Writing code

If you're familiar with object-oriented programming, then it's fairly easy to start writing Apex code. The rigid structure of the language provides easy-to-follow rules to ensure that your code is valid. In fact, the compiler even prevents you from saving code that is syntactically invalid and provides easy-to-understand error messages that include line numbers. On top of this, all transactions are monitored by the platform and will be automatically terminated if an exception is thrown or a limit is exceeded. Essentially, you can write code in your Developer Edition without having to worry about breaking anything.

Irrespective of whether you are using the Force.com IDE or the web-based editor, there are only two places in the Salesforce1 Platform where you can store Apex code: triggers and classes.

Just like all of Apex, triggers are compiled into a set of instructions stored as metadata. However, what makes triggers unique is that these instructions can only be invoked by the platform itself. Each trigger is tied to the operation on records of a specific sObject type in the database. The individual operation (insert, update, and so on) for the data type can also be specified. Triggers allow you to modify the default behavior of the platform when a record is saved.

As triggers are only called by the platform, classes end up being the main stomping ground for Apex code. Classes aren't limited to being called by the platform. You can interact with them from other classes and triggers, call methods on demand from Execute Anonymous (similar to a command line editor), and even expose them outside as web services.

Classes can be instantiated just like any other object. However, as the Salesforce1 Platform is a multitenant service, we typically use a lot of static methods to avoid creating a bunch of short-lived instances. These methods can be called directly without first instantiating a class. The built-in Apex classes include a lot of static methods, and the best practice is for your code to do the same as it will use fewer resources and this reduces the chances of hitting the platform limits. This is shown in the following code:

public class myClass{
public void myInstanceMethod(){
  system.debug('This is my instance method');
}

public static void myStaticMethod(){
  system.debug('This is my static method');
}
}

//Construct an instance of myClass and call a method
myClass instanceOfMyClass = new myClass();
instanceOfMyClass.myInstanceMethod();
//Call a static method without instantiating the Class
myClass.myStaticMethod();

We'll discuss triggers and classes in detail in Chapter 4, Triggers and Classes. For now, you just need to be familiar with their names and general use.