Find Repeated-Value Patterns in a Matrix

Find every placement of a small symbol matrix inside a larger integer matrix that preserves the pattern of equal and different values.

Implement find_pattern_positions(grid: int[][], pattern: string[]) -> int[][]. Each pattern string is one row of uppercase letter symbols. Return the zero-based top-left coordinate of every matching window, ordered by row and then column.

Matching contract

For each candidate window independently, there must be a one-to-one mapping between symbols and integer values:

• Every occurrence of the same symbol must match the same integer.
• Different symbols must match different integers.
• The mapping can differ between windows.

The report illustrates AT/TB matching 1,2/2,3 but does not explicitly resolve whether different symbols may share a value. This practice version chooses the one-to-one interpretation above.

Constraints & Assumptions

• The grid is rectangular, with 1 through 50 rows and columns.
• The pattern is rectangular, with 1 through 5 rows and columns, and symbols A through Z.
• Grid values are integers between -1,000,000,000 and 1,000,000,000.
• A window must fit fully inside the grid. If the pattern is larger in either dimension, return an empty list.
• Do not rotate, reflect, or resize the pattern.

Examples

grid = [[1,2,4,5],[2,3,6,7],[6,4,2,2]]
pattern = ["AT","TB"]
result = [[0,0]]

grid = [[7,7],[7,7]]
pattern = ["AA","AA"]
result = [[0,0]]

The first example's upper-left window maps A -> 1, T -> 2, and B -> 3.