How should I practice Effective Access Control?

Start by reading the full statement, solve the visible examples, then compare the pattern against related Snowflake phone screen problems in FastPrep.

Is Effective Access Control associated with Snowflake?

FastPrep tags this as a Snowflake phone screen coding practice problem when the problem catalog has that company and stage context.

Problem · Graph

Effective Access Control

Learn this problem

● Medium

SnowflakeFULLTIMEPHONE SCREEN

See Snowflake hiring insights

Problem statement

You are given n access-control nodes numbered from 0 through n - 1. The arrays allowLists and denyLists have one row per node. Each row contains the permission names that the node directly allows or denies.

Each row [parent, child] in edges means that child inherits access-control entries from parent. The directed graph is acyclic, and a node inherits entries from every ancestor reachable through any parent.

For each node, combine all local and inherited allow entries and all local and inherited deny entries. A deny entry overrides every allow entry for the same permission, regardless of whether either entry is local or inherited. Therefore, a permission is effective exactly when it appears in the combined allow set and does not appear in the combined deny set.

Return one permission list per node in node-index order. Remove duplicates and sort each node's effective permission names lexicographically.

Function

getEffectiveAccess(allowLists: String[][], denyLists: String[][], edges: int[][]) → String[][]

Examples

Example 1

allowLists = [["READ"],["WRITE"],["DEPLOY"]]denyLists = [[],["READ"],[]]edges = [[0,1],[1,2]]return = [["READ"],["WRITE"],["DEPLOY","WRITE"]]

Node 1 inherits READ but directly denies it. Node 2 inherits both the allow and the deny, so READ remains excluded while WRITE and DEPLOY are effective.

Example 2

allowLists = [["READ"],["WRITE"],["AUDIT"],["DEPLOY"]]denyLists = [[],[],["WRITE"],["READ"]]edges = [[0,2],[1,2],[2,3]]return = [["READ"],["WRITE"],["AUDIT","READ"],["AUDIT","DEPLOY"]]

Node 2 inherits from two parents and denies inherited WRITE. Node 3 inherits that deny and directly denies READ, leaving AUDIT and DEPLOY.

Example 3

allowLists = [["A","B"],["B","C"],[],["D"]]denyLists = [["B"],[],["C"],[]]edges = [[0,2],[1,2],[2,3]]return = [["A"],["B","C"],["A"],["A","D"]]

At node 2, inherited deny B and local deny C override all matching allows. Both denies continue to node 3.

Constraints

1 <= allowLists.length = denyLists.length <= 500.
0 <= edges.length <= 5000.
Every edge is [parent, child] with valid, distinct node indices.
The inheritance graph is a directed acyclic graph.
Every permission name is a non-empty ASCII token of at most 30 characters and contains no whitespace.
The total number of local allow and deny entries is at most 20000.
There are at most 2000 distinct permission names, and the total number of returned permission entries is at most 200000.