Skip to content

Changelog

2.5.0

  • added annotation @Ref for classes and properties: reference external schema instead of generating schema from type. Handled in the handleCoreAnnotation-step for Swagger and JSON-Schemas. #66
  • schema-kenerator-serialization: added step convertToKotlinxSerializationTypes to convert compiled schema to kotlinx-serialization types. #64
  • schema-kenerator-jsonschema: removed deprecated option OptionalHandling (was previously replaced by RequiredHandling)
  • stricter handling of value classes to prevent infinite loops for specific types ktor-openapi-tools#220
  • fixed bug: double and float schemas had incorrect min and max values. #65
  • fixed bug: null pointer exception in swagger schemas for specific types. #61

2.4.0

  • added more (core) annotations for swagger and json schemas #59
  • @Min, @Max, @ExclusiveMin, @ExclusiveMax
  • @MinLength, @MaxLength for strings and arrays
  • @Pattern for strings
  • steps addDiscriminatorProperty, addJsonClassDiscriminatorProperty and addJacksonTypeInfoDiscriminatorProperty have new optional parameter asEnum: Boolean. Specifies whether the discriminator properties should be generated as simple strings or enums with only the possible types as options. #60
  • expanded RefType and SimpleType for swagger and json schema
  • SIMPLE, OPENAPI_SIMPLE: now contain the names of outer classes for nested classes.
  • MINIMAL, OPENAPI_MINIMAL: new; same behavior as previous SIMPLE & OPENAPI_SIMPLE. Contains only the name of the class without any outer classes.
  • fixed a null pointer exception when using swagger @Schema annotation "required" property #61

2.3.0

  • type parameters are more often correctly detected when using kotlinx-serialization for analysis
  • behavior can be disabled with findTypeParametersUsingReflection = false in the analysis step config
  • deprecate markNotParameterized for kotlinx-serialization type analysis - property is no longer used

2.2.0

Improvements to JSON schema generation:

  • explicit nullables handling now configurable (now on parity with swagger schema). nullables property in generateJsonSchema-step.
  • renamed optionalHandling property to optionals in generateJsonSchema-step (now on parity with swagger schema).
  • added mergePropertyAttributesIntoType-step (now on parity with swagger schema)
  • changed default reference path from #/definitions/MySchema to #/$defs/MySchema to better match jsonschema spec (path is configurable at compile step)
  • added new step merge that creates a single json schema (from the output of the compile steps) with all separate referenced schemas placed in a $defs section (configurable) of the root schema.

Improvements to Swagger schema generation:

  • added previously missing support for @Swagger#requiredMode property
  • improved handling of schemas with multiple types (e.g. due to @Type-annotation) in specific situations. This change prevents results like "type": ["object", "myType"] where the type "object" should have been replaced by "myType".

Other improvements and fixes:

  • kotlinx-serialization can now distinguish between lists (with any items) and sets (with unique items) and set the flag accordingly (now on parity with reflection type analysis)
  • improved check determining whether a schema should be inlined or referenced. This should lead to simpler and more predictable schemas.

2.1.4

Upgrade dependency versions - Jackson 2.15.3 -> 2.19.1 - Jackson Serialization Json 1.7.3 -> 1.8.1 - Swagger Parser 2.1.24 -> 2.1.30 - Jakarta Validation 3.0.0 -> 3.1.1 - Kotest 5.8.0 -> 5.9.1 - forced commons-codec 1.11 -> 1.13 (transitive dependency)

2.1.3

  • improved handling of recursion with inlined schemas: if the root type of a schema is used recursively, it is referenced instead of inlined.
  • fixed: null pointer exception when using multiple @Contextual annotations
  • fixed: fixed handling of @Description annotation when added to properties and classes

2.1.2

  • fixed bug: duplicate "string" type added by type discriminator handling
  • fixed bug: incorrect comparison check in "mergePropertyAttributesIntoType" step caused infinite loop

2.1.1

  • fix infinite loop in mergePropertyAttributesIntoType-step (#50, #51)

2.1.0

  • add initial step extension functions to all InitialTypeData instead of specific subclasses

