airbyte-cdk 6.60.13__py3-none-any.whl → 6.60.14__py3-none-any.whl

airbyte_cdk/connector_builder/test_reader/reader.py

@@ -120,7 +120,11 @@ class TestReader:
         deprecation_warnings: List[LogMessage] = source.deprecation_warnings()
 
         schema_inferrer = SchemaInferrer(
-            self._pk_to_nested_and_composite_field(stream.primary_key) if stream else None,
+            self._pk_to_nested_and_composite_field(
+                stream.primary_key if hasattr(stream, "primary_key") else stream._primary_key  # type: ignore  # We are accessing the private property here as the primary key is not exposed. We should either expose it or use `as_airbyte_stream` to retrieve it as this is the "official" way where it is exposed in the Airbyte protocol
+            )
+            if stream
+            else None,
             self._cursor_field_to_nested_and_composite_field(stream.cursor_field)
             if stream
             else None,

airbyte_cdk/sources/declarative/concurrent_declarative_source.py

@@ -3,7 +3,7 @@
 #
 
 import logging
-from typing import Any, Generic, Iterator, List, Mapping, MutableMapping, Optional, Tuple
+from typing import Any, Generic, Iterator, List, Mapping, MutableMapping, Optional, Tuple, Union
 
 from airbyte_cdk.models import (
     AirbyteCatalog,
@@ -15,10 +15,6 @@ from airbyte_cdk.sources.concurrent_source.concurrent_source import ConcurrentSo
 from airbyte_cdk.sources.connector_state_manager import ConnectorStateManager
 from airbyte_cdk.sources.declarative.concurrency_level import ConcurrencyLevel
 from airbyte_cdk.sources.declarative.declarative_stream import DeclarativeStream
-from airbyte_cdk.sources.declarative.extractors import RecordSelector
-from airbyte_cdk.sources.declarative.extractors.record_filter import (
-    ClientSideIncrementalRecordFilterDecorator,
-)
 from airbyte_cdk.sources.declarative.incremental import (
     ConcurrentPerPartitionCursor,
     GlobalSubstreamCursor,
@@ -28,7 +24,6 @@ from airbyte_cdk.sources.declarative.incremental.per_partition_with_global impor
     PerPartitionWithGlobalCursor,
 )
 from airbyte_cdk.sources.declarative.manifest_declarative_source import ManifestDeclarativeSource
-from airbyte_cdk.sources.declarative.models import FileUploader
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     ConcurrencyLevel as ConcurrencyLevelModel,
 )
@@ -84,7 +79,6 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
         # incremental streams running in full refresh.
         component_factory = component_factory or ModelToComponentFactory(
             emit_connector_builder_messages=emit_connector_builder_messages,
-            disable_resumable_full_refresh=True,
             connector_state_manager=self._connector_state_manager,
             max_concurrent_async_job_count=source_config.get("max_concurrent_async_job_count"),
         )
@@ -180,7 +174,7 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
             ]
         )
 
-    def streams(self, config: Mapping[str, Any]) -> List[Stream]:
+    def streams(self, config: Mapping[str, Any]) -> List[Union[Stream, AbstractStream]]:  # type: ignore  # we are migrating away from the AbstractSource and are expecting that this will only be called by ConcurrentDeclarativeSource or the Connector Builder
         """
         The `streams` method is used as part of the AbstractSource in the following cases:
         * ConcurrentDeclarativeSource.check -> ManifestDeclarativeSource.check -> AbstractSource.check -> DeclarativeSource.check_connection -> CheckStream.check_connection -> streams
@@ -210,6 +204,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
             # these legacy Python streams the way we do low-code streams to determine if they are concurrent compatible,
             # so we need to treat them as synchronous
 
+            if isinstance(declarative_stream, AbstractStream):
+                concurrent_streams.append(declarative_stream)
+                continue
+
             supports_file_transfer = (
                 isinstance(declarative_stream, DeclarativeStream)
                 and "file_uploader" in name_to_stream_mapping[declarative_stream.name]
@@ -278,10 +276,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
 
                         partition_generator = StreamSlicerPartitionGenerator(
                             partition_factory=DeclarativePartitionFactory(
-                                declarative_stream.name,
-                                declarative_stream.get_json_schema(),
-                                retriever,
-                                self.message_repository,
+                                stream_name=declarative_stream.name,
+                                schema_loader=declarative_stream._schema_loader,  # type: ignore  # We are accessing the private property but the public one is optional and we will remove this code soonish
+                                retriever=retriever,
+                                message_repository=self.message_repository,
                             ),
                             stream_slicer=declarative_stream.retriever.stream_slicer,
                         )
@@ -309,10 +307,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                             )
                         partition_generator = StreamSlicerPartitionGenerator(
                             partition_factory=DeclarativePartitionFactory(
-                                declarative_stream.name,
-                                declarative_stream.get_json_schema(),
-                                retriever,
-                                self.message_repository,
+                                stream_name=declarative_stream.name,
+                                schema_loader=declarative_stream._schema_loader,  # type: ignore  # We are accessing the private property but the public one is optional and we will remove this code soonish
+                                retriever=retriever,
+                                message_repository=self.message_repository,
                             ),
                             stream_slicer=cursor,
                         )
@@ -339,10 +337,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                 ) and hasattr(declarative_stream.retriever, "stream_slicer"):
                     partition_generator = StreamSlicerPartitionGenerator(
                         DeclarativePartitionFactory(
-                            declarative_stream.name,
-                            declarative_stream.get_json_schema(),
-                            declarative_stream.retriever,
-                            self.message_repository,
+                            stream_name=declarative_stream.name,
+                            schema_loader=declarative_stream._schema_loader,  # type: ignore  # We are accessing the private property but the public one is optional and we will remove this code soonish
+                            retriever=declarative_stream.retriever,
+                            message_repository=self.message_repository,
                         ),
                         declarative_stream.retriever.stream_slicer,
                     )
@@ -399,10 +397,10 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
 
                     partition_generator = StreamSlicerPartitionGenerator(
                         DeclarativePartitionFactory(
-                            declarative_stream.name,
-                            declarative_stream.get_json_schema(),
-                            retriever,
-                            self.message_repository,
+                            stream_name=declarative_stream.name,
+                            schema_loader=declarative_stream._schema_loader,  # type: ignore  # We are accessing the private property but the public one is optional and we will remove this code soonish
+                            retriever=retriever,
+                            message_repository=self.message_repository,
                         ),
                         perpartition_cursor,
                     )

airbyte_cdk/sources/declarative/manifest_declarative_source.py

@@ -8,7 +8,7 @@ import pkgutil
 from copy import deepcopy
 from importlib import metadata
 from types import ModuleType
-from typing import Any, Dict, Iterator, List, Mapping, Optional, Set
+from typing import Any, Dict, Iterator, List, Mapping, Optional, Set, Union
 
 import orjson
 import yaml
@@ -66,6 +66,7 @@ from airbyte_cdk.sources.declarative.parsers.model_to_component_factory import (
 from airbyte_cdk.sources.declarative.resolvers import COMPONENTS_RESOLVER_TYPE_MAPPING
 from airbyte_cdk.sources.declarative.spec.spec import Spec
 from airbyte_cdk.sources.message import MessageRepository
+from airbyte_cdk.sources.streams.concurrent.abstract_stream import AbstractStream
 from airbyte_cdk.sources.streams.core import Stream
 from airbyte_cdk.sources.types import Config, ConnectionDefinition
 from airbyte_cdk.sources.utils.slice_logger import (
@@ -297,7 +298,12 @@ class ManifestDeclarativeSource(DeclarativeSource):
                 f"Expected to generate a ConnectionChecker component, but received {check_stream.__class__}"
             )
 
-    def streams(self, config: Mapping[str, Any]) -> List[Stream]:
+    def streams(self, config: Mapping[str, Any]) -> List[Union[Stream, AbstractStream]]:  # type: ignore  # we are migrating away from the AbstractSource and are expecting that this will only be called by ConcurrentDeclarativeSource or the Connector Builder
+        """
+        As a migration step, this method will return both legacy stream (Stream) and concurrent stream (AbstractStream).
+        Once the migration is done, we can probably have this method throw "not implemented" as we figure out how to
+        fully decouple this from the AbstractSource.
+        """
         if self._spec_component:
             self._spec_component.validate_config(config)
 

airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py

@@ -7,6 +7,7 @@ from __future__ import annotations
 import datetime
 import importlib
 import inspect
+import logging
 import re
 from functools import partial
 from typing import (
@@ -543,6 +544,10 @@ from airbyte_cdk.sources.declarative.stream_slicers import (
     StreamSlicer,
     StreamSlicerTestReadDecorator,
 )
+from airbyte_cdk.sources.declarative.stream_slicers.declarative_partition_generator import (
+    DeclarativePartitionFactory,
+    StreamSlicerPartitionGenerator,
+)
 from airbyte_cdk.sources.declarative.transformations import (
     AddFields,
     RecordTransformation,
@@ -594,6 +599,7 @@ from airbyte_cdk.sources.streams.call_rate import (
     Rate,
     UnlimitedCallRatePolicy,
 )
+from airbyte_cdk.sources.streams.concurrent.abstract_stream import AbstractStream
 from airbyte_cdk.sources.streams.concurrent.clamping import (
     ClampingEndProvider,
     ClampingStrategy,
@@ -603,7 +609,14 @@ from airbyte_cdk.sources.streams.concurrent.clamping import (
     WeekClampingStrategy,
     Weekday,
 )
-from airbyte_cdk.sources.streams.concurrent.cursor import ConcurrentCursor, Cursor, CursorField
+from airbyte_cdk.sources.streams.concurrent.cursor import (
+    ConcurrentCursor,
+    Cursor,
+    CursorField,
+    FinalStateCursor,
+)
+from airbyte_cdk.sources.streams.concurrent.default_stream import DefaultStream
+from airbyte_cdk.sources.streams.concurrent.helpers import get_primary_key_from_stream
 from airbyte_cdk.sources.streams.concurrent.state_converters.datetime_stream_state_converter import (
     CustomFormatConcurrentStreamStateConverter,
     DateTimeStreamStateConverter,
@@ -633,7 +646,6 @@ class ModelToComponentFactory:
         emit_connector_builder_messages: bool = False,
         disable_retries: bool = False,
         disable_cache: bool = False,
-        disable_resumable_full_refresh: bool = False,
         message_repository: Optional[MessageRepository] = None,
         connector_state_manager: Optional[ConnectorStateManager] = None,
         max_concurrent_async_job_count: Optional[int] = None,
@@ -644,7 +656,6 @@ class ModelToComponentFactory:
         self._emit_connector_builder_messages = emit_connector_builder_messages
         self._disable_retries = disable_retries
         self._disable_cache = disable_cache
-        self._disable_resumable_full_refresh = disable_resumable_full_refresh
         self._message_repository = message_repository or InMemoryMessageRepository(
             self._evaluate_log_level(emit_connector_builder_messages)
         )
@@ -1920,8 +1931,8 @@ class ModelToComponentFactory:
         )
 
     def create_declarative_stream(
-        self, model: DeclarativeStreamModel, config: Config, **kwargs: Any
-    ) -> DeclarativeStream:
+        self, model: DeclarativeStreamModel, config: Config, is_parent: bool = False, **kwargs: Any
+    ) -> Union[DeclarativeStream, AbstractStream]:
         # When constructing a declarative stream, we assemble the incremental_sync component and retriever's partition_router field
         # components if they exist into a single CartesianProductStreamSlicer. This is then passed back as an argument when constructing the
         # Retriever. This is done in the declarative stream not the retriever to support custom retrievers. The custom create methods in
@@ -2027,15 +2038,6 @@ class ModelToComponentFactory:
             file_uploader=file_uploader,
             incremental_sync=model.incremental_sync,
         )
-        cursor_field = model.incremental_sync.cursor_field if model.incremental_sync else None
-
-        if model.state_migrations:
-            state_transformations = [
-                self._create_component_from_model(state_migration, config, declarative_stream=model)
-                for state_migration in model.state_migrations
-            ]
-        else:
-            state_transformations = []
 
         schema_loader: Union[
             CompositeSchemaLoader,
@@ -2063,6 +2065,56 @@ class ModelToComponentFactory:
                 options["name"] = model.name
             schema_loader = DefaultSchemaLoader(config=config, parameters=options)
 
+        if (
+            isinstance(combined_slicers, PartitionRouter)
+            and not self._emit_connector_builder_messages
+            and not is_parent
+        ):
+            # We are starting to migrate streams to instantiate directly the DefaultStream instead of instantiating the
+            # DeclarativeStream and assembling the DefaultStream from that. The plan is the following:
+            # * Streams without partition router nor cursors and streams with only partition router. This is the `isinstance(combined_slicers, PartitionRouter)` condition as the first kind with have a SinglePartitionRouter
+            # * Streams without partition router but with cursor
+            # * Streams with both partition router and cursor
+            # We specifically exclude parent streams here because SubstreamPartitionRouter has not been updated yet
+            # We specifically exclude Connector Builder stuff for now as Brian is working on this anyway
+            stream_name = model.name or ""
+            partition_generator = StreamSlicerPartitionGenerator(
+                DeclarativePartitionFactory(
+                    stream_name,
+                    schema_loader,
+                    retriever,
+                    self._message_repository,
+                ),
+                stream_slicer=cast(
+                    StreamSlicer,
+                    StreamSlicerTestReadDecorator(
+                        wrapped_slicer=combined_slicers,
+                        maximum_number_of_slices=self._limit_slices_fetched or 5,
+                    ),
+                ),
+            )
+            return DefaultStream(
+                partition_generator=partition_generator,
+                name=stream_name,
+                json_schema=schema_loader.get_json_schema,
+                primary_key=get_primary_key_from_stream(primary_key),
+                cursor_field=None,
+                # FIXME we should have the cursor field has part of the interface of cursor
+                logger=logging.getLogger(f"airbyte.{stream_name}"),
+                # FIXME this is a breaking change compared to the old implementation,
+                cursor=FinalStateCursor(stream_name, None, self._message_repository),
+                supports_file_transfer=hasattr(model, "file_uploader")
+                and bool(model.file_uploader),
+            )
+
+        cursor_field = model.incremental_sync.cursor_field if model.incremental_sync else None
+        if model.state_migrations:
+            state_transformations = [
+                self._create_component_from_model(state_migration, config, declarative_stream=model)
+                for state_migration in model.state_migrations
+            ]
+        else:
+            state_transformations = []
         return DeclarativeStream(
             name=model.name or "",
             primary_key=primary_key,
@@ -2083,7 +2135,7 @@ class ModelToComponentFactory:
         ],
         config: Config,
         stream_name: Optional[str] = None,
-    ) -> Optional[PartitionRouter]:
+    ) -> PartitionRouter:
         if (
             hasattr(model, "partition_router")
             and isinstance(model, SimpleRetrieverModel | AsyncRetrieverModel)
@@ -2104,7 +2156,7 @@ class ModelToComponentFactory:
                 return self._create_component_from_model(  # type: ignore[no-any-return] # Will be created PartitionRouter as stream_slicer_model is model.partition_router
                     model=stream_slicer_model, config=config, stream_name=stream_name or ""
                 )
-        return None
+        return SinglePartitionRouter(parameters={})
 
     def _build_incremental_cursor(
         self,
@@ -2121,7 +2173,9 @@ class ModelToComponentFactory:
             else []
         )
 
-        if model.incremental_sync and stream_slicer:
+        if model.incremental_sync and (
+            stream_slicer and not isinstance(stream_slicer, SinglePartitionRouter)
+        ):
             if model.retriever.type == "AsyncRetriever":
                 stream_name = model.name or ""
                 stream_namespace = None
@@ -2194,7 +2248,11 @@ class ModelToComponentFactory:
         else:
             state_transformations = []
 
-        if model.incremental_sync and stream_slicer:
+        if (
+            model.incremental_sync
+            and stream_slicer
+            and not isinstance(stream_slicer, SinglePartitionRouter)
+        ):
             return self.create_concurrent_cursor_from_perpartition_cursor(  # type: ignore # This is a known issue that we are creating and returning a ConcurrentCursor which does not technically implement the (low-code) StreamSlicer. However, (low-code) StreamSlicer and ConcurrentCursor both implement StreamSlicer.stream_slices() which is the primary method needed for checkpointing
                 state_manager=self._connector_state_manager,
                 model_type=DatetimeBasedCursorModel,
@@ -2233,28 +2291,6 @@ class ModelToComponentFactory:
                 )
         return None
 
-    def _build_resumable_cursor(
-        self,
-        model: Union[
-            AsyncRetrieverModel,
-            CustomRetrieverModel,
-            SimpleRetrieverModel,
-        ],
-        stream_slicer: Optional[PartitionRouter],
-    ) -> Optional[StreamSlicer]:
-        if hasattr(model, "paginator") and model.paginator and not stream_slicer:
-            # For the regular Full-Refresh streams, we use the high level `ResumableFullRefreshCursor`
-            return ResumableFullRefreshCursor(parameters={})
-        elif stream_slicer:
-            # For the Full-Refresh sub-streams, we use the nested `ChildPartitionResumableFullRefreshCursor`
-            return PerPartitionCursor(
-                cursor_factory=CursorFactory(
-                    create_function=partial(ChildPartitionResumableFullRefreshCursor, {})
-                ),
-                partition_router=stream_slicer,
-            )
-        return None
-
     def _merge_stream_slicers(
         self, model: DeclarativeStreamModel, config: Config
     ) -> Optional[StreamSlicer]:
@@ -2291,11 +2327,7 @@ class ModelToComponentFactory:
         if model.incremental_sync:
             return self._build_incremental_cursor(model, stream_slicer, config)
 
-        return (
-            stream_slicer
-            if self._disable_resumable_full_refresh
-            else self._build_resumable_cursor(retriever_model, stream_slicer)
-        )
+        return stream_slicer
 
     def create_default_error_handler(
         self, model: DefaultErrorHandlerModel, config: Config, **kwargs: Any
@@ -2577,9 +2609,6 @@ class ModelToComponentFactory:
     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(model.retriever, stream_slicer)
-
         schema_transformations = []
         if model.schema_transformations:
             for transformation_model in model.schema_transformations:
@@ -2592,7 +2621,7 @@ class ModelToComponentFactory:
             config=config,
             name=name,
             primary_key=None,
-            stream_slicer=combined_slicers,
+            stream_slicer=self._build_stream_slicer_from_partition_router(model.retriever, config),
             transformations=[],
             use_cache=True,
             log_formatter=(
@@ -2945,7 +2974,10 @@ class ModelToComponentFactory:
         self, model: ParentStreamConfigModel, config: Config, **kwargs: Any
     ) -> ParentStreamConfig:
         declarative_stream = self._create_component_from_model(
-            model.stream, config=config, **kwargs
+            model.stream,
+            config=config,
+            is_parent=True,
+            **kwargs,
         )
         request_option = (
             self._create_component_from_model(model.request_option, config=config)
@@ -3855,15 +3887,12 @@ class ModelToComponentFactory:
     def create_http_components_resolver(
         self, model: HttpComponentsResolverModel, config: Config, stream_name: Optional[str] = None
     ) -> Any:
-        stream_slicer = self._build_stream_slicer_from_partition_router(model.retriever, config)
-        combined_slicers = self._build_resumable_cursor(model.retriever, stream_slicer)
-
         retriever = self._create_component_from_model(
             model=model.retriever,
             config=config,
             name=f"{stream_name if stream_name else '__http_components_resolver'}",
             primary_key=None,
-            stream_slicer=stream_slicer if stream_slicer else combined_slicers,
+            stream_slicer=self._build_stream_slicer_from_partition_router(model.retriever, config),
             transformations=[],
         )
 

airbyte_cdk/sources/declarative/stream_slicers/declarative_partition_generator.py

@@ -3,6 +3,7 @@
 from typing import Any, Iterable, Mapping, Optional
 
 from airbyte_cdk.sources.declarative.retrievers import Retriever
+from airbyte_cdk.sources.declarative.schema import SchemaLoader
 from airbyte_cdk.sources.message import MessageRepository
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
 from airbyte_cdk.sources.streams.concurrent.partitions.partition_generator import PartitionGenerator
@@ -11,11 +12,23 @@ from airbyte_cdk.sources.types import Record, StreamSlice
 from airbyte_cdk.utils.slice_hasher import SliceHasher
 
 
+class SchemaLoaderCachingDecorator(SchemaLoader):
+    def __init__(self, schema_loader: SchemaLoader):
+        self._decorated = schema_loader
+        self._loaded_schema: Optional[Mapping[str, Any]] = None
+
+    def get_json_schema(self) -> Mapping[str, Any]:
+        if self._loaded_schema is None:
+            self._loaded_schema = self._decorated.get_json_schema()
+
+        return self._loaded_schema  # type: ignore  # at that point, we assume the schema will be populated
+
+
 class DeclarativePartitionFactory:
     def __init__(
         self,
         stream_name: str,
-        json_schema: Mapping[str, Any],
+        schema_loader: SchemaLoader,
         retriever: Retriever,
         message_repository: MessageRepository,
     ) -> None:
@@ -25,17 +38,17 @@ class DeclarativePartitionFactory:
         In order to avoid these problems, we will create one retriever per thread which should make the processing thread-safe.
         """
         self._stream_name = stream_name
-        self._json_schema = json_schema
+        self._schema_loader = SchemaLoaderCachingDecorator(schema_loader)
         self._retriever = retriever
         self._message_repository = message_repository
 
     def create(self, stream_slice: StreamSlice) -> Partition:
         return DeclarativePartition(
-            self._stream_name,
-            self._json_schema,
-            self._retriever,
-            self._message_repository,
-            stream_slice,
+            stream_name=self._stream_name,
+            schema_loader=self._schema_loader,
+            retriever=self._retriever,
+            message_repository=self._message_repository,
+            stream_slice=stream_slice,
         )
 
 
@@ -43,20 +56,22 @@ class DeclarativePartition(Partition):
     def __init__(
         self,
         stream_name: str,
-        json_schema: Mapping[str, Any],
+        schema_loader: SchemaLoader,
         retriever: Retriever,
         message_repository: MessageRepository,
         stream_slice: StreamSlice,
     ):
         self._stream_name = stream_name
-        self._json_schema = json_schema
+        self._schema_loader = schema_loader
         self._retriever = retriever
         self._message_repository = message_repository
         self._stream_slice = stream_slice
         self._hash = SliceHasher.hash(self._stream_name, self._stream_slice)
 
     def read(self) -> Iterable[Record]:
-        for stream_data in self._retriever.read_records(self._json_schema, self._stream_slice):
+        for stream_data in self._retriever.read_records(
+            self._schema_loader.get_json_schema(), self._stream_slice
+        ):
             if isinstance(stream_data, Mapping):
                 record = (
                     stream_data

airbyte_cdk/sources/streams/concurrent/adapters.py

@@ -6,7 +6,7 @@ import copy
 import json
 import logging
 from functools import lru_cache
-from typing import Any, Iterable, List, Mapping, MutableMapping, Optional, Tuple, Union
+from typing import Any, Iterable, List, Mapping, MutableMapping, Optional, Union
 
 from typing_extensions import deprecated
 
@@ -196,6 +196,7 @@ class StreamFacade(AbstractStreamFacade[DefaultStream], Stream):
     def cursor(self) -> Optional[Cursor]:  # type: ignore[override] # StreamFaced expects to use only airbyte_cdk.sources.streams.concurrent.cursor.Cursor
         return self._cursor
 
+    # FIXME the lru_cache seems to be mostly there because of typing issue
     @lru_cache(maxsize=None)
     def get_json_schema(self) -> Mapping[str, Any]:
         return self._abstract_stream.get_json_schema()

airbyte_cdk/sources/streams/concurrent/default_stream.py

@@ -2,9 +2,8 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 
-from functools import lru_cache
 from logging import Logger
-from typing import Any, Iterable, List, Mapping, Optional
+from typing import Any, Callable, Iterable, List, Mapping, Optional, Union
 
 from airbyte_cdk.models import AirbyteStream, SyncMode
 from airbyte_cdk.sources.streams.concurrent.abstract_stream import AbstractStream
@@ -20,7 +19,7 @@ class DefaultStream(AbstractStream):
         self,
         partition_generator: PartitionGenerator,
         name: str,
-        json_schema: Mapping[str, Any],
+        json_schema: Union[Mapping[str, Any], Callable[[], Mapping[str, Any]]],
         primary_key: List[str],
         cursor_field: Optional[str],
         logger: Logger,
@@ -53,14 +52,13 @@ class DefaultStream(AbstractStream):
     def cursor_field(self) -> Optional[str]:
         return self._cursor_field
 
-    @lru_cache(maxsize=None)
     def get_json_schema(self) -> Mapping[str, Any]:
-        return self._json_schema
+        return self._json_schema() if callable(self._json_schema) else self._json_schema
 
     def as_airbyte_stream(self) -> AirbyteStream:
         stream = AirbyteStream(
             name=self.name,
-            json_schema=dict(self._json_schema),
+            json_schema=dict(self.get_json_schema()),
             supported_sync_modes=[SyncMode.full_refresh],
             is_resumable=False,
             is_file_based=self._supports_file_transfer,

{airbyte_cdk-6.60.13.dist-info → airbyte_cdk-6.60.14.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 6.60.13
+Version: 6.60.14
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT

{airbyte_cdk-6.60.13.dist-info → airbyte_cdk-6.60.14.dist-info}/RECORD

@@ -21,7 +21,7 @@ airbyte_cdk/connector_builder/models.py,sha256=9pIZ98LW_d6fRS39VdnUOf3cxGt4TkC5M
 airbyte_cdk/connector_builder/test_reader/__init__.py,sha256=iTwBMoI9vaJotEgpqZbFjlxRcbxXYypSVJ9YxeHk7wc,120
 airbyte_cdk/connector_builder/test_reader/helpers.py,sha256=vqoHpZeQ0BLIw2NiTNGXr0euA8gI_X0pcNRcHOv8sHM,27942
 airbyte_cdk/connector_builder/test_reader/message_grouper.py,sha256=LDNl-xFQwA4RsUpn7684KbWaVH-SWWBIwhHvIgduLTE,7090
-airbyte_cdk/connector_builder/test_reader/reader.py,sha256=3jLy3tUUHkG1rmGWrZuo4SmPYNVD9oiAqy8mdaUwzvo,21301
+airbyte_cdk/connector_builder/test_reader/reader.py,sha256=DugoqS6SMrtOJ--2Y0F0h_9x8m632i7fSOPMAA0JHnc,21654
 airbyte_cdk/connector_builder/test_reader/types.py,sha256=hPZG3jO03kBaPyW94NI3JHRS1jxXGSNBcN1HFzOxo5Y,2528
 airbyte_cdk/destinations/__init__.py,sha256=FyDp28PT_YceJD5HDFhA-mrGfX9AONIyMQ4d68CHNxQ,213
 airbyte_cdk/destinations/destination.py,sha256=CIq-yb8C_0QvcKCtmStaHfiqn53GEfRAIGGCkJhKP1Q,5880
@@ -86,7 +86,7 @@ airbyte_cdk/sources/declarative/checks/check_stream.py,sha256=sV-ZY7dZ03V8GdAxPY
 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=_C0IKcPlumc3HAGxccmeyS8lSQo7HxPr4jcNFok_9oU,26637
+airbyte_cdk/sources/declarative/concurrent_declarative_source.py,sha256=HQxvDEoMWtVdVRZgJylrT0YLx-R8sOgICjY3HnifvWs,27391
 airbyte_cdk/sources/declarative/datetime/__init__.py,sha256=4Hw-PX1-VgESLF16cDdvuYCzGJtHntThLF4qIiULWeo,61
 airbyte_cdk/sources/declarative/datetime/datetime_parser.py,sha256=_zGNGq31RNy_0QBLt_EcTvgPyhj7urPdx6oA3M5-r3o,3150
 airbyte_cdk/sources/declarative/datetime/min_max_datetime.py,sha256=0BHBtDNQZfvwM45-tY5pNlTcKAFSGGNxemoi0Jic-0E,5785
@@ -128,7 +128,7 @@ airbyte_cdk/sources/declarative/interpolation/interpolated_string.py,sha256=CQkH
 airbyte_cdk/sources/declarative/interpolation/interpolation.py,sha256=9IoeuWam3L6GyN10L6U8xNWXmkt9cnahSDNkez1OmFY,982
 airbyte_cdk/sources/declarative/interpolation/jinja.py,sha256=oFGKs3oX0xO6DOL4E9x8rhxwbEoRcgx4HJVIL1RQ9c4,7269
 airbyte_cdk/sources/declarative/interpolation/macros.py,sha256=RpsAYG75bW0js2fQCzAN1nf3oeGyXwyt0LhJCHnlaUA,6031
-airbyte_cdk/sources/declarative/manifest_declarative_source.py,sha256=NNywOq7i0N0oEBvZWbKmo09jHEZVsrY_TAiYR-P4m5k,26558
+airbyte_cdk/sources/declarative/manifest_declarative_source.py,sha256=VqR3lti_RLRRe0_1EwUn8_OsJTxQrGqU3n-T9GowAKk,27154
 airbyte_cdk/sources/declarative/migrations/__init__.py,sha256=47DEQpj8HBSa-_TImW-5JCeuQeRkm5NMpJWZG3hSuFU,0
 airbyte_cdk/sources/declarative/migrations/legacy_to_per_partition_state_migration.py,sha256=V2lpYE9LJKvz6BUViHk4vaRGndxNABmPbDCtyYdkqaE,4013
 airbyte_cdk/sources/declarative/migrations/state_migration.py,sha256=KWPjealMLKSMtajXgkdGgKg7EmTLR-CqqD7UIh0-eDU,794
@@ -141,7 +141,7 @@ airbyte_cdk/sources/declarative/parsers/custom_exceptions.py,sha256=wnRUP0Xeru9R
 airbyte_cdk/sources/declarative/parsers/manifest_component_transformer.py,sha256=2UdpCz3yi7ISZTyqkQXSSy3dMxeyOWqV7OlAS5b9GVg,11568
 airbyte_cdk/sources/declarative/parsers/manifest_normalizer.py,sha256=EtKjS9c94yNp3AwQC8KUCQaAYW5T3zvFYxoWYjc_buI,19729
 airbyte_cdk/sources/declarative/parsers/manifest_reference_resolver.py,sha256=pJmg78vqE5VfUrF_KJnWjucQ4k9IWFULeAxHCowrHXE,6806
-airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py,sha256=zr4oBy1MCbsJtOf4gffEQOFo5wX5oq9lhIpVUmyNJoE,181452
+airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py,sha256=3oJ7JdtlpkzeTB5PAu0FRH0iVfBlk2wcFGB1U_0owdI,183073
 airbyte_cdk/sources/declarative/partition_routers/__init__.py,sha256=TBC9AkGaUqHm2IKHMPN6punBIcY5tWGULowcLoAVkfw,1109
 airbyte_cdk/sources/declarative/partition_routers/async_job_partition_router.py,sha256=VelO7zKqKtzMJ35jyFeg0ypJLQC0plqqIBNXoBW1G2E,3001
 airbyte_cdk/sources/declarative/partition_routers/cartesian_product_stream_slicer.py,sha256=c5cuVFM6NFkuQqG8Z5IwkBuwDrvXZN1CunUOM_L0ezg,6892
@@ -220,7 +220,7 @@ airbyte_cdk/sources/declarative/schema/schema_loader.py,sha256=kjt8v0N5wWKA5zyLn
 airbyte_cdk/sources/declarative/spec/__init__.py,sha256=9FYO-fVOclrwjAW4qwRTbZRVopTc9rOaauAJfThdNCQ,177
 airbyte_cdk/sources/declarative/spec/spec.py,sha256=SwL_pfXZgcLYLJY-MAeFMHug9oYh2tOWjgG0C3DoLOY,3602
 airbyte_cdk/sources/declarative/stream_slicers/__init__.py,sha256=UX-cP_C-9FIFFPL9z8nuxu_rglssRsMOqQmQHN8FLB8,341
-airbyte_cdk/sources/declarative/stream_slicers/declarative_partition_generator.py,sha256=cjKGm4r438dd1GxrFHJ4aYrdzG2bkncnwaWxAwlXR3M,3585
+airbyte_cdk/sources/declarative/stream_slicers/declarative_partition_generator.py,sha256=Wk7P-Jpy3f3a59mwnc9ycJbpA3zVcgykNt2grBSXhBA,4272
 airbyte_cdk/sources/declarative/stream_slicers/stream_slicer.py,sha256=SOkIPBi2Wu7yxIvA15yFzUAB95a3IzA8LPq5DEqHQQc,725
 airbyte_cdk/sources/declarative/stream_slicers/stream_slicer_test_read_decorator.py,sha256=aUSleOw9elq3-5TaDUvp7H8W-2qUKqpr__kaJd8-ZFA,983
 airbyte_cdk/sources/declarative/transformations/__init__.py,sha256=CPJ8TlMpiUmvG3624VYu_NfTzxwKcfBjM2Q2wJ7fkSA,919
@@ -316,12 +316,12 @@ airbyte_cdk/sources/streams/concurrent/README.md,sha256=0nvgnlCBfZJiPDAofT8yFmUh
 airbyte_cdk/sources/streams/concurrent/__init__.py,sha256=4Hw-PX1-VgESLF16cDdvuYCzGJtHntThLF4qIiULWeo,61
 airbyte_cdk/sources/streams/concurrent/abstract_stream.py,sha256=aalik3FvyEjoeA1S3wUYEV3bgQLGrTnhYKPvT-rgy9E,3919
 airbyte_cdk/sources/streams/concurrent/abstract_stream_facade.py,sha256=QTry1QCBUwJDw1QSCEvz23s7zIEx_7QMxkPq9j-oPIQ,1358
-airbyte_cdk/sources/streams/concurrent/adapters.py,sha256=R9EZza1jF0e_HaLgN9Q_VREjvmlk0p9UfBLsnHB2y48,13936
+airbyte_cdk/sources/streams/concurrent/adapters.py,sha256=h4ZewhWn2PzPTt0lZZjcUL4rrpW9E_of7prnI3bm-c4,14004
 airbyte_cdk/sources/streams/concurrent/availability_strategy.py,sha256=M0XmvF3vjlr4GbCM0XH1hAj7udiAONM9SnmXjqufzLM,1035
 airbyte_cdk/sources/streams/concurrent/clamping.py,sha256=i26GVyui2ScEXSP-IP_61K2HaTp1-6lTlYHsZVYpuZA,3240
 airbyte_cdk/sources/streams/concurrent/cursor.py,sha256=xFFB8eEbtjGUdb42vkyWT5JB-WTUsaJlZ0gjKoVEycc,22307
 airbyte_cdk/sources/streams/concurrent/cursor_types.py,sha256=ZyWLPpeLX1qXcP5MwS-wxK11IBMsnVPCw9zx8gA2_Ro,843
-airbyte_cdk/sources/streams/concurrent/default_stream.py,sha256=cVBMwCM8OVlsLA39e1YN_eGb7qApqH-GQtnWMs19sGU,4687
+airbyte_cdk/sources/streams/concurrent/default_stream.py,sha256=SSufbo5f7OOYS8DZaABXeJVvodcfp9wb8J9lT5Xik3s,4744
 airbyte_cdk/sources/streams/concurrent/exceptions.py,sha256=JOZ446MCLpmF26r9KfS6OO_6rGjcjgJNZdcw6jccjEI,468
 airbyte_cdk/sources/streams/concurrent/helpers.py,sha256=S6AW8TgIASCZ2UuUcQLE8OzgYUHWt2-KPOvNPwnQf-Q,1596
 airbyte_cdk/sources/streams/concurrent/partition_enqueuer.py,sha256=2t64b_z9cEPmlHZnjSiMTO8PEtEdiAJDG0JcYOtUqAE,3363
@@ -424,9 +424,9 @@ airbyte_cdk/utils/slice_hasher.py,sha256=EDxgROHDbfG-QKQb59m7h_7crN1tRiawdf5uU7G
 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-6.60.13.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
-airbyte_cdk-6.60.13.dist-info/LICENSE_SHORT,sha256=aqF6D1NcESmpn-cqsxBtszTEnHKnlsp8L4x9wAh3Nxg,55
-airbyte_cdk-6.60.13.dist-info/METADATA,sha256=Zy-y3rjiuGIGzc3BLW-7F8zJJAxZMsX85KCPZHIxUXc,6478
-airbyte_cdk-6.60.13.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
-airbyte_cdk-6.60.13.dist-info/entry_points.txt,sha256=AKWbEkHfpzzk9nF9tqBUaw1MbvTM4mGtEzmZQm0ZWvM,139
-airbyte_cdk-6.60.13.dist-info/RECORD,,
+airbyte_cdk-6.60.14.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
+airbyte_cdk-6.60.14.dist-info/LICENSE_SHORT,sha256=aqF6D1NcESmpn-cqsxBtszTEnHKnlsp8L4x9wAh3Nxg,55
+airbyte_cdk-6.60.14.dist-info/METADATA,sha256=m-YkMQwaHLjes8d92IMFWcX4rh0zNCQcOkMK0VkCWzI,6478
+airbyte_cdk-6.60.14.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
+airbyte_cdk-6.60.14.dist-info/entry_points.txt,sha256=AKWbEkHfpzzk9nF9tqBUaw1MbvTM4mGtEzmZQm0ZWvM,139
+airbyte_cdk-6.60.14.dist-info/RECORD,,