Grouping Classes with Packages in Java

Large Java projects have a lot of classes and it's difficult to manage them if they are stored in the same directory. Packages provide a mechanism for grouping classes together in the same module (or package). In this topic, you will learn how to use them to better systemize your classes and keep your app tidy.

The basics of packages

Packages have many advantages, for example, they allow us to:

• group related classes together, which makes it easier to figure out where a certain class is;
• avoid conflicting class names;
• control access to classes and members with access modifiers.

According to the naming convention, package names are always lowercase, for example:

model
collection
utils

A package can contain other packages, and the whole structure resembles directories in a file system. Here is an example project with a simple tree of packages and classes.

At the top of the tree is the src directory. This is the source root directory. In this tree, the full name of the User class is org.company.webapp.data.User.

You can output the full name using the getName() method as shown below:

System.out.println(User.class.getName()); // org.company.webapp.data.User

Classes declared inside a package have a special package keyword at the top of the file.

package org.company.webapp.data;

public class User {
}

Avoiding conflicting class names

When you use external libraries, two classes may have the same name. Packages allow us to avoid a conflict of class names because the full class name includes the package name. So even if two classes from different packages have the same name, their full names will be different. That is, of course, if there are no conflicts between the package names.

To avoid creating packages with the same names as other public packages, it is generally recommended to start your package hierarchy with the reversed domain name of your company (or another type of organization), for example:

org.company
org.hyperskill
net.labs

Importing classes

If two classes are located in the same package, using one class inside the other is no problem. If this is not the case and the classes are in different packages, you need to write an import statement to use one class inside the other. The import statement is defined by the import keyword.

Here is an example: we have two public classes in different packages:

org.hyperskill.java.packages.theory.p1.A
org.hyperskill.java.packages.theory.p2.B

To use class B inside class A we should use an import statement.

package org.hyperskill.java.packages.theory.p1;  // current package

import org.hyperskill.java.packages.theory.p2.B; // it's required to use the import

public class A {

    public static void method() {

        B b = new B();
    }
}

If both the package declaration and import statements are present, the package must come before all imports! Otherwise, we get a compilation error.

It is possible to import all classes from a package and to do this, we need to write * in the import statement instead of a particular class name.

import org.hyperskill.java.packages.theory.p3.*; // import all classes from the package

Don't do this too often, in some cases this is considered a bad practice and can break the compatibility of your program with new versions of Java. Here is an interesting discussion about such statements.

If two classes belong to the same package, you don't need to import them to each other.

There is also a way to use a class from another package without the import statement. In this case, you should write the full class name (including the full package path) instead of just the class name (short name). This is how we would use the Scanner class without explicitly importing it first:

java.util.Scanner scanner = new java.util.Scanner(System.in);
java.util.Date now = new java.util.Date();

Let's polish the information about access to classes inside the package with the following example:

The classes City and Area are located in the same subpackage – state, so you can use one class inside the other with a short name. The same thing is true for the classes Urban and Rural in the territory subpackage.

If you want to use any class of the territory subpackage inside the class of the state subpackage or vice versa, you need to write the full name of this class, import the class, or import the whole subpackage. Moreover, if you want to use classes from the subpackages state or territory inside the Republic class from the country package or vice versa, you also need to write a full name or import the class. This should be done even if these packages are in the same root package (here all packages are a part of the country package).

Importing standard classes

There is no difference between importing standard or custom classes. For example, many Java developers use java.util.Scanner to work with the standard input/output. In their programs, they include the following import:

import java.util.Scanner;

After this, we can create an instance of the Scanner class like in the examples above and use it in our programs.

Even though we have to import most packages to use their classes, in Java there are built-in packages that is always automatically imported, namely, java.lang. This package contains many widely used classes, such as String, System, Long, Integer, NullPointerException and others.

Static imports

We can also import static members (methods and fields) of a class inside another class. If we write an asterisk * in the import statement, we don't need to write the imported class name before invoking static methods or reading static fields.

Here is an example of the static import of the class Arrays, which contains a lot of useful methods for processing arrays.

package org.hyperskill.java.packages.theory;

import static java.util.Arrays.*; // instead of the statement "import java.util.Arrays;"

public class Main {

    public static void main(String[] args) {
        int[] numbers = { 10, 4, 5, 47, 5, 12 }; // an array

        sort(numbers); // instead of writing Arrays.sort(...)

        int[] copy = copyOf(numbers, numbers.length); // instead of writing Arrays.copyOf(...)
    }
}

Default package

If we do not write a package statement before defining our class, it will be placed inside the default package. This package has a big disadvantage — classes located here can't be imported to classes located inside named packages.

The following class cannot be used in a class located inside packages since there is no package declaration.

// no package declaration

public class Person {
    String firstName;
    String lastName;
}

Do not use the default package for real applications. It is perfectly fine for simple, educational applications, like "Hello world", but more complex projects will be better in named packages.

Conclusion

Packages are a very useful tool for OOP projects. They allow grouping classes into packages and to structure the source code better, and they make it more maintainable. It is very important for big projects that can consist of thousands of classes. Packages are also very helpful for avoiding conflicting class names because the full class name includes the path of the whole package. If we are careful with the naming of the package itself, there should be no conflicts!

Another thing is that packages affect the visibility of classes and class members to each other. That means that we should keep in mind the imports, static members, and the default package.

Creating packages even for small applications is great training for your great future projects!