airbyte-cdk 6.9.0__py3-none-any.whl → 6.9.0.dev0__py3-none-any.whl

airbyte_cdk/sources/declarative/concurrent_declarative_source.py

@@ -86,10 +86,23 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
             component_factory=component_factory,
         )
 
-        # todo: We could remove state from initialization. Now that streams are grouped during the read(), a source
-        #  no longer needs to store the original incoming state. But maybe there's an edge case?
         self._state = state
 
+        self._concurrent_streams: Optional[List[AbstractStream]]
+        self._synchronous_streams: Optional[List[Stream]]
+
+        # If the connector command was SPEC, there is no incoming config, and we cannot instantiate streams because
+        # they might depend on it. Ideally we want to have a static method on this class to get the spec without
+        # any other arguments, but the existing entrypoint.py isn't designed to support this. Just noting this
+        # for our future improvements to the CDK.
+        if config:
+            self._concurrent_streams, self._synchronous_streams = self._group_streams(
+                config=config or {}
+            )
+        else:
+            self._concurrent_streams = None
+            self._synchronous_streams = None
+
         concurrency_level_from_manifest = self._source_config.get("concurrency_level")
         if concurrency_level_from_manifest:
             concurrency_level_component = self._constructor.create_component(
@@ -123,20 +136,17 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
         logger: logging.Logger,
         config: Mapping[str, Any],
         catalog: ConfiguredAirbyteCatalog,
-        state: Optional[List[AirbyteStateMessage]] = None,
+        state: Optional[Union[List[AirbyteStateMessage]]] = None,
     ) -> Iterator[AirbyteMessage]:
-        concurrent_streams, _ = self._group_streams(config=config)
-
-        # ConcurrentReadProcessor pops streams that are finished being read so before syncing, the names of
-        # the concurrent streams must be saved so that they can be removed from the catalog before starting
-        # synchronous streams
-        if len(concurrent_streams) > 0:
+        # ConcurrentReadProcessor pops streams that are finished being read so before syncing, the names of the concurrent
+        # streams must be saved so that they can be removed from the catalog before starting synchronous streams
+        if self._concurrent_streams:
             concurrent_stream_names = set(
-                [concurrent_stream.name for concurrent_stream in concurrent_streams]
+                [concurrent_stream.name for concurrent_stream in self._concurrent_streams]
             )
 
             selected_concurrent_streams = self._select_streams(
-                streams=concurrent_streams, configured_catalog=catalog
+                streams=self._concurrent_streams, configured_catalog=catalog
             )
             # It would appear that passing in an empty set of streams causes an infinite loop in ConcurrentReadProcessor.
             # This is also evident in concurrent_source_adapter.py so I'll leave this out of scope to fix for now
@@ -155,7 +165,8 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
         yield from super().read(logger, config, filtered_catalog, state)
 
     def discover(self, logger: logging.Logger, config: Mapping[str, Any]) -> AirbyteCatalog:
-        concurrent_streams, synchronous_streams = self._group_streams(config=config)
+        concurrent_streams = self._concurrent_streams or []
+        synchronous_streams = self._synchronous_streams or []
         return AirbyteCatalog(
             streams=[
                 stream.as_airbyte_stream() for stream in concurrent_streams + synchronous_streams
@@ -195,7 +206,7 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
             # so we need to treat them as synchronous
             if (
                 isinstance(declarative_stream, DeclarativeStream)
-                and name_to_stream_mapping[declarative_stream.name]["retriever"]["type"]
+                and name_to_stream_mapping[declarative_stream.name].get("retriever")["type"]
                 == "SimpleRetriever"
             ):
                 incremental_sync_component_definition = name_to_stream_mapping[
@@ -204,7 +215,7 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
 
                 partition_router_component_definition = (
                     name_to_stream_mapping[declarative_stream.name]
-                    .get("retriever", {})
+                    .get("retriever")
                     .get("partition_router")
                 )
                 is_without_partition_router_or_cursor = not bool(
@@ -226,7 +237,7 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                     cursor = self._constructor.create_concurrent_cursor_from_datetime_based_cursor(
                         state_manager=state_manager,
                         model_type=DatetimeBasedCursorModel,
-                        component_definition=incremental_sync_component_definition,  # type: ignore  # Not None because of the if condition above
+                        component_definition=incremental_sync_component_definition,
                         stream_name=declarative_stream.name,
                         stream_namespace=declarative_stream.namespace,
                         config=config or {},
@@ -309,11 +320,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
     def _is_datetime_incremental_without_partition_routing(
         self,
         declarative_stream: DeclarativeStream,
-        incremental_sync_component_definition: Mapping[str, Any] | None,
+        incremental_sync_component_definition: Mapping[str, Any],
     ) -> bool:
         return (
-            incremental_sync_component_definition is not None
-            and bool(incremental_sync_component_definition)
+            bool(incremental_sync_component_definition)
             and incremental_sync_component_definition.get("type", "")
             == DatetimeBasedCursorModel.__name__
             and self._stream_supports_concurrent_partition_processing(

