• 设为首页
  • 点击收藏
  • 手机版
    手机扫一扫访问
    迪恩网络手机版
  • 关注官方公众号
    微信扫一扫关注
    迪恩网络公众号

voxpupuli/json-schema: Ruby JSON Schema Validator

原作者: [db:作者] 来自: 网络 收藏 邀请

开源软件名称:

voxpupuli/json-schema

开源软件地址:

https://github.com/voxpupuli/json-schema

开源编程语言:

Ruby 100.0%

开源软件介绍:

Ruby JSON Schema Validator

License Test Release RubyGem Version RubyGem Downloads Donated by Iain Beeston

This library is intended to provide Ruby with an interface for validating JSON objects against a JSON schema conforming to JSON Schema Draft 4. Legacy support for JSON Schema Draft 3, JSON Schema Draft 2, and JSON Schema Draft 1 is also included.

Additional Resources

Version 2.0.0 Upgrade Notes

Please be aware that the upgrade to version 2.0.0 will use Draft-04 by default, so schemas that do not declare a validator using the $schema keyword will use Draft-04 now instead of Draft-03. This is the reason for the major version upgrade.

Version 3.0.0 Upgrade Notes

All individual changes are documented in the CHANGELOG.md. The biggest change is that the new version only supports Ruby 2.5 and newer. Take a look into the gemspec file to see the currently supported Ruby version and also .github/workflows/test.yml to see the Ruby versions we test on.

Installation

From rubygems.org:

gem install json-schema

From the git repo:

gem build json-schema.gemspec
gem install json-schema-*.gem

Validation

Three base validation methods exist:

  1. validate: returns a boolean on whether a validation attempt passes
  2. validate!: throws a JSON::Schema::ValidationError with an appropriate message/trace on where the validation failed
  3. fully_validate: builds an array of validation errors return when validation is complete

All methods take two arguments, which can be either a JSON string, a file containing JSON, or a Ruby object representing JSON data. The first argument to these methods is always the schema, the second is always the data to validate. An optional third options argument is also accepted; available options are used in the examples below.

By default, the validator uses the JSON Schema Draft 4 specification for validation; however, the user is free to specify additional specifications or extend existing ones. Legacy support for Draft 1, Draft 2, and Draft 3 is included by either passing an optional :version parameter to the validate method (set either as :draft1 or draft2), or by declaring the $schema attribute in the schema and referencing the appropriate specification URI. Note that the $schema attribute takes precedence over the :version option during parsing and validation.

For further information on json schema itself refer to Understanding JSON Schema.

Basic Usage

require "json-schema"

schema = {
  "type" => "object",
  "required" => ["a"],
  "properties" => {
    "a" => {"type" => "integer"}
  }
}

#
# validate ruby objects against a ruby schema
#

# => true
JSON::Validator.validate(schema, { "a" => 5 })
# => false
JSON::Validator.validate(schema, {})

#
# validate a json string against a json schema file
#

require "json"
File.write("schema.json", JSON.dump(schema))

# => true
JSON::Validator.validate('schema.json', '{ "a": 5 }')

#
# raise an error when validation fails
#

# => "The property '#/a' of type String did not match the following type: integer"
begin
  JSON::Validator.validate!(schema, { "a" => "taco" })
rescue JSON::Schema::ValidationError => e
  e.message
end

#
# return an array of error messages when validation fails
#

# => ["The property '#/a' of type String did not match the following type: integer in schema 18a1ffbb-4681-5b00-bd15-2c76aee4b28f"]
JSON::Validator.fully_validate(schema, { "a" => "taco" })

Advanced Options

require "json-schema"

schema = {
  "type"=>"object",
  "required" => ["a"],
  "properties" => {
    "a" => {
      "type" => "integer",
      "default" => 42
    },
    "b" => {
      "type" => "object",
      "properties" => {
        "x" => {
          "type" => "integer"
        }
      }
    }
  }
}

#
# with the `:list` option, a list can be validated against a schema that represents the individual objects
#

# => true
JSON::Validator.validate(schema, [{"a" => 1}, {"a" => 2}, {"a" => 3}], :list => true)
# => false
JSON::Validator.validate(schema, [{"a" => 1}, {"a" => 2}, {"a" => 3}])

