Course Scheduler — Print Path (Simplified)

A simplified course-scheduler variant: no multi-prerequisite courses, and the answer is the explicit ordered path of courses (not just a count). Asked as the 20-minute coding portion of an Oracle Health phone screen, paired with a job-scheduler design in the same round. Requirements Input: a set of courses, each with at most one direct prerequisite (the prompt explicitly states there are no multi-prerequisite courses). Output: the ordered sequence of courses such that every course's prerequisite (if any) appears earlier in the sequence; print the path. The simplification (single-prerequisite-max) reduces the problem from a general DAG topological sort to a forest where each node has at most one parent. Notes With at-most-one-parent, the structure is a forest. Each tree's correct order is just root → ... → leaf, recursive or iterative. Algorithm: find all nodes with no prerequisite (roots). For each root, walk its tree in depth-first order, appending each node to the output list. If multiple roots exist (forest), emit them in any consistent order (alphabetical is the safest interviewer-friendly choice). This is materially simpler than LeetCode 210; the simplification was the interviewer's deliberate framing for a 20-minute coding window paired with a job-scheduler design afterwards. If the simplification is unexpectedly relaxed mid-round ("now allow multiple prerequisites"), pivot to the general topological-sort algorithm — same skeleton, in-degree tracking and a BFS queue. Output format: print the path on stdout, one course per line or comma-separated; the round did not specify, so confirm at the start. Preparation Implement the at-most-one-parent forest traversal in under 10 minutes. Have the general Kahn's-algorithm topological sort ready as the fallback in case the constraint is relaxed. This round paired coding (20 min) with system design (20 min) within a single 60-minute slot; do not over-engineer the coding solution at the expense of the design portion.