JSON Schema validation and resolution library for Squiz DX platform, with specialized support for primitive types like `SquizLink` and `SquizImage`.
JSON Schema validation and resolution library for Squiz DX platform, with specialized support for primitive types like SquizLink and SquizImage.
- ✅ JSON Schema validation and resolution
- ✅ Support for custom primitive types (SquizLink, SquizImage, FormattedText)
- ✅ AllOf conditional schema support with data preservation
- ✅ Resolvable types integration (Matrix Assets, etc.)
- ✅ High performance with comprehensive test coverage
``typescript
import { JSONSchemaService } from '@squiz/dx-json-schema-lib';
import { TypeResolverBuilder } from '@squiz/dx-json-schema-lib/jsonTypeResolution';
import { SquizLinkType, SquizImageType } from '@squiz/dx-json-schema-lib/primitiveTypes';
// Create service with primitive type support
const typeResolver = new TypeResolverBuilder()
.addPrimitive(SquizLinkType)
.addPrimitive(SquizImageType)
.build();
const service = new JSONSchemaService(typeResolver, schema);
// Resolve input data
const resolvedData = await service.resolveInput(inputData, inputSchema);
`
This library provides robust support for allOf conditional schemas, including data preservation for SquizLink and SquizImage arrays within conditional structures:
`javascript`
// ✅ FULLY SUPPORTED: Single-level allOf conditions
{
"type": "object",
"properties": {
"contentType": { "type": "string", "enum": ["basic", "advanced"] }
},
"allOf": [
{
"if": { "properties": { "contentType": { "const": "advanced" } } },
"then": {
"properties": {
"links": {
"type": "array",
"items": { "type": "SquizLink" }
}
}
}
}
]
}
#### Complex Deeply Nested AllOf Structures
Limitation: The current implementation has a known edge case with deeply nested allOf conditional schemas within array structures.
Example of unsupported pattern:
`javascript`
// ❌ COMPLEX EDGE CASE: Multi-level nested allOf in arrays
{
"type": "object",
"properties": {
"sections": {
"type": "array",
"items": {
"type": "object",
"properties": {
"sectionType": { "type": "string" }
},
"allOf": [ // ✅ Level 1 allOf - WORKS
{
"if": { "properties": { "sectionType": { "const": "content" } } },
"then": {
"properties": {
"contentGroups": {
"type": "array",
"items": {
"type": "object",
"properties": {
"groupType": { "type": "string" }
},
"allOf": [ // ❌ Level 2 allOf - EDGE CASE
{
"if": { "properties": { "groupType": { "const": "links" } } },
"then": {
"properties": {
"groupLinks": {
"type": "array",
"items": { "type": "SquizLink" }
}
}
}
}
]
}
}
}
}
}
]
}
}
}
}
Impact:
- This pattern represents a rare edge case in real-world schemas
- Known practical production use cases work perfectly
Workaround:
- Flatten deeply nested conditional structures where possible
- Use separate schema definitions for complex nested conditions
- Consider restructuring schemas to avoid multiple levels of array-embedded allOf patterns
Future Enhancement:
This limitation could be addressed in a future version by implementing:
- Recursive allOf` detection across multiple nesting levels
- Enhanced pointer-based data preservation for deeply nested structures
- Advanced JSON Schema library processing overrides for complex recursive scenarios