How to rebuild a tree from PostgreSQL without duplicate nodes? Closed

I'm developing a Spring Boot application that stores a tree structure in PostgreSQL.

The table schema is:

CREATE TABLE nodes (
    id BIGINT PRIMARY KEY,
    value VARCHAR(150),
    parent_id BIGINT
);

Example data:

1 | Management    | NULL
2 | IT Department | 1
3 | Developer     | 2

To rebuild the tree in memory, I'm using Java Collections:

private final Map<Long, List<Long>> tree = new HashMap<>();

public void loadTree(List<NodeEntity> nodes) {

    for (NodeEntity node : nodes) {

        if (node.getParentId() != null) {
            tree.putIfAbsent(node.getParentId(), new ArrayList<>());
            tree.get(node.getParentId()).add(node.getId());
        }
    }
}

The problem is that after reloading the data several times, some child nodes appear duplicated in the generated tree structure.

Observed behavior:

Management
 ├─ IT Department
 │   ├─ Developer
 │   └─ Developer

Expected behavior:

Management
 └─ IT Department
     └─ Developer

I suspect that the in-memory collections should be cleared before rebuilding the tree, but I'm not sure if that's considered the best approach.

What is the recommended way to rebuild a hierarchical tree structure from database records without introducing duplicate nodes?