Data Selection — JavaScript Challenge | FineFrontend

Practice workspaceSign in

JavaScript/Data Selectionhard

00:00

Read here. Code on desktop. The coding workspace opens on screens 1024px and wider.

JavaScript interview question

Data Selection

hard30 min

Implement dataSelection(rows, requirement), a small filtering engine for an array of records. A requirement can test a field or combine other requirements with nested AND, OR, and NOT groups.

Use { all: [...] } for AND, { any: [...] } for OR, and { not: requirement } for negation. A field rule has { field, operator, value }. For example, a dashboard could select active developers whose experience is at least two years.

Requirements

Export dataSelection as both a named and default export. Accept a dense array of records and one requirement tree; return a new array of matching rows in their original order, preserving row references.
A field rule has exactly field, operator, and value. field is a non-empty dotted path such as profile.age or scores.0. Traverse only own properties of objects or arrays; a missing segment must not throw.
Support eq and neq using Object.is equality, and in using Object.is against each value in an array. For this exercise equality values are primitives. There is no coercion; NaN equals NaN and negative zero differs from positive zero.
Support gt, gte, lt, and lte for number values only. Their rule value must be a number other than NaN; a row with a non-number or NaN does not satisfy an ordered comparison.
Support exists with a boolean rule value. An own field containing null or undefined still exists. Missing fields fail all other field operators, including neq.
Groups contain exactly one key: all with an array of requirements, any with an array of requirements, or not with one requirement. Groups may nest arbitrarily. Empty all matches every row; empty any matches none.
Validate the complete requirement tree before filtering, including when rows is empty or a branch would short-circuit. Throw TypeError for malformed groups, empty path segments, unknown operators, missing rule fields, or invalid in/exists/ordered-comparison values.
Do not modify rows or requirements. Inputs are acyclic data records with no accessors. Sorting, projection, escaped dots in field names, and SQL parsing are outside scope.

Examples

Example 1

Input: const rows = [{ id: 1, active: true, profile: { years: 3 } }, { id: 2, active: false, profile: { years: 5 } }];

dataSelection(rows, { all: [
  { field: "active", operator: "eq", value: true },
  { field: "profile.years", operator: "gte", value: 2 },
] });
Output: [{ id: 1, active: true, profile: { years: 3 } }]

Example 2

Input: dataSelection([{ id: 1 }, { id: 2, note: undefined }], { field: "note", operator: "exists", value: false });
Output: [{ id: 1 }]

Example 3

Input: dataSelection([{ id: 1 }], { any: [] });
Output: []

Use the same documented API in both language variants.

What the tests check

Filters without coercion and preserves row references and order
Rejects missing fields for neq
Supports all four ordered comparison operators
Uses Object.is equality for NaN and signed zero
Supports in lists including an empty list
Distinguishes missing properties from null and undefined
Reads nested fields and array indices safely
Ignores inherited properties while supporting special own keys
Combines nested all, any, and not groups
Handles empty groups and empty row arrays
Validates malformed rules even when no rows are provided
Validates operator-specific values and every nested branch
Does not mutate frozen rows or requirements

References

Array.prototype.filter()Object.is()

Data FilteringRecursionPredicate CompositionValidation

Back

Preparing your workspace…