#
# with the `:errors_as_objects` option, `#fully_validate` returns errors as hashes instead of strings
#

# => [{:schema=>#<Addressable::URI:0x3ffa69cbeed8 URI:18a1ffbb-4681-5b00-bd15-2c76aee4b28f>, :fragment=>"#/a", :message=>"The property '#/a' of type String did not match the following type: integer in schema 18a1ffbb-4681-5b00-bd15-2c76aee4b28f", :failed_attribute=>"TypeV4"}]
JSON::Validator.fully_validate(schema, { "a" => "taco" }, :errors_as_objects => true)

#
# with the `:strict` option, all properties are considered to have `"required": true` and all objects `"additionalProperties": false`
#

# => true
JSON::Validator.validate(schema, { "a" => 1, "b" => { "x" => 2 } }, :strict => true)
# => false
JSON::Validator.validate(schema, { "a" => 1, "b" => { "x" => 2 }, "c" => 3 }, :strict => true)
# => false
JSON::Validator.validate(schema, { "a" => 1 }, :strict => true)

#
# with the `:fragment` option, only a fragment of the schema is used for validation
#

# => true
JSON::Validator.validate(schema, { "x" => 1 }, :fragment => "#/properties/b")
# => false
JSON::Validator.validate(schema, { "x" => 1 })

#
# with the `:validate_schema` option, the schema is validated (against the json schema spec) before the json is validated (against the specified schema)
#

# => true
JSON::Validator.validate(schema, { "a" => 1 }, :validate_schema => true)
# => false
JSON::Validator.validate({ "required" => true }, { "a" => 1 }, :validate_schema => true)

#
# with the `:insert_defaults` option, any undefined values in the json that have a default in the schema are replaced with the default before validation
#

# => true
JSON::Validator.validate(schema, {}, :insert_defaults => true)
# => false
JSON::Validator.validate(schema, {})

#
# with the `:version` option, schemas conforming to older drafts of the json schema spec can be used
#

v2_schema = {
  "type" => "object",
  "properties" => {
    "a" => {
      "type" => "integer"
    }
  }
}

# => false
JSON::Validator.validate(v2_schema, {}, :version => :draft2)
# => true
JSON::Validator.validate(v2_schema, {})

#
# with the `:parse_data` option set to false, the json must be a parsed ruby object (not a json text, a uri or a file path)
#

# => true
JSON::Validator.validate(schema, { "a" => 1 }, :parse_data => false)
# => false
JSON::Validator.validate(schema, '{ "a": 1 }', :parse_data => false)

#
# with the `:json` option, the json must be an unparsed json text (not a hash, a uri or a file path)
#

# => true
JSON::Validator.validate(schema, '{ "a": 1 }', :json => true)
# => "no implicit conversion of Hash into String"
begin
  JSON::Validator.validate(schema, { "a" => 1 }, :json => true)
rescue TypeError => e
  e.message
end

#
# with the `:uri` option, the json must be a uri or file path (not a hash or a json text)
#

File.write("data.json", '{ "a": 1 }')

# => true
JSON::Validator.validate(schema, "data.json", :uri => true)
# => "Can't convert Hash into String."
begin
  JSON::Validator.validate(schema, { "a"  => 1 }, :uri => true)
rescue TypeError => e
  e.message
end

#
# with the `:clear_cache` option set to true, the internal cache of schemas is
# cleared after validation (otherwise schemas are cached for efficiency)
#

File.write("schema.json", v2_schema.to_json)

# => true
JSON::Validator.validate("schema.json", {})

File.write("schema.json", schema.to_json)

# => true
JSON::Validator.validate("schema.json", {}, :clear_cache => true)

# => false
JSON::Validator.validate("schema.json", {})

Extending Schemas

For this example, we are going to extend the JSON Schema Draft 3 specification by adding a 'bitwise-and' property for validation.

require "json-schema"

class BitwiseAndAttribute < JSON::Schema::Attribute
  def self.validate(current_schema, data, fragments, processor, validator, options = {})
    if data.is_a?(Integer) && data & current_schema.schema['bitwise-and'].to_i == 0
      message = "The property '#{build_fragment(fragments)}' did not evaluate  to true when bitwise-AND'd with  #{current_schema.schema['bitwise-or']}"
      validation_error(processor, message, fragments, current_schema, self, options[:record_errors])
    end
  end
end

class ExtendedSchema < JSON::Schema::Draft3
  def initialize
    super
    @attributes["bitwise-and"] = BitwiseAndAttribute
    @uri = JSON::Util::URI.parse("http://test.com/test.json")
    @names = ["http://test.com/test.json"]
  end

