airbyte-cdk 7.3.2.dev0__tar.gz → 7.3.3__tar.gz

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 7.3.2.dev0
+Version: 7.3.3
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py

@@ -122,6 +122,9 @@ from airbyte_cdk.sources.declarative.models.base_model_with_deprecations import
     DEPRECATION_LOGS_TAG,
     BaseModelWithDeprecations,
 )
+from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
+    Action1 as PaginationResetActionModel,
+)
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     AddedFieldDefinition as AddedFieldDefinitionModel,
 )
@@ -2213,6 +2216,14 @@ class ModelToComponentFactory:
             and stream_slicer
             and not isinstance(stream_slicer, SinglePartitionRouter)
         ):
+            if isinstance(model.incremental_sync, IncrementingCountCursorModel):
+                # We don't currently support usage of partition routing and IncrementingCountCursor at the
+                # same time because we didn't solve for design questions like what the lookback window would
+                # be as well as global cursor fall backs. We have not seen customers that have needed both
+                # at the same time yet and are currently punting on this until we need to solve it.
+                raise ValueError(
+                    f"The low-code framework does not currently support usage of a PartitionRouter and an IncrementingCountCursor at the same time. Please specify only one of these options for stream {stream_name}."
+                )
             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,
@@ -2400,21 +2411,12 @@ class ModelToComponentFactory:
 
         api_budget = self._api_budget
 
-        # Removes QueryProperties components from the interpolated mappings because it has been designed
-        # to be used by the SimpleRetriever and will be resolved from the provider from the slice directly
-        # instead of through jinja interpolation
-        request_parameters: Optional[Union[str, Mapping[str, str]]]
-        if isinstance(model.request_parameters, Mapping):
-            request_parameters = self._remove_query_properties(model.request_parameters)
-        else:
-            request_parameters = model.request_parameters
-
         request_options_provider = InterpolatedRequestOptionsProvider(
             request_body=model.request_body,
             request_body_data=model.request_body_data,
             request_body_json=model.request_body_json,
             request_headers=model.request_headers,
-            request_parameters=request_parameters,
+            request_parameters=model.request_parameters,  # type: ignore  # QueryProperties have been removed in `create_simple_retriever`
             query_properties_key=query_properties_key,
             config=config,
             parameters=model.parameters or {},
@@ -3208,7 +3210,8 @@ class ModelToComponentFactory:
 
         query_properties: Optional[QueryProperties] = None
         query_properties_key: Optional[str] = None
-        if self._query_properties_in_request_parameters(model.requester):
+        self._ensure_query_properties_to_model(model.requester)
+        if self._has_query_properties_in_request_parameters(model.requester):
             # It is better to be explicit about an error if PropertiesFromEndpoint is defined in multiple
             # places instead of default to request_parameters which isn't clearly documented
             if (
@@ -3220,7 +3223,7 @@ class ModelToComponentFactory:
                 )
 
             query_properties_definitions = []
-            for key, request_parameter in model.requester.request_parameters.items():  # type: ignore # request_parameters is already validated to be a Mapping using _query_properties_in_request_parameters()
+            for key, request_parameter in model.requester.request_parameters.items():  # type: ignore # request_parameters is already validated to be a Mapping using _has_query_properties_in_request_parameters()
                 if isinstance(request_parameter, QueryPropertiesModel):
                     query_properties_key = key
                     query_properties_definitions.append(request_parameter)
@@ -3234,6 +3237,16 @@ class ModelToComponentFactory:
                 query_properties = self._create_component_from_model(
                     model=query_properties_definitions[0], config=config
                 )
+
+            # Removes QueryProperties components from the interpolated mappