In some programming API's I see a list of methods to call, like getBoolean(String key, getDouble(String key), and getString(String key). Some other API's use a general get(String key) method, and return an Object which you are supposed to cast to an appropriate type yourself.
Now I am writing my own data accessing point, and I am wondering which approach to use. What are the advantages and disadvantages of each approach? When would you choose one over the other?
Advantage: getBoolean(), getDouble(), etc. allow you to return the respective primitive types. As far as I've seen, that's the primary reason anyone writes methods like that.
new Long(1).equals(1)?Provide getters for the types that are most likely to be used. There's no real right or wrong way.
It depends on the purpose of the library. When the output is a predictable set of items, then toy can have specific names. As in ResultSet. If it is generic, then you will need generic get methods. Like ObjectOutputStream
In a very high level sense, you might need both: getBoolean, getDouble, getIngeter for the primitives (or their respective wrappers), a getString, for Strings and a generic get or a getObject for getting out objects.
However, this is a very generic answer for a very generic questions. What your tries to do very much decided such stuff.
Two questions: 1) Why don't you use general Properties instead, à là:
String getName() Address getAddress() Date getDateOfBirth() etc?
2) If you want to use methods like:
String getString(String key) Double getDouble(String key) Address getAddress(String key) How on earth would I, as the user, know, which key's are associated with objects of type String, which are associated with objects of type Double, etc?
I would recommend going with a solution similar to 1). If I didn't misunderstand your question, that is.