2.0.0

  • re-structured reflection and kotlinx-serialization type analysis
    • allow more powerful custom type analyzers
    • open up more internals for custom processors while still being able to re-use common functions
  • re-structures swagger and json schema generation
    • allow more powerful custom schema generators
  • reworked/replaced bundles & overhauled input data classes for steps to make flow of data and order of steps more clear
    • generic bundles replaced with specific classes "InitialTypeData", "TypeDataGroup", "IntermediateXYSchemaData", "CompiledXYSchemaData"
    • start of pipeline changed from extension functions on raw kotlin types (e.g. typeOf, T::class, etc.) to initial, initial(T::clas), etc.
  • overhauled type redirection
    • enhance configuration dsl
    • more options to define exact behavior
      • matching
        • MATCH: whether the nullability of the "from" has to match exactly
        • IGNORE: nullability of the "from" type is ignored when matching redirects
      • replacing
        • KEEP: whether the nullability of the "from" type should be kept with the replaced type
        • REPLACE: whether the nullability is also overwritten
  • reworked base data models to better fit requirements and make them more flexible
  • added option to enable/disable explicit "null"-types for properties (e.g. .compileInlining(explicitNullTypes = true)
  • added option to allow special float values (i.e. "NaN", "Infinity", "-Infinity") for Float and Double (e.g. .generateSwaggerSchema { allowSpecialFloatingPointValues = true })
  • added option to generate schemas for maps with complex keys as array with the map key and map values as valid item types (e.g. .generateSwaggerSchema { mapsWithStructuredKeysAsArrays = true })
  • added option to swagger schema generation to restrict values of discriminator properties to the name of the type (strictDiscriminatorProperty))
  • option to handle nullable properties as required or non-required (e.g. .generateSwaggerSchema { nullables = RequiredHandling.REQUIRED })
  • added support for kotlinx-serialization @Contextual annotation and @Serializable(with = ...)
  • improved support for annotations on fields that modify the type/schema (e.g. @Format on properties)
  • renamed steps to better reflect their function and purpose:
    • connectSubTypes to addMissingSupertypeSubtypeRelations
    • mergeGetters to gettersToProperties
    • renameProperties to renameMembers
    • processReflection to analyzeTypeUsingReflection
    • processSerialization to analyzeTypeUsingKotlinxSerialization
  • removed deprecated "withAutoTitle" step
  • shortened & simplified package structure
  • make classes "internal" to reduce namespace pollution
  • simplified and removed unnecessary dependencies, opened up transitive dependencies to library consumers
  • fixed bug: type parameter nullability information was lost
  • fixed bug: @Type-annotation not handled correctly
  • fixed renaming swagger schema properties #44

1.6.4

  • fixed bug: when using reflection to analyze generic types, too many interfaces might be detected as subtypes #43
  • fixed bug: when generating schema for a type overwriting a property with a more specific type, the wrong type was used in the resulting schema #43

1.6.3

  • fixed: unchanged properties of swagger @Schema and @ArraySchema annotations were not ignored during schema generation

1.5.1

  • fixed: unchanged properties of swagger @Schema and @ArraySchema annotations were not ignored during schema generation

1.6.2

  • fixed/add support for top-level generics with kotlinx

1.6.1

  • upgrade schema io.swagger.parser.v3:swagger-parser from 2.1.20 to 2.1.24

1.6.0

  • added core annotations to specify schema 'type' and 'format': @Schema, @Type. Can be added on classes and fields
  • exposes swagger library via gradle "api"
  • swagger compilation step treats both openapi30 type and openapi31 types as valid types
  • fixed bug: NPE during swagger compile inline step when using classes that have a nullable sealed class property

1.5.0

  • added support for discriminator mapping in swagger schemas
  • added support for example property on swagger @Schema-annotation #32
  • added new core annotation @Format #36
  • upgrade kotlin to 2.0.21
  • fixed: naming conflict in core @Default annotation #35

1.4.3

  • fixed: swagger enum types were missing type "string"

1.4.2

  • fixed bug: swagger schema now internally uses types (for openapi 3.1) instead of type (for openapi 3.0) for everything #28
  • fixed bug: ignore default/unchanged values in swagger "Schema"-annotation #31
  • fixed bug: properly merge annotations on properties with annotations on types #27
  • fixed bug: null-pointer exception #26
  • fixed bug: (json-schema) properties with same types share modifications even though they should be applied to specific properties

