goschema

Define data-structure validation in go code. Can be used to validate data from json/yaml files. Uses https://github.com/xeipuuv/gojsonschema to do the actual validation.

usage

It's implemented in a fluent interface style:

personSchema := NewObjectType("Example person schema", func(p ObjectType) {
	p.String("firstName", "")
	p.String("lastName", "").Optional()
	p.Int("age", "Age in years").Min(0).Max(5).Optional()
	p.Enum("gender", "").Add("male", "A male specimen").Add("female", "A female specimen")
})
// person must be a map with string keys
// it can also be loaded from a json file with json.Unmarshal(data, mapPointer)
goschema.Validate(personSchema, person)

The validation is executed by generating a jsonschema, which is then validated against the datastructure:

{
  "additionalProperties":false,
  "properties":{
    "age":{
      "description":"Age in years",
      "maximum":5,
      "minimum":0,
      "type":"integer"
    },
    "firstName":{
      "type":"string"
    },
    "gender":{
      "enum":[
        "male",
        "female"
      ],
      "type":"string"
    },
    "lastName":{
      "type":"string"
    }
  },
  "required":[
    "firstName",
    "gender"
  ],
  "title":"Example Schema",
  "type":"object"
}

doc generation

Creating plain text output with documentation of the validation format in the sense of:

Example person schema
---------------------
age       //  optional, Age in years as int from 0 to 5
firstName // firstName as string
gender    // 
  male    // A male speciemen
  female  // A female speciemen
lastName  //  optional, lastName as string

can be done with

str := goschema.Doc(personSchema)

This data can be extracted from the defined values above including the help string which is usually the second parameter in data-type definition calls