When creating my object I pass it an array of a row from my database. (everything in the array we will need, disregarding unnecessary elements at sql query level)

When I need to access certain array elements from within my class, I do so like


However, As I continue development, I sometimes forget what exactly is in this passed array.(this itself doesn't seem good)

I am wondering if their is a professional approach to dealing with this, Or am I the only one who has these "I wonder whats in the array" thoughts.

One approach to tackling this could be that when we originally pass the array, in the constructor, we assign each element of the array to its own variable, but is this considered professional practice?

Additionally by doing this, we could make those variables constants, in a attempt at immutability.

Overall I am trying to adhere to good software craftsmanship.


  • 1
    They're called "magic strings." In a proper OO language, you would create a class that holds the elements of your row. You could then take advantage of intellisense in your IDE to pick the element you want from the row object. Commented Jan 30, 2014 at 0:15
  • Just read about that term, So you're saying I should dissect the array into singular variables, Curious as to why all the OOP and programming books have never explicitly touched on this, nor mentioned that term in relation with the issue.
    – cecilli0n
    Commented Jan 30, 2014 at 0:22
  • Probably because PHP programmers have been doing it the way you described for a long time, and online PHP tutorials (and most books, I would imagine) just parrot the same approach. Have a look here: css.dzone.com/books/practical-php/practical-php-patterns-data Commented Jan 30, 2014 at 0:26
  • ...although I do note that is a whole lot of ceremony for the simple problem of packaging data rows. In c#, we would simply use Properties, not getters and setters. Commented Jan 30, 2014 at 0:31
  • The books/tutorials I have been reading mostly use Java/C++ as their language of example, But you are right it does seem alot of overhead to define each element in the array into its own property, which is why I haven't started using this approach and decided to seek advice here. Going to read that link you gave now. Thanks
    – cecilli0n
    Commented Jan 30, 2014 at 0:38

1 Answer 1


The best way is to thoroughly describe the structure of your arrays within comments in your code, but I use var_dump() at times that I forget. It will dump a detailed list of each element within an array or object (name, type, key, value, etc) including deeply nested arrays. I've been using PHP for more than 10 years so this sort of thing is second nature to me (PHP didn't really have debuggers or high-end IDE's back than, so you were stuck digging through vars manually)

And yes, it's OK to manually name elements within array properties when constructing a class object. This is actually common practice in configuration files. For example, a config class can have the property $database, and then set $this->database['user'], $this->database['pass'], etc within the constructor.

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service and acknowledge you have read our privacy policy.

Not the answer you're looking for? Browse other questions tagged or ask your own question.