Coding Interview Patterns

The Go you need for coding interviews: slices, maps, heaps, runes and common gotchas.

Estimate time and space cost, and know what complexity the input size demands.

Solve a problem with a smaller copy of itself: base case, call stack and why recursion can repeat work.

Trade memory for speed: look things up in O(1) instead of scanning again.

Walk two indexes through sorted or symmetric data instead of nesting loops.

Keep a moving range valid and update it incrementally instead of recomputing.

Precompute running totals so any range sum is a subtraction.

Last in, first out: matching pairs, evaluating expressions, undo and a running minimum.

Use a stack to remember unresolved items; keep it ordered to find next greater/smaller.

Halve the search space whenever a yes/no condition is monotonic.

Rewire pointers carefully; use a fast and slow pointer to find cycles and middles.

Sort by start, then sweep and merge overlapping ranges.

Recurse on subtrees, or sweep level by level with a queue.

Keep only the best K items; the heap tells you the worst of them in O(log k).

Model relationships as nodes and edges; traverse with a visited set.

Build answers one choice at a time; undo the choice and try the next.

Break a problem into overlapping subproblems and solve each only once.

Grids, two strings and ranges: when a subproblem needs two numbers, fill a table cell by cell.

Take the best local choice and never undo it, when you can argue it is safe.

Track which items are in the same group and merge groups in near-constant time.

Weighted graphs: cheapest paths with a heap, cheapest ways to connect everything, and ordering by rules.

Store words letter by letter so any prefix question is a short walk down a tree.

XOR tricks, bit counting, in-place matrix moves, fast exponentiation and digit-by-digit arithmetic.