File: reshape-the-matrix/solution.py

Date: 2026-06-06

Time: 18:52

reshape-the-matrix/solution.py

Purpose

This file solves LeetCode 566 - Reshape the Matrix. It owns the single responsibility of converting an m×n matrix into an r×c matrix by reading elements in row-major order, or returning the original matrix unchanged if the reshape is dimensionally impossible.

Key Components

Solution.matrixReshape(mat, r, c) — The sole public method. Contract:

• Input: mat (non-empty 2D list of ints), r (target rows), c (target columns)
• Output: A new r×c matrix if m*n == r*c, otherwise the original mat reference
• Purity: Does not mutate mat. Returns a new list-of-lists on success, or the same object on failure.

Patterns

Flatten-then-slice: The solution uses the classic 2D reshape idiom — flatten the matrix into a 1D list, then slice it into rows of width c. This is the same conceptual operation as numpy.reshape but implemented with pure list comprehensions.

The flatten step ([val for row in mat for val in row]) produces a row-major traversal. The rebuild step (flat[i*c:(i+1)*c]) partitions it back into chunks of size c.

Dependencies

Imports: None beyond builtins. The solution uses only Python list operations.

Imported by: reshape-the-matrix/test_solution.py directly. The "Imported By" list in the prompt is misleading — those ~400+ test files are unrelated problems that likely share a common test harness or conftest, not actual importers of this module's Solution class.

Flow

1. Extract dimensions m, n from the input matrix.

2. Guard: if total element count m*n != r*c, return mat unchanged — reshape is impossible.

3. Flatten mat into a 1D list flat via nested list comprehension.

4. Slice flat into r rows of c elements each via list comprehension with index arithmetic.

5. Return the new 2D list.

Invariants

• Element conservation: The reshape is only attempted when m*n == r*c, guaranteeing no elements are lost or invented.
• Row-major ordering: Elements are read left-to-right, top-to-bottom from the source and written in the same order to the target. This matches the problem's specification.
• No partial results: Either a complete r×c matrix is returned or the original matrix is returned untouched. There's no intermediate failure state.

Error Handling

There is none beyond the dimension check. The code assumes mat is non-empty and rectangular (guaranteed by LeetCode constraints). If mat were empty or jagged, len(mat[0]) would raise IndexError or produce incorrect results — but that's outside the problem's contract.

Beliefs

• reshape-returns-original-on-mismatch — When m*n != r*c, matrixReshape returns the exact same mat object (identity, not a copy)
• reshape-uses-O(mn)-space — The flatten step allocates a full copy of all elements, so space complexity is O(m·n) beyond the output
• reshape-preserves-row-major-order — Elements appear in the output in the same row-major order as the input, guaranteed by the nested list comprehension iteration order
• reshape-no-mutation — The input matrix mat is never modified; the result is always a freshly constructed list-of-lists (on the success path)