How should I practice Feature Flag Evaluation?

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

Is Feature Flag Evaluation associated with CVS Health?

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

Feature Flag Evaluation

CVS Health● EasyFULLTIMEPHONE SCREEN

Problem statement

A feature-flag system decides whether a feature is on for a user. Flag i is named flagNames[i] and has three settings:

enabled[i]: a master switch. When it is false, the feature is off for everyone.
allowLists[i]: user ids that always get the feature while the switch is on.
rolloutPercents[i]: the percentage of other users who get it. A user's bucket is userId % 100, and the user gets the feature when the bucket is less than the percentage.

For each query, decide whether flag queryFlags[j] is on for user queryUsers[j]. Check the rules in this order: a flag name that does not exist is off; a switched-off flag is off; a user on the allow list is on; otherwise the bucket rule decides.

Return the answers in query order.

Function

evaluateFeatureFlags(flagNames: String[], enabled: boolean[], rolloutPercents: int[], allowLists: int[][], queryFlags: String[], queryUsers: int[]) → boolean[]

Examples

Example 1

flagNames = ["new_checkout","dark_mode"]enabled = [true,false]rolloutPercents = [30,100]allowLists = [[7],[5]]queryFlags = ["new_checkout","new_checkout","new_checkout","dark_mode","beta_search"]queryUsers = [1229,1250,7,5,1]return = [true,false,true,false,false]

new_checkout is on with 30% rollout. User 1229 is in bucket 1229 % 100 = 29, which is below 30, so it is on. User 1250 is in bucket 50, so it is off. User 7 is on the allow list, so it is on, and bucket 7 would also pass. dark_mode is turned off, so it is off for everyone, even user 5 on its allow list. beta_search does not exist, so it is off.

Example 2

flagNames = ["pilot"]enabled = [true]rolloutPercents = [0]allowLists = [[42,1000000000]]queryFlags = ["pilot","pilot","pilot"]queryUsers = [42,43,1000000000]return = [true,false,true]

With 0% rollout, only the allow-listed users 42 and 1000000000 get the feature. User 43 does not.

Constraints

1 <= flagNames.length <= 50, and enabled, rolloutPercents and allowLists have the same length.
Flag names (in flagNames and queryFlags) are 1 to 20 lowercase English letters or underscores, and the names in flagNames are all different.
0 <= rolloutPercents[i] <= 100
Each allow list has at most 20 different user ids.
User ids are integers from 0 to 10^9.
1 <= queryFlags.length <= 1000 and queryUsers.length == queryFlags.length