Mark and Sweep: Find the Garbage
Problem
A program has n objects numbered 0 to n-1. The first line is n. The second line lists the root objects (objects the running program can use directly), or a single - if there are none. Each next line a b means: object a holds a reference to object b. An object is garbage if it cannot be reached from any root by following references. Print the garbage objects in ascending order, or none.
Constraints
- 1 <= n <= 10^5
Examples
Input
5 0 0 1 1 2 3 4
Output
3 4
Solutions
Garbage collectors do not count references (which fails for cycles like 2 and 3 in the example). They mark everything reachable from the roots, then sweep the rest. Two objects that point to each other are still garbage if no root reaches them.
/*
* Problem : Find the garbage.
* Input : n, the roots, then reference lines.
* Output : The garbage objects.
* Approach: Mark and sweep.
* Time : O(n + m)
* Space : O(n + m)
* Java : 8 or newer.
*/
import java.util.Scanner;
public class Main {
// This method solves the task. It receives the input and returns the answer.
static String solve(String[] lines) {
int n = Integer.parseInt(lines[0]);
// Step 1: Read the roots (the objects the program can reach directly). A line with only "-" means: no roots.
java.util.ArrayDeque<Integer> pending = new java.util.ArrayDeque<>();
if (!lines[1].equals("-")) {
for (String root : lines[1].split(" ")) {
pending.push(Integer.parseInt(root));
}
}
// Step 2: Read the references. references.get(a) holds the objects that object a points to.
java.util.List<java.util.List<Integer>> references = new java.util.ArrayList<>();
for (int i = 0; i < n; i++) {
references.add(new java.util.ArrayList<>());
}
for (int i = 2; i < lines.length; i++) {
String[] edge = lines[i].split(" ");
references.get(Integer.parseInt(edge[0])).add(Integer.parseInt(edge[1]));
}
// Step 3: MARK. Walk from the roots and mark every object we can reach.
boolean[] marked = new boolean[n];
while (!pending.isEmpty()) {
int object = pending.pop();
if (marked[object]) {
continue;
}
marked[object] = true;
for (int next : references.get(object)) {
pending.push(next);
}
}
// Step 4: SWEEP. Every object that is not marked is garbage.
StringBuilder garbage = new StringBuilder();
for (int i = 0; i < n; i++) {
if (!marked[i]) {
garbage.append(garbage.length() == 0 ? "" : " ").append(i);
}
}
return garbage.length() == 0 ? "none" : garbage.toString();
}
public static void main(String[] args) {
Scanner scanner = new Scanner(System.in);
java.util.List<String> all = new java.util.ArrayList<>();
while (scanner.hasNextLine()) {
String line = scanner.nextLine().trim();
if (!line.isEmpty()) {
all.add(line);
}
}
String[] lines = all.toArray(new String[0]);
System.out.println(solve(lines));
}
}
Hints (1)
- Why does reference counting fail for cycles but mark and sweep does not?