Have created following schema:
{
"type": "object",
"properties": {
"name": {
"type": "string",
"enum": [
"full",
"partial"
]
}
},
"required": [
"name"
],
"if": {
"properties": {
"name": {
"const": "full"
}
}
},
"then": {
"properties": {
"status": {
"type": "string",
"enum": [
"success",
"failure"
]
}
},
"required": [
"status"
]
},
"else": {
"properties": {
"status": {
"type": "string",
"enum": [
"success",
"failure",
"partial success"
]
},
"message": {
"type": "string"
},
"created": {
"type": "array",
"items": [
{
"type": "integer"
}
]
},
"deleted": {
"type": "array",
"items": [
{
"type": "integer"
}
]
}
},
"required": [
"name",
"status",
"created",
"deleted"
]
}
}
I am trying to have two kind of json such that based on key 'name', there will be different sub validations for keys - 'full' and 'partial'
So, two sample valid json possible are:
When name is 'full'
{"name": "full", "status": "success"}
When name is 'partial'
{
"name": "partial",
"status":"success",
"created": [6],
"deleted": [4]
}
On validating using this schema in python, it's not validating the part in if/then/else.
validator = Validator(json.load(open(path, 'r')))
validator.validate({"name": "full"})
[]
validator.validate({"name": "full", "status": "success"})
[]
It gives both of them as valid while first one should have been invalid.
Similarly for second json, it doesn't fail for invalid one:
validator.validate({"name": "partial"})
[]
validator.validate({"name": "partial", "stauts": "success", "created": [6], "deleted": [4]})
[]
Python validator code:
class Validator(object):
def __init__(self, schema):
self.schema = schema
def validate(self, json):
validator = Draft4Validator(self.schema)
errors = sorted(validator.iter_errors(json), key=str)
return errors
