Refactoring code to use Stream pipelines

Converting code to use Stream pipelines can be slightly difficult due to multiple reasons, such as unfamiliarity with the API, or an inability to visualise.

This is a common question for PEs (specifically PE2).

An approach could be:

1. Consider how the question would be approached without considering functional programming.
2. Think of the relevant API calls.

---

Example

Consider a list of students, where each student has an attribute marks which is a double[] containing all their marks for all their assessments, as well as a name which can be obtained from getName(). The Student class also contains a method getMarks(double[] weights), which obtains the total mark based on the weights passed in.

Given another double[], weights which contains the weights of the assessments, get the list of students that score above a given mark threshold.

Doing this normally, we would

1. calculate all the marks for each student AND
2. remove all students not above threshold/add all students above threshold to a new list
3. get the names

 
List<String> getStudentsAboveThreshold(List<Student> students, 
	double[] weights, double threshold) {
	
	List<String> result = new ArrayList<String>();
	
	for (Student s : students) {
		if (s.getMarks(weights) > threshold) {
			result.add(student.getName());
		}
	}
 
	return result;
}
 

Now, if we think of it from the Stream pipeline perspective,

1. we can filter all the students above threshold
2. we can map the Student objects into their names.

 
List<String> getStudentsAboveThreshold(List<Student> students, 
	double[] weights, double threshold) {
	
	return Stream.of(students)
		.filter(s -> s.getMarks(weights) > threshold) // step 1
		.map(s -> s.getName()) // step 2
		.toList(); // since they want a List<String>
}