airbyte_cdk/sources/declarative/declarative_component_schema.yaml

@@ -1684,6 +1684,92 @@ definitions:
       $parameters:
         type: object
         additionalProperties: true
+  TypesMap:
+    title: Types Map
+    description: (This component is experimental. Use at your own risk.) Represents a mapping between a current type and its corresponding target type.
+    type: object
+    required:
+      - target_type
+      - current_type
+    properties:
+      target_type:
+        anyOf:
+          - type: string
+          - type: array
+            items:
+              type: string
+      current_type:
+        anyOf:
+          - type: string
+          - type: array
+            items:
+              type: string
+  SchemaTypeIdentifier:
+    title: Schema Type Identifier
+    description: (This component is experimental. Use at your own risk.) Identifies schema details for dynamic schema extraction and processing.
+    type: object
+    required:
+      - key_pointer
+    properties:
+      type:
+        type: string
+        enum: [SchemaTypeIdentifier]
+      schema_pointer:
+        title: Schema Path
+        description: List of nested fields defining the schema field path to extract. Defaults to [].
+        type: array
+        default: []
+        items:
+          - type: string
+        interpolation_content:
+          - config
+      key_pointer:
+        title: Key Path
+        description: List of potentially nested fields describing the full path of the field key to extract.
+        type: array
+        items:
+          - type: string
+        interpolation_content:
+          - config
+      type_pointer:
+        title: Type Path
+        description: List of potentially nested fields describing the full path of the field type to extract.
+        type: array
+        items:
+          - type: string
+        interpolation_content:
+          - config
+      types_mapping:
+        type: array
+        items:
+          - "$ref": "#/definitions/TypesMap"
+      $parameters:
+        type: object
+        additionalProperties: true
+  DynamicSchemaLoader:
+    title: Dynamic Schema Loader
+    description: (This component is experimental. Use at your own risk.) Loads a schema by extracting data from retrieved records.
+    type: object
+    required:
+      - type
+      - retriever
+      - schema_type_identifier
+    properties:
+      type:
+        type: string
+        enum: [DynamicSchemaLoader]
+      retriever:
+        title: Retriever
+        description: Component used to coordinate how records are extracted across stream slices and request pages.
+        anyOf:
+          - "$ref": "#/definitions/AsyncRetriever"
+          - "$ref": "#/definitions/CustomRetriever"
+          - "$ref": "#/definitions/SimpleRetriever"
+      schema_type_identifier:
+        "$ref": "#/definitions/SchemaTypeIdentifier"
+      $parameters:
+        type: object
+        additionalProperties: true
   InlineSchemaLoader:
     title: Inline Schema Loader
     description: Loads a schema that is defined directly in the manifest file.

