JSON Schema を生成する方法

问题

JSON データがあり、バリデーションや API ドキュメントのために構造を記述するスキーマを生成したい。

前提条件

• JSON の理解
• JSON Schema の概念への習熟

ステップ

1. 1. Analyze the sample JSON

   The generator inspects the JSON to determine types, required fields, array item types, and nested object structures. It infers the schema from the data. Multiple samples can improve accuracy by detecting optional fields and enum values.

   入力

   {"name":"Alice","age":30}

   输出

   inferred structure

2. 2. Generate the schema

   Produce a JSON Schema document with type constraints for each field. Strings get type string, numbers get type number or integer, arrays get items with the element type, and objects get properties with nested schemas. Mark fields present in all samples as required.

   入力

   sample JSON

   输出

   JSON Schema draft 2020-12

3. 3. Refine and validate

   Review the generated schema and refine it: add constraints (minLength, maximum, pattern), identify optional vs required fields, and set enum values for fields with known options. Validate the original JSON against the schema to confirm it passes.

   入力

   schema + original JSON

   输出

   validation passes

期待される結果

サンプルデータを検証でき、ランタイムバリデーションや API ドキュメントに利用できる JSON Schema。

よくあるエラー

Schema is too permissive

A single sample cannot infer all constraints. Add more samples or manually refine the schema with tighter constraints.

Array item type is wrong

If the array contains mixed types, the schema may infer a union type. Provide more samples or specify the expected item type manually.

トラブルシューティング指引

• If validation fails against the generated schema, the schema may have over-constrained a field. Relax the constraint.
• Use multiple samples to detect optional fields and enum values.

セキュリティ上の制限

• Schema generation is a read-only analysis. No data is transmitted.