go-schema-processor

General description:

Library goal is to create claim data slots according to the claim specification (https://idocs.iden3.io/#/core/spec/spec)

We use a common approach to describe data for Iden3 credentials by utilizing the concept of JSON-LD and JSON schemas.

Repository of claim schema vocabulary: https://github.com/iden3/claim-schema-vocab

The library includes three main components of any processor:

1. Schema Loaders
2. Data Validators
3. Data Parsers

Schema loader's purpose is to load schema (JSON / JSON-LD) from a given address.

Implemented loaders:

• HTTP loaders
• IPFS loader

Schema examples:

JSON-ld:

• Example of JSON-LD schema for KYC

  {
    "@context": [
      {
        "@version": 1.1,
        "@protected": true,
        "id": "@id",
        "type": "@type",
        "KYCCountryOfResidenceCredential": {
          "@id": "https://raw.githubusercontent.com/iden3/claim-schema-vocab/main/schemas/json-ld/kyc.json-ld#KYCCountryOfResidenceCredential",
          "@context": {
            "@version": 1.1,
            "@protected": true,
            "id": "@id",
            "type": "@type",
            "kyc-vocab": "https://github.com/iden3/claim-schema-vocab/blob/main/credentials/kyc.md#",
            "serialization": "https://github.com/iden3/claim-schema-vocab/blob/main/credentials/serialization.md#",
            "countryCode": {
              "@id": "kyc-vocab:countryCode",
              "@type": "serialization:IndexDataSlotA"
            },
            "documentType": {
              "@id": "kyc-vocab:documentType",
              "@type": "serialization:IndexDataSlotB"
            }
          }
        }
      }
    ]
  }
  

Each schema defines the type of credential e.g. KYCCountryOfResidenceCredential

KYCCountryOfResidenceCredential has next fields : countryCode and documentType

Type of countryCode is determined by the "serialization:IndexDataSlotA" field.

That means, claim processor library will process data with property countryCode and will put it to the Index slot (A) of the claim.

JSON schemas

• Example of JSON schema for KYC

  {
    "$schema": "http://json-schema.org/draft-07/schema#",
    "type": "object",
    "properties": {
      "countryCode": {
        "type": "integer"
      },
      "documentType": {
        "type": "integer"
      },
      "index": {
        "type": "array",
        "default":["countryCode","documentType"]
      },
      "value": {
        "type": "array",
        "default":[]
      }
    },
    "required": [
      "countryCode","documentType"
    ]
  }
  

From this schema, JSON processor will extract fields name for Index slots ("countryCode","documentType") and for Value slots (none)

Validators:

Validators implement two methods: ValidateData and ValidateDocument

Their purpose is to restrict possible invalid data to be processed by the parser.

Parsers:

The parser is the main part of this library. There are two implementations: JSON and JSON-LD

For each parser two parsing strategies exist: OneFieldPerSlotStrategy and SlotFullfilmentStrategy

If the parser is initialized with OneFieldPerSlotStrategy it will assign only one field from data to the claim a slot, in the case of SlotFullfilmentStrategy all capacity of the claim slot will be used (for several fields).

For JSON-LD parser claim type is required.

Examples:

loader := loaders.HTTP{}
validator := json.Validator{}
parser := jsonld.Parser{ClaimType: "KYCAgeCredential", ParsingStrategy: processor.OneFieldPerSlotStrategy}

jsonLdProcessor := New(processor.WithValidator(validator), processor.WithParser(parser), processor.WithSchemaLoader(loader))

schema, _, err := jsonLdProcessor.Load(url) // Load schema

data := make(map[string]interface{})
data["birthday"] = 828522341
data["documentType"] = 1
dataBytes, err := commonJSON.Marshal(data)
assert.Nil(t, err)

err = p.ValidateData(dataBytes, schema)  // Validation of data
parsedData, err := jsonLdProcessor.ParseSlots(dataBytes, schema) // parsing data

As result, output of processor will be 4 claim slots

type ParsedSlots struct {
	IndexA, IndexB []byte
	ValueA, ValueB []byte
}