1.4.1

  • fix bug: wrong formatting with swagger TitleType/PathType "OPENAPI_FULL" and "OPENAPI_SIMPLE"

1.4.0

  • added step renameProperties to rename properties (including support for kotlinx-serialization JsonNamingStrategy)
  • support for discriminators
  • addDiscriminatorProperty-step to add discriminator property with specified name to all types with subtypes
  • addJacksonTypeInfoDiscriminatorProperty-step to add discriminator property with name specified by jackson JsonTypeInfo-annotation to all types with subtypes
  • addJsonClassDiscriminatorProperty-step to add discriminator property with name specified by kotlinx-serialization JsonClassDiscriminator-annotation to all types with subtypes
  • discriminator property is included in swagger schema-generation
  • add project schema-kenerator-validations with support for javax and jackarta validation annotation in swagger-schemas
  • supported jakarta annotations:
    • NotNull
    • NotEmpty
    • NotBlank
    • Size
    • Min
    • Max

  • supported javax annotations:

    • NotNull
    • NotEmpty
    • NotBlank
    • Size
    • Min
    • Max

  • annotations on constructor parameters are added to annotations of matching property

  • handle nullability of fields according to openapi 3.1.0, e.g. nullable string as type: ["string", "null"]
  • fixed bug: required-annotation is ignored when all fields of a class would be nullable/optional
  • fixed bug: nullpointer-exception (see issue #22)

1.3.0

please ignore this version - i messed up during release here

1.2.3

  • fix issues with nullability of fields

1.2.2

  • fixed bug: wrong format for openapi TitleType and RefType

1.2.1

  • fixed bug: wrong format for TitleType.OPENAPI_SIMPLE

1.2.0

  • added TitleType.OPENAPI_SIMPLE, TitleType.OPENAPI_FULL, RefType.OPENAPI_SIMPLE, RefType.OPENAPI_FULL as options for swagger that generate results confirming to the openapi-spec
  • RefType.OPENAPI_FULL replaces RefType.FULL as the default for swagger schemas
  • renames step withAutoTitle(...) to withTitle(...)
  • step withTitle(...) can take a builder function to provide a custom title
  • step compileReferencing(...) and compileReferencingRoot(...) can take a builder function to provide a custom path
  • fixed bug: redirecting to nullable types did not work correctly

1.1.1

  • fixed bug: swagger example property not set correctly. Used examples instead of example.
  • fixed bug: exception when using Schema-annotation with blank minimum and maximum
  • fixed bug: Default-annotation not handled for properties (swagger and json-schema)

1.1.0

  • added steps customizeTypes and customizeProperties for easier manual customization of json-schema and swagger-schema

  • added support for optional properties

  • detect optional parameters (non nullable, but with a default value provided)

  • handle in schema generation step as required or non required (configurable)

  • added support for inline value classes

  • detect value classes

  • inline "wrapped" type in schema generation

  • added config option knownNotParameterized for type processing with kotlinx-serialization to manually prevent single type appearing multiple times in some setups

  • added support for annotations with kotlinx-serialization

  • fixed: type redirects not working with kotlinx-serialization

  • fixed: exception when qualified name is missing, e.g. for local classes

1.0.0

  • 🥳

0.4.0

  • reworked naming of steps to be more consistent and clearer
  • detect whether a collection allows only unique elements (i.e. is a set) and include uniqueItems in json and swagger schema generation
  • implemented type redirects to allow processing a custom type instead of the real one
  • fix: exception when filtering members of some types
  • fix: exception when processing type parameters of some types

0.3.0

  • merge annotations from kotlin property and java field to fix annotations not being able to be detected
  • add support for Jackson annotations
  • @JsonPropertyDescription
  • @JsonIgnore
  • @JsonIgnoreType
  • @JsonIgnoreProperties
  • @JsonProperty
  • add support for swagger annotations
  • @Schema
  • @ArraySchema
  • add new step to merge getters with their properties

0.2.0

  • added SchemaName annotation to "rename" types (qualified and simple name)
  • added options to choose whether to use the name or the toString-function for enum constants
  • fix type parameters missing for maps and collections
  • fix schema reference paths
  • improved dsl of custom processors

0.1.1

  • fixed bug with typeId parser
  • fixed build config

0.1.0

  • initial version