Schema JSON Reference

This page documents the JSON format accepted by TokenSchemasBuilder.fromJson in phantasma-sdk-ts.

JSON shape

{
  "seriesMetadata": [
    { "name": "name", "type": "String" }
  ],
  "rom": [
    { "name": "artist", "type": "String" }
  ],
  "ram": [
    { "name": "level", "type": "Int32" }
  ]
}

Rules:

• seriesMetadata, rom, and ram must be JSON arrays (they can be empty).

• Each entry must be an object with name (string) and type (string).

• type is a VmType name string; see the list below.

• Field names are case-sensitive.

Accepted type strings

The type value must be one of the strings accepted by vmTypeFromString:

• Dynamic, Array

• Bytes, Bytes16, Bytes32, Bytes64

• Struct

• Int8, Int16, Int32, Int64, Int256

• String

• Array_Dynamic, Array_Bytes, Array_Struct

• Array_Int8, Array_Int16, Array_Int32, Array_Int64, Array_Int256

• Array_Bytes16, Array_Bytes32, Array_Bytes64

• Array_String

Notes:

• The JSON arrays above describe schema fields. They are not related to Array_* type strings.

• Array_* types are field types; use them when a field itself stores an array.

Default/system fields

TokenSchemasBuilder always adds default fields, so you should not include them in JSON:

• Series metadata: _i, mode, rom.

• NFT ROM schema: _i, rom.

Standard NFT fields requirement

When using fromJson, the builder verifies that the standard NFT fields exist in either seriesMetadata or the NFT rom schema:

• name

• description

• imageURL

• infoURL

• royalties

RAM and DynamicExtras

If ram is an empty array, the SDK sets the RAM schema flag DynamicExtras. This causes extra RAM fields (not defined in the schema) to be serialized. When you define RAM fields, DynamicExtras is not set, so only schema-defined RAM fields are serialized. This means any extra RAM fields you include are dropped and cannot be read or updated later.