Word Search II with Trie

Find words from a dictionary inside a 2D character board. The expected solution builds a trie over the word list, then runs backtracking DFS from each board cell with in-place visited marking. Requirements Given a 2D board of characters and a list of words, return the words that can be formed by walking adjacent cells on the board. A cell may be used at most once in one word path. def findWords(board: List[List[str]], words: List[str]) -> List[str]: ... Expected approach: Build a trie from words. Start DFS from every board cell whose character exists at the trie root. Move in four directions while following trie edges. Deduplicate found words and avoid revisiting a board cell in the same path. Run at least one self-built test case in the coding pad. Notes Trie pruning is the main signal; a per-word board search is usually too slow. Mark visited cells in-place or with a set, but restore state on backtrack. Avoid returning duplicates when the same word can be found through multiple paths. The prompt was paired with a senior-level loop, so expect follow-ups on advanced data structures and runtime trade-offs. Preparation Write the trie node structure from scratch without importing a specialized library. Drill board DFS with backtracking and state restoration. Prepare a small handmade test board and word list so you can prove the code in the interview editor.