多态数组对象JSON Schema验证异常：未检测必填字段缺失

使用com.networknt json-schema-validator（版本1.0.79）验证包含多态类型车辆数组的JSON数据时，发现当bicycle类型的对象缺失必填的petName字段时，未触发预期的验证错误。

使用的依赖 #

<dependency>
    <groupId>com.networknt</groupId>
    <artifactId>json-schema-validator</artifactId>
    <version>1.0.79</version>
    <exclusions>
        <exclusion>
            <groupId>org.apache.commons</groupId>
            <artifactId>commons-lang3</artifactId>
        </exclusion>
    </exclusions>
</dependency>

测试JSON数据 #

{
  "firstname" : "Jack",
  "lastname" : "Sparrow",
  "age" : "42",
  "vehicles" : [
    {
      "type" : "bicycle",
      "company" : "Hercules",
      "isElectric" : false
    },
    {
      "type" : "motorcycle",
      "company" : "Hero",
      "hp" : 340,
      "isRich" : true
    }
  ]
}

原JSON Schema #

{
    "$schema": "http://json-schema.org/draft-06/schema#",
    "type": "object",
    "required": [
        "firstname",
        "lastname",
        "age",
        "vehicles"
    ],
    "properties": {
        "firstname": {
            "type": "string",
            "minLength": 2,
            "maxLength": 10
        },
        "lastname": {
            "type": "string"
        },
        "age": {
            "type": [
                "integer",
                "string"
            ]
        },
        "vehicles": {
            "type": "array",
            "items": {
                "type": "object",
                "required": [
                    "type"
                ],
                "properties": {
                    "company": {
                        "type": "string",
                        "minLength": 2,
                        "maxLength": 10
                    },
                    "isElectric": {
                        "type": "boolean"
                    },
                    "hp": {
                        "type": [
                            "string",
                            "integer"
                        ]
                    },
                    "isRich": {
                        "type": "boolean"
                    },
                    "type": {
                        "type": "string",
                        "enum": [
                            "bicycle",
                            "motorcycle"
                        ]
                    }
                },
                "additionalProperties": false
            },
            "anyOf": [
                {
                    "properties": {
                        "type": {
                            "const": "bicycle"
                        },
                        "vehicles": {
                            "items": {
                                "required": [
                                    "company",
                                    "isElectric",
                                    "petName"
                                ]
                            }
                        }
                    }
                },
                {
                    "properties": {
                        "type": {
                            "const": "motorcycle"
                        },
                        "vehicles": {
                            "items": {
                                "required": [
                                    "company",
                                    "hp",
                                    "isRich"
                                ]
                            }
                        }
                    }
                }
            ]
        },
        "additionalProperties": false
    }
}

问题原因分析 #

原Schema存在两个核心错误：

• 层级错误：将anyOf放在了vehicles数组的层级下，但type是数组中每个车辆对象的属性，不是数组本身的属性，导致校验规则无法匹配到数组内的元素。
• 规则错位：anyOf分支内错误地再次定义vehicles属性，完全偏离了针对单个车辆对象校验的逻辑，相当于试图给整个数组重复定义规则，而非针对数组中的每个元素。

修正后的JSON Schema #

正确的做法是在数组的items中使用oneOf（或anyOf），针对每个车辆对象根据type字段区分校验规则：

{
    "$schema": "http://json-schema.org/draft-06/schema#",
    "type": "object",
    "required": [
        "firstname",
        "lastname",
        "age",
        "vehicles"
    ],
    "properties": {
        "firstname": {
            "type": "string",
            "minLength": 2,
            "maxLength": 10
        },
        "lastname": {
            "type": "string"
        },
        "age": {
            "type": [
                "integer",
                "string"
            ]
        },
        "vehicles": {
            "type": "array",
            "items": {
                "type": "object",
                "required": ["type"],
                "properties": {
                    "type": {
                        "type": "string",
                        "enum": ["bicycle", "motorcycle"]
                    },
                    "company": {
                        "type": "string",
                        "minLength": 2,
                        "maxLength": 10
                    }
                },
                "additionalProperties": false,
                "oneOf": [
                    {
                        "properties": {
                            "type": {"const": "bicycle"},
                            "isElectric": {"type": "boolean"},
                            "petName": {"type": "string"}
                        },
                        "required": ["company", "isElectric", "petName"]
                    },
                    {
                        "properties": {
                            "type": {"const": "motorcycle"},
                            "hp": {"type": ["string", "integer"]},
                            "isRich": {"type": "boolean"}
                        },
                        "required": ["company", "hp", "isRich"]
                    }
                ]
            }
        },
        "additionalProperties": false
    }
}

修正说明 #

• 将oneOf移动到items内部，确保每个数组元素（车辆对象）都会被校验。
• 每个oneOf分支针对特定type的车辆定义专属的必填字段和属性规则，确保bicycle必须包含petName，motorcycle必须包含hp和isRich。
• 公共属性（如company）提取到外层properties中，避免重复定义。

内容的提问来源于stack exchange，提问作者pjsagar