airbyte_cdk/sources/declarative/models/declarative_component_schema.py

@@ -650,6 +650,32 @@ class HttpResponseFilter(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
+class TypesMap(BaseModel):
+    target_type: Union[str, List[str]]
+    current_type: Union[str, List[str]]
+
+
+class SchemaTypeIdentifier(BaseModel):
+    type: Optional[Literal["SchemaTypeIdentifier"]] = None
+    schema_pointer: Optional[List[str]] = Field(
+        [],
+        description="List of nested fields defining the schema field path to extract. Defaults to [].",
+        title="Schema Path",
+    )
+    key_pointer: List[str] = Field(
+        ...,
+        description="List of potentially nested fields describing the full path of the field key to extract.",
+        title="Key Path",
+    )
+    type_pointer: Optional[List[str]] = Field(
+        None,
+        description="List of potentially nested fields describing the full path of the field type to extract.",
+        title="Type Path",
+    )
+    types_mapping: Optional[List[TypesMap]] = None
+    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
+
+
 class InlineSchemaLoader(BaseModel):
     type: Literal["InlineSchemaLoader"]
     schema_: Optional[Dict[str, Any]] = Field(
@@ -822,13 +848,13 @@ class OauthConnectorInputSpecification(BaseModel):
     )
     extract_output: List[str] = Field(
         ...,
-        description="The DeclarativeOAuth Specific list of strings to indicate which keys should be extracted and returned back to the input config.                ",
+        description="The DeclarativeOAuth Specific list of strings to indicate which keys should be extracted and returned back to the input config.",
         examples=[{"extract_output": ["access_token", "refresh_token", "other_field"]}],
         title="DeclarativeOAuth Extract Output",
     )
     state: Optional[State] = Field(
         None,
-        description="The DeclarativeOAuth Specific object to provide the criteria of how the `state` query param should be constructed,\nincluding length and complexity.                ",
+        description="The DeclarativeOAuth Specific object to provide the criteria of how the `state` query param should be constructed,\nincluding length and complexity.",
         examples=[{"state": {"min": 7, "max": 128}}],
         title="(Optional) DeclarativeOAuth Configurable State Query Param",
     )
@@ -852,13 +878,13 @@ class OauthConnectorInputSpecification(BaseModel):
     )
     state_key: Optional[str] = Field(
         None,
-        description="The DeclarativeOAuth Specific optional override to provide the custom `state` key name, if required by data-provider.                ",
+        description="The DeclarativeOAuth Specific optional override to provide the custom `state` key name, if required by data-provider.",
         examples=[{"state_key": "my_custom_state_key_key_name"}],
         title="(Optional) DeclarativeOAuth State Key Override",
     )
     auth_code_key: Optional[str] = Field(
         None,
-        description="The DeclarativeOAuth Specific optional override to provide the custom `code` key name to something like `auth_code` or `custom_auth_code`, if required by data-provider.                ",
+        description="The DeclarativeOAuth Specific optional override to provide the custom `code` key name to something like `auth_code` or `custom_auth_code`, if required by data-provider.",
         examples=[{"auth_code_key": "my_custom_auth_code_key_name"}],
         title="(Optional) DeclarativeOAuth Auth Code Key Override",
     )
@@ -1774,6 +1800,17 @@ class HttpRequester(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
+class DynamicSchemaLoader(BaseModel):
+    type: Literal["DynamicSchemaLoader"]
+    retriever: Union[AsyncRetriever, CustomRetriever, SimpleRetriever] = Field(
+        ...,
+        description="Component used to coordinate how records are extracted across stream slices and request pages.",
+        title="Retriever",
+    )
+    schema_type_identifier: SchemaTypeIdentifier
+    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
+
+
 class ParentStreamConfig(BaseModel):
     type: Literal["ParentStreamConfig"]
     parent_key: str = Field(
@@ -1981,5 +2018,6 @@ DeclarativeSource2.update_forward_refs()
 SelectiveAuthenticator.update_forward_refs()
 DeclarativeStream.update_forward_refs()
 SessionTokenAuthenticator.update_forward_refs()
+DynamicSchemaLoader.update_forward_refs()
 SimpleRetriever.update_forward_refs()
 AsyncRetriever.update_forward_refs()

airbyte_cdk/sources/declarative/parsers/manifest_component_transformer.py

@@ -64,6 +64,10 @@ DEFAULT_MODEL_TYPES: Mapping[str, str] = {
     "AddFields.fields": "AddedFieldDefinition",
     # CustomPartitionRouter
     "CustomPartitionRouter.parent_stream_configs": "ParentStreamConfig",
+    # DynamicSchemaLoader
+    "DynamicSchemaLoader.retriever": "SimpleRetriever",
+    # SchemaTypeIdentifier
+    "SchemaTypeIdentifier.types_map": "TypesMap",
 }
 
 # We retain a separate registry for custom components to automatically insert the type if it is missing. This is intended to

airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py

@@ -188,6 +188,9 @@ from airbyte_cdk.sources.declarative.models.declarative_component_schema import
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     DpathExtractor as DpathExtractorModel,
 )
+from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
+    DynamicSchemaLoader as DynamicSchemaLoaderModel,
+)
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     ExponentialBackoffStrategy as ExponentialBackoffStrategyModel,
 )
@@ -278,6 +281,9 @@ from airbyte_cdk.sources.declarative.models.declarative_component_schema import
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     ResponseToFileExtractor as ResponseToFileExtractorModel,
 )
+from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
+    SchemaTypeIdentifier as SchemaTypeIdentifierModel,
+)
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     SelectiveAuthenticator as SelectiveAuthenticatorModel,
 )
@@ -291,6 +297,9 @@ from airbyte_cdk.sources.declarative.models.declarative_component_schema import
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     SubstreamPartitionRouter as SubstreamPartitionRouterModel,
 )
+from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
+    TypesMap as TypesMapModel,
+)
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import ValueType
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     WaitTimeFromHeader as WaitTimeFromHeaderModel,
@@ -356,8 +365,11 @@ from airbyte_cdk.sources.declarative.retrievers import (
 )
 from airbyte_cdk.sources.declarative.schema import (
     DefaultSchemaLoader,
+    DynamicSchemaLoader,
     InlineSchemaLoader,
     JsonFileSchemaLoader,
+    SchemaTypeIdentifier,
+    TypesMap,
 )
 from airbyte_cdk.sources.declarative.spec import Spec
 from airbyte_cdk.sources.declarative.stream_slicers import StreamSlicer
@@ -455,6 +467,9 @@ class ModelToComponentFactory:
             IterableDecoderModel: self.create_iterable_decoder,
             XmlDecoderModel: self.create_xml_decoder,
             JsonFileSchemaLoaderModel: self.create_json_file_schema_loader,
+            DynamicSchemaLoaderModel: self.create_dynamic_schema_loader,
+            SchemaTypeIdentifierModel: self.create_schema_type_identifier,
+            TypesMapModel: self.create_types_map,
             JwtAuthenticatorModel: self.create_jwt_authenticator,
             LegacyToPerPartitionStateMigrationModel: self.create_legacy_to_per_partition_state_migration,
             ListPartitionRouterModel: self.create_list_partition_router,
@@ -1574,6 +1589,63 @@ class ModelToComponentFactory:
     ) -> InlineSchemaLoader:
         return InlineSchemaLoader(schema=model.schema_ or {}, parameters={})
 
+    @staticmethod
+    def create_types_map(model: TypesMapModel, **kwargs: Any) -> TypesMap:
+        return TypesMap(target_type=model.target_type, current_type=model.current_type)
+
+    def create_schema_type_identifier(
+        self, model: SchemaTypeIdentifierModel, config: Config, **kwargs: Any
+    ) -> SchemaTypeIdentifier:
+        types_mapping = []
+        if model.types_mapping:
+            types_mapping.extend(
+                [
+                    self._create_component_from_model(types_map, config=config)
+                    for types_map in model.types_mapping
+                ]
+            )
+        model_schema_pointer: List[Union[InterpolatedString, str]] = (
+            [x for x in model.schema_pointer] if model.schema_pointer else []
+        )
+        model_key_pointer: List[Union[InterpolatedString, str]] = [x for x in model.key_pointer]
+        model_type_pointer: Optional[List[Union[InterpolatedString, str]]] = (
+            [x for x in model.type_pointer] if model.type_pointer else None
+        )
+
+        return SchemaTypeIdentifier(
+            schema_pointer=model_schema_pointer,
+            key_pointer=model_key_pointer,
+            type_pointer=model_type_pointer,
+            types_mapping=types_mapping,
+            parameters=model.parameters or {},
+        )
+
+    def create_dynamic_schema_loader(
+        self, model: DynamicSchemaLoaderModel, config: Config, **kwargs: Any
+    ) -> DynamicSchemaLoader:
+        stream_slicer = self._build_stream_slicer_from_partition_router(model.retriever, config)
+        combined_slicers = self._build_resumable_cursor_from_paginator(
+            model.retriever, stream_slicer
+        )
+
+        retriever = self._create_component_from_model(
+            model=model.retriever,
+            config=config,
+            name="",
+            primary_key=None,
+            stream_slicer=combined_slicers,
+            transformations=[],
+        )
+        schema_type_identifier = self._create_component_from_model(
+            model.schema_type_identifier, config=config, parameters=model.parameters or {}
+        )
+        return DynamicSchemaLoader(
+            retriever=retriever,
+            config=config,
+            schema_type_identifier=schema_type_identifier,
+            parameters=model.parameters or {},
+        )
+
     @staticmethod
     def create_json_decoder(model: JsonDecoderModel, config: Config, **kwargs: Any) -> JsonDecoder:
         return JsonDecoder(parameters={})

airbyte_cdk/sources/declarative/schema/__init__.py

@@ -6,5 +6,6 @@ from airbyte_cdk.sources.declarative.schema.default_schema_loader import Default
 from airbyte_cdk.sources.declarative.schema.inline_schema_loader import InlineSchemaLoader
 from airbyte_cdk.sources.declarative.schema.json_file_schema_loader import JsonFileSchemaLoader
 from airbyte_cdk.sources.declarative.schema.schema_loader import SchemaLoader
+from airbyte_cdk.sources.declarative.schema.dynamic_schema_loader import DynamicSchemaLoader, TypesMap, SchemaTypeIdentifier
 
-__all__ = ["JsonFileSchemaLoader", "DefaultSchemaLoader", "SchemaLoader", "InlineSchemaLoader"]
+__all__ = ["JsonFileSchemaLoader", "DefaultSchemaLoader", "SchemaLoader", "InlineSchemaLoader", "DynamicSchemaLoader", "TypesMap", "SchemaTypeIdentifier"]

airbyte_cdk/sources/declarative/schema/dynamic_schema_loader.py

@@ -0,0 +1,219 @@
+#
+# Copyright (c) 2024 Airbyte, Inc., all rights reserved.
+#
+
+
+from copy import deepcopy
+from dataclasses import InitVar, dataclass
+from typing import Any, List, Mapping, MutableMapping, Optional, Union
+
+import dpath
+from typing_extensions import deprecated
+
+from airbyte_cdk.sources.declarative.interpolation.interpolated_string import InterpolatedString
+from airbyte_cdk.sources.declarative.retrievers.retriever import Retriever
+from airbyte_cdk.sources.declarative.schema.schema_loader import SchemaLoader
+from airbyte_cdk.sources.source import ExperimentalClassWarning
+from airbyte_cdk.sources.types import Config
+
+AIRBYTE_DATA_TYPES: Mapping[str, Mapping[str, Any]] = {
+    "string": {"type": ["null", "string"]},
+    "boolean": {"type": ["null", "boolean"]},
+    "date": {"type": ["null", "string"], "format": "date"},
+    "timestamp_without_timezone": {
+        "type": ["null", "string"],
+        "format": "date-time",
+        "airbyte_type": "timestamp_without_timezone",
+    },
+    "timestamp_with_timezone": {"type": ["null", "string"], "format": "date-time"},
+    "time_without_timezone": {
+        "type": ["null", "string"],
+        "format": "time",
+        "airbyte_type": "time_without_timezone",
+    },
+    "time_with_timezone": {
+        "type": ["null", "string"],
+        "format": "time",
+        "airbyte_type": "time_with_timezone",
+    },
+    "integer": {"type": ["null", "integer"]},
+    "number": {"type": ["null", "number"]},
+    "array": {"type": ["null", "array"]},
+    "object": {"type": ["null", "object"]},
+}
+
+
+@deprecated("This class is experimental. Use at your own risk.", category=ExperimentalClassWarning)
+@dataclass(frozen=True)
+class TypesMap:
+    """
+    Represents a mapping between a current type and its corresponding target type.
+    """
+
+    target_type: Union[List[str], str]
+    current_type: Union[List[str], str]
+
+
+@deprecated("This class is experimental. Use at your own risk.", category=ExperimentalClassWarning)
+@dataclass
+class SchemaTypeIdentifier:
+    """
+    Identifies schema details for dynamic schema extraction and processing.
+    """
+
+    key_pointer: List[Union[InterpolatedString, str]]
+    parameters: InitVar[Mapping[str, Any]]
+    type_pointer: Optional[List[Union[InterpolatedString, str]]] = None
+    types_mapping: Optional[List[TypesMap]] = None
+    schema_pointer: Optional[List[Union[InterpolatedString, str]]] = None
+
+    def __post_init__(self, parameters: Mapping[str, Any]) -> None:
+        self.schema_pointer = (
+            self._update_pointer(self.schema_pointer, parameters) if self.schema_pointer else []
+        )  # type: ignore[assignment]  # This is reqired field in model
+        self.key_pointer = self._update_pointer(self.key_pointer, parameters)  # type: ignore[assignment]  # This is reqired field in model
+        self.type_pointer = (
+            self._update_pointer(self.type_pointer, parameters) if self.type_pointer else None
+        )
+
+    @staticmethod
+    def _update_pointer(
+        pointer: Optional[List[Union[InterpolatedString, str]]], parameters: Mapping[str, Any]
+    ) -> Optional[List[Union[InterpolatedString, str]]]:
+        return (
+            [
+                InterpolatedString.create(path, parameters=parameters)
+                if isinstance(path, str)
+                else path
+                for path in pointer
+            ]
+            if pointer
+            else None
+        )
+
+
+@deprecated("This class is experimental. Use at your own risk.", category=ExperimentalClassWarning)
+@dataclass
+class DynamicSchemaLoader(SchemaLoader):
+    """
+    Dynamically loads a JSON Schema by extracting data from retrieved records.
+    """
+
+    retriever: Retriever
+    config: Config
+    parameters: InitVar[Mapping[str, Any]]
+    schema_type_identifier: SchemaTypeIdentifier
+
+    def get_json_schema(self) -> Mapping[str, Any]:
+        """
+        Constructs a JSON Schema based on retrieved data.
+        """
+        properties = {}
+        retrieved_record = next(self.retriever.read_records({}), None)  # type: ignore[call-overload] # read_records return Iterable data type
+
+        raw_schema = (
+            self._extract_data(
+                retrieved_record,  # type: ignore[arg-type] # Expected that retrieved_record will be only Mapping[str, Any]
+                self.schema_type_identifier.schema_pointer,
+            )
+            if retrieved_record
+            else []
+        )
+
+        for property_definition in raw_schema:
+            key = self._get_key(property_definition, self.schema_type_identifier.key_pointer)
+            value = self._get_type(
+                property_definition,
+                self.schema_type_identifier.type_pointer,
+            )
+            properties[key] = value
+
+        return {
+            "$schema": "http://json-schema.org/draft-07/schema#",
+            "type": "object",
+            "properties": properties,
+        }
+
+    def _get_key(
+        self,
+        raw_schema: MutableMapping[str, Any],
+        field_key_path: List[Union[InterpolatedString, str]],
+    ) -> str:
+        """
+        Extracts the key field from the schema using the specified path.
+        """
+        field_key = self._extract_data(raw_schema, field_key_path)
+        if not isinstance(field_key, str):
+            raise ValueError(f"Expected key to be a string. Got {field_key}")
+        return field_key
+
+    def _get_type(
+        self,
+        raw_schema: MutableMapping[str, Any],
+        field_type_path: Optional[List[Union[InterpolatedString, str]]],
+    ) -> Union[Mapping[str, Any], List[Mapping[str, Any]]]:
+        """
+        Determines the JSON Schema type for a field, supporting nullable and combined types.
+        """
+        raw_field_type = (
+            self._extract_data(raw_schema, field_type_path, default="string")
+            if field_type_path
+            else "string"
+        )
+        mapped_field_type = self._replace_type_if_not_valid(raw_field_type)
+        if (
+            isinstance(mapped_field_type, list)
+            and len(mapped_field_type) == 2
+            and all(isinstance(item, str) for item in mapped_field_type)
+        ):
+            first_type = self._get_airbyte_type(mapped_field_type[0])
+            second_type = self._get_airbyte_type(mapped_field_type[1])
+            return {"oneOf": [first_type, second_type]}
+        elif isinstance(mapped_field_type, str):
+            return self._get_airbyte_type(mapped_field_type)
+        else:
+            raise ValueError(
+                f"Invalid data type. Available string or two items list of string. Got {mapped_field_type}."
+            )
+
+    def _replace_type_if_not_valid(
+        self, field_type: Union[List[str], str]
+    ) -> Union[List[str], str]:
+        """
+        Replaces a field type if it matches a type mapping in `types_map`.
+        """
+        if self.schema_type_identifier.types_mapping:
+            for types_map in self.schema_type_identifier.types_mapping:
+                if field_type == types_map.current_type:
+                    return types_map.target_type
+        return field_type
+
+    @staticmethod
+    def _get_airbyte_type(field_type: str) -> Mapping[str, Any]:
+        """
+        Maps a field type to its corresponding Airbyte type definition.
+        """
+        if field_type not in AIRBYTE_DATA_TYPES:
+            raise ValueError(f"Invalid Airbyte data type: {field_type}")
+
+        return deepcopy(AIRBYTE_DATA_TYPES[field_type])
+
+    def _extract_data(
+        self,
+        body: Mapping[str, Any],
+        extraction_path: Optional[List[Union[InterpolatedString, str]]] = None,
+        default: Any = None,
+    ) -> Any:
+        """
+        Extracts data from the body based on the provided extraction path.
+        """
+
+        if not extraction_path:
+            return body
+
+        path = [
+            node.eval(self.config) if not isinstance(node, str) else node
+            for node in extraction_path
+        ]
+
+        return dpath.get(body, path, default=default)  # type: ignore # extracted will be a MutableMapping, given input data structure

{airbyte_cdk-6.9.0.dist-info → airbyte_cdk-6.9.0.dev0.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 6.9.0
+Version: 6.9.0.dev0
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT
@@ -18,10 +18,12 @@ Classifier: Programming Language :: Python :: 3.12
 Classifier: Topic :: Scientific/Engineering
 Classifier: Topic :: Software Development :: Libraries :: Python Modules
 Provides-Extra: file-based
+Provides-Extra: sphinx-docs
 Provides-Extra: sql
 Provides-Extra: vector-db-based
 Requires-Dist: Jinja2 (>=3.1.2,<3.2.0)
 Requires-Dist: PyYAML (>=6.0.1,<7.0.0)
+Requires-Dist: Sphinx (>=4.2,<4.3) ; extra == "sphinx-docs"
 Requires-Dist: airbyte-protocol-models-dataclasses (>=0.14,<0.15)
 Requires-Dist: avro (>=1.11.2,<1.12.0) ; extra == "file-based"
 Requires-Dist: backoff
@@ -61,6 +63,7 @@ Requires-Dist: rapidfuzz (>=3.10.1,<4.0.0)
 Requires-Dist: requests
 Requires-Dist: requests_cache
 Requires-Dist: serpyco-rs (>=1.10.2,<2.0.0)
+Requires-Dist: sphinx-rtd-theme (>=1.0,<1.1) ; extra == "sphinx-docs"
 Requires-Dist: sqlalchemy (>=2.0,<3.0,!=2.0.36) ; extra == "sql"
 Requires-Dist: tiktoken (==0.8.0) ; extra == "vector-db-based"
 Requires-Dist: unstructured.pytesseract (>=0.3.12) ; extra == "file-based"

{airbyte_cdk-6.9.0.dist-info → airbyte_cdk-6.9.0.dev0.dist-info}/RECORD

@@ -62,11 +62,11 @@ airbyte_cdk/sources/declarative/checks/check_stream.py,sha256=dAA-UhmMj0WLXCkRQr
 airbyte_cdk/sources/declarative/checks/connection_checker.py,sha256=MBRJo6WJlZQHpIfOGaNOkkHUmgUl_4wDM6VPo41z5Ss,1383
 airbyte_cdk/sources/declarative/concurrency_level/__init__.py,sha256=5XUqrmlstYlMM0j6crktlKQwALek0uiz2D3WdM46MyA,191
 airbyte_cdk/sources/declarative/concurrency_level/concurrency_level.py,sha256=YIwCTCpOr_QSNW4ltQK0yUGWInI8PKNY216HOOegYLk,2101
-airbyte_cdk/sources/declarative/concurrent_declarative_source.py,sha256=9OXOFzG5PBL_MHEJg4ETE0TXfXeuhvv-at38AN3wWEM,23323
+airbyte_cdk/sources/declarative/concurrent_declarative_source.py,sha256=hgKamhOh1B8RA4Fx8FmCl4ORc7eO2h_RhxbkQovh3FM,23724
 airbyte_cdk/sources/declarative/datetime/__init__.py,sha256=l9LG7Qm6e5r_qgqfVKnx3mXYtg1I9MmMjomVIPfU4XA,177
 airbyte_cdk/sources/declarative/datetime/datetime_parser.py,sha256=SX9JjdesN1edN2WVUVMzU_ptqp2QB1OnsnjZ4mwcX7w,2579
 airbyte_cdk/sources/declarative/datetime/min_max_datetime.py,sha256=0BHBtDNQZfvwM45-tY5pNlTcKAFSGGNxemoi0Jic-0E,5785
-airbyte_cdk/sources/declarative/declarative_component_schema.yaml,sha256=kZkGoasVzufDE2BkFo_7DVO6xHL9kueNBjddtl-7kaU,124134
+airbyte_cdk/sources/declarative/declarative_component_schema.yaml,sha256=ZCYl6v0miacvpIt6M8FakkGZpEsY8SmB4_436sHEw9Y,126841
 airbyte_cdk/sources/declarative/declarative_source.py,sha256=nF7wBqFd3AQmEKAm4CnIo29CJoQL562cJGSCeL8U8bA,1531
 airbyte_cdk/sources/declarative/declarative_stream.py,sha256=JRyNeOIpsFu4ztVZsN6sncqUEIqIE-bUkD2TPgbMgk0,10375
 airbyte_cdk/sources/declarative/decoders/__init__.py,sha256=hNlhaB5FjNC6IfJyglj5ZJWkYD2nEAukMDmzRz5PC6o,671
@@ -104,12 +104,12 @@ airbyte_cdk/sources/declarative/migrations/__init__.py,sha256=47DEQpj8HBSa-_TImW
 airbyte_cdk/sources/declarative/migrations/legacy_to_per_partition_state_migration.py,sha256=iemy3fKLczcU0-Aor7tx5jcT6DRedKMqyK7kCOp01hg,3924
 airbyte_cdk/sources/declarative/migrations/state_migration.py,sha256=KWPjealMLKSMtajXgkdGgKg7EmTLR-CqqD7UIh0-eDU,794
 airbyte_cdk/sources/declarative/models/__init__.py,sha256=nUFxNCiKeYRVXuZEKA7GD-lTHxsiKcQ8FitZjKhPIvE,100
-airbyte_cdk/sources/declarative/models/declarative_component_schema.py,sha256=6iAzpGmUrhwEUQcCL5bW-FXuLXPMeFqs_GR4B1rS3ZE,88511
+airbyte_cdk/sources/declarative/models/declarative_component_schema.py,sha256=qjzXe162aUcaq1n6B8KhA6Z9B8boM9yY8dAsLXll5-g,89872
 airbyte_cdk/sources/declarative/parsers/__init__.py,sha256=ZnqYNxHsKCgO38IwB34RQyRMXTs4GTvlRi3ImKnIioo,61
 airbyte_cdk/sources/declarative/parsers/custom_exceptions.py,sha256=Rir9_z3Kcd5Es0-LChrzk-0qubAsiK_RSEnLmK2OXm8,553
-airbyte_cdk/sources/declarative/parsers/manifest_component_transformer.py,sha256=0jfi-ogL-rOVORTIYnu64wNfh1L8fYaLVDWzJ2zGdi8,8799
+airbyte_cdk/sources/declarative/parsers/manifest_component_transformer.py,sha256=KflzFl_ZKRSW9XkH16sfr5-9HnAAI0T5s8CVBrJK2Ao,8958
 airbyte_cdk/sources/declarative/parsers/manifest_reference_resolver.py,sha256=IWUOdF03o-aQn0Occo1BJCxU0Pz-QILk5L67nzw2thw,6803
-airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py,sha256=lS2oKfkND54u66hocc2BycS-AIYIbkn4npq6CFRNokc,99573
+airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py,sha256=kGU0re3R-Ujtn6Gp96KpEVRHJB5P-B8sRa0aMR_jDdk,102536
 airbyte_cdk/sources/declarative/partition_routers/__init__.py,sha256=1NjaZoGAIefvWwj6wx-LOKIXXWS-UnBlZFnuR7y6uYA,745
 airbyte_cdk/sources/declarative/partition_routers/cartesian_product_stream_slicer.py,sha256=c5cuVFM6NFkuQqG8Z5IwkBuwDrvXZN1CunUOM_L0ezg,6892
 airbyte_cdk/sources/declarative/partition_routers/list_partition_router.py,sha256=t7pRdFWfFWJtQQG19c9PVeMODyO2BknRTakpM5U9N-8,4844
@@ -159,8 +159,9 @@ airbyte_cdk/sources/declarative/retrievers/__init__.py,sha256=FVQpUGVwp2Gibk4gp0
 airbyte_cdk/sources/declarative/retrievers/async_retriever.py,sha256=WDFnjrXLz3-YEjFhmlMkWAn9AJvnZ0mk9FyC8DAhEYk,4976
 airbyte_cdk/sources/declarative/retrievers/retriever.py,sha256=XPLs593Xv8c5cKMc37XzUAYmzlXd1a7eSsspM-CMuWA,1696
 airbyte_cdk/sources/declarative/retrievers/simple_retriever.py,sha256=N4swGw5mfuTXJ2R7AKX18CHzizsr69pXwt5uSHLPi48,24172
-airbyte_cdk/sources/declarative/schema/__init__.py,sha256=ul8L9S0-__AMEdbCLHBq-PMEeA928NVp8BB83BMotfU,517
+airbyte_cdk/sources/declarative/schema/__init__.py,sha256=Io9vninzlEjQ2uFmWklxfwNM0cXfljtzOz5zL1OVyT4,701
 airbyte_cdk/sources/declarative/schema/default_schema_loader.py,sha256=KTACrIE23a83wsm3Rd9Eb4K6-20lrGqYxTHNp9yxsso,1820
+airbyte_cdk/sources/declarative/schema/dynamic_schema_loader.py,sha256=tP5DIEMn-k2JshWeXmo53ZEudDAVb4AJ50Z5tfme_ZU,8063
 airbyte_cdk/sources/declarative/schema/inline_schema_loader.py,sha256=bVETE10hRsatRJq3R3BeyRR0wIoK3gcP1gcpVRQ_P5U,464
 airbyte_cdk/sources/declarative/schema/json_file_schema_loader.py,sha256=5Wl-fqW-pVf_dxJ4yGHMAFfC4JjKHYJhqFJT1xA57F4,4177
 airbyte_cdk/sources/declarative/schema/schema_loader.py,sha256=kjt8v0N5wWKA5zyLnrDLxf1PJKdUqvQq2RVnAOAzNSY,379
@@ -333,8 +334,8 @@ airbyte_cdk/utils/slice_hasher.py,sha256=-pHexlNYoWYPnXNH-M7HEbjmeJe9Zk7SJijdQ7d
 airbyte_cdk/utils/spec_schema_transformations.py,sha256=-5HTuNsnDBAhj-oLeQXwpTGA0HdcjFOf2zTEMUTTg_Y,816
 airbyte_cdk/utils/stream_status_utils.py,sha256=ZmBoiy5HVbUEHAMrUONxZvxnvfV9CesmQJLDTAIWnWw,1171
 airbyte_cdk/utils/traced_exception.py,sha256=C8uIBuCL_E4WnBAOPSxBicD06JAldoN9fGsQDp463OY,6292
-airbyte_cdk-6.9.0.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
-airbyte_cdk-6.9.0.dist-info/METADATA,sha256=zfXBcI2WKOmEe38cVGYmW8cngDts67fzR94zJQ3oNrU,5949
-airbyte_cdk-6.9.0.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
-airbyte_cdk-6.9.0.dist-info/entry_points.txt,sha256=fj-e3PAQvsxsQzyyq8UkG1k8spunWnD4BAH2AwlR6NM,95
-airbyte_cdk-6.9.0.dist-info/RECORD,,
+airbyte_cdk-6.9.0.dev0.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
+airbyte_cdk-6.9.0.dev0.dist-info/METADATA,sha256=VS_vbZgaobGxk6eCGkFZuQgCR-xzMrHzJWL2FpCnyM8,6112
+airbyte_cdk-6.9.0.dev0.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
+airbyte_cdk-6.9.0.dev0.dist-info/entry_points.txt,sha256=fj-e3PAQvsxsQzyyq8UkG1k8spunWnD4BAH2AwlR6NM,95
+airbyte_cdk-6.9.0.dev0.dist-info/RECORD,,