2-D Matrix Range Sum with Updates (OA SumTable)

HackerRank OA for the SWE AI Trainer (LLM rater) role. Five TypeScript multiple-choice questions on code-quality judgement plus one Python implementation: process a stream of `get`/`set` queries against a 2-D matrix returning the rectangle sum for `get` and applying point updates for `set`. Requirements Reported OA format (HackerRank, camera-on, ~90 minutes): 5 multiple-choice TypeScript questions — pick the better-written of two snippets each. The implicit role context is LLM training / data annotation. 1 Python implementation — the SumTable problem below. SumTable: read stdin shaped like 2 3 1 2 3 4 5 6 5 get 0 2 1 3 set 0 1 10 get 0 2 1 3 set 0 0 -3 get 0 1 0 2 The first three lines specify a 2 × 3 matrix [[1, 2, 3], [4, 5, 6]]. The fifth line is the number of subsequent queries. Each query is either: set r c v — matrix[r][c] = v. get r1 c1 r2 c2 — print the sum of matrix[r1..r2][c1..c2] inclusive on the lower bound, exclusive on the upper? Verify boundary inclusivity from the examples — given get 0 2 1 3 on the starting matrix returns 1 + 2 + 5 + 6 = 14, the indices map to rows [0, 2) and columns [1, 3). Expected output for the input above: 14, 24, 7. Given the mixed update/query workload, the textbook answer is a 2-D Binary Indexed Tree (Fenwick) with O(log² N) per operation. A naive prefix-sum recompute on every set is O(R × C) per update and likely TLEs on large inputs. Notes This OA is associated with a part-time AI trainer role — the work is rating LLM-generated code, not building product. The TypeScript multiple-choice section reflects that. Camera-on monitoring is enforced; no external assistance is allowed. Parsing is part of the grade — the comma-vs-space-separated rows in the prompt are a deliberate trap; the actual stdin uses spaces. Preparation Implement 2-D Fenwick tree from scratch; expect to need update(r, c, delta) and query(r1, c1, r2, c2) with inclusion-exclusion. Drill clean stdin parsing in Python (sys.stdin.read().split()) — the OA's I/O harness is unforgiving on stray newlines. For the TypeScript multiple-choice portion, brush up on idiomatic vs anti-pattern code (mutation-in-map, missing await, untyped any, missing error handling).