Topic-wise pattern references: recognition cues, the core template, key variations, and the pitfalls that cost interviews.
Hash maps, sets, prefix sums, and frequency counting — the pattern behind a third of all easy/medium problems.
Converging, parallel, and fast/slow pointers — O(n) scans that replace nested loops on sorted or structured data.
Fixed and variable windows over strings and arrays — longest/shortest substring problems in O(n).
The bulletproof template, boundary (first/last) searches, rotated arrays, and binary search on the answer space.
Matching, simulation, expression evaluation, and the monotonic stack behind every "next greater" problem.
Dummy heads, fast/slow pointers, reversal — the small toolkit that solves every list problem.
DFS recursion shapes, BFS levels, BST invariants, and the global-variable trick for path problems.
Prefix trees for autocomplete, dictionaries, and wildcard search — when a hash set of words isn't enough.
Size-k heaps, two-heap medians, and when bucket sort or quickselect beats the heap.
Sort-then-sweep — merging, inserting, and scheduling ranges without off-by-one wounds.
The choose–explore–unchoose template for subsets, permutations, combinations, and constraint search.
BFS, DFS, topological sort, union-find, Dijkstra — and how to pick between them in ten seconds.
State, transition, base case — the five DP families and a recipe for deriving each from brute force.
When the locally best choice is globally safe — recognizing greedy problems and defending the choice out loud.
The XOR family, bit tricks worth memorizing, and adding without the plus sign.