, you might want to make `(regex "[a-z][a-z0-9]+")` compile the string to a regular expression at compile-time instead of run-time.

Posted by roryokane on May 8, 2015, 8:32 PM

Tags

• reliability (Other)

Related issues

• This is my biggest problem with agents, they constantly defensively code for cases that have no chance of happening and in most cases a p...
• I wish there was something like LEAN for biology.
• The issue is really the extreme difficulty of programming them, especially for anything sequential (rather than combinatorial).
• I probably could have beared with it if it was only limited to programming, but it showing up in creative fields has made me feel all sor...
• I hate that I'm in deep enough to know what you mean.
• The biggest problem with “AI” is that big corps are shoving it down our throats.