  JSON::Validator.register_validator(self.new)
end

schema = {
  "$schema" => "http://test.com/test.json",
  "properties" => {
    "a" => {
      "bitwise-and" => 1
    },
    "b" => {
      "type" => "string"
    }
  }
}

data = {
  "a" => 0
}

data = {"a" => 1, "b" => "taco"}
JSON::Validator.validate(schema,data) # => true
data = {"a" => 1, "b" => 5}
JSON::Validator.validate(schema,data) # => false
data = {"a" => 0, "b" => "taco"}
JSON::Validator.validate(schema,data) # => false

Custom format validation

The JSON schema standard allows custom formats in schema definitions which should be ignored by validators that do not support them. JSON::Schema allows registering procs as custom format validators which receive the value to be checked as parameter and must raise a JSON::Schema::CustomFormatError to indicate a format violation. The error message will be prepended by the property name, e.g. The property '#a'

require "json-schema"

format_proc = -> value {
  raise JSON::Schema::CustomFormatError.new("must be 42") unless value == "42"
}

# register the proc for format 'the-answer' for draft4 schema
JSON::Validator.register_format_validator("the-answer", format_proc, ["draft4"])

# omitting the version parameter uses ["draft1", "draft2", "draft3", "draft4"] as default
JSON::Validator.register_format_validator("the-answer", format_proc)

# deregistering the custom validator
# (also ["draft1", "draft2", "draft3", "draft4"] as default version)
JSON::Validator.deregister_format_validator('the-answer', ["draft4"])

# shortcut to restore the default formats for validators (same default as before)
JSON::Validator.restore_default_formats(["draft4"])

# with the validator registered as above, the following results in
# ["The property '#a' must be 42"] as returned errors
schema = {
  "$schema" => "http://json-schema.org/draft-04/schema#",
  "properties" => {
    "a" => {
      "type" => "string",
      "format" => "the-answer",
    }
  }
}
errors = JSON::Validator.fully_validate(schema, {"a" => "23"})

Validating a JSON Schema

To validate that a JSON Schema conforms to the JSON Schema standard, you need to validate your schema against the metaschema for the appropriate JSON Schema Draft. All of the normal validation methods can be used for this. First retrieve the appropriate metaschema from the internal cache (using JSON::Validator.validator_for_name() or JSON::Validator.validator_for_uri()) and then simply validate your schema against it.

require "json-schema"

schema = {
  "type" => "object",
  "properties" => {
    "a" => {"type" => "integer"}
  }
}

metaschema = JSON::Validator.validator_for_name("draft4").metaschema
# => true
JSON::Validator.validate(metaschema, schema)

Controlling Remote Schema Reading

In some cases, you may wish to prevent the JSON Schema library from making HTTP calls or reading local files in order to resolve $ref schemas. If you fully control all schemas which should be used by validation, this could be accomplished by registering all referenced schemas with the validator in advance:

schema = JSON::Schema.new(some_schema_definition, Addressable::URI.parse('http://example.com/my-schema'))
JSON::Validator.add_schema(schema)

If more extensive control is necessary, the JSON::Schema::Reader instance used can be configured in a few ways:

# Change the default schema reader used
JSON::Validator.schema_reader = JSON::Schema::Reader.new(:accept_uri => true, :accept_file => false)

# For this validation call, use a reader which only accepts URIs from my-website.com
schema_reader = JSON::Schema 

鲜花

握手

雷人

路过

鸡蛋
该文章已有0人参与评论

请发表评论

全部评论

专题导读
热门推荐
阅读排行榜

扫描微信二维码

查看手机版网站

随时了解更新最新资讯

139-2527-9053

在线客服(服务时间 9:00~18:00)

在线QQ客服
地址:深圳市南山区西丽大学城创智工业园
电邮:jeky_zhao#qq.com
移动电话:139-2527-9053

Powered by 互联科技 X3.4© 2001-2213 极客世界.|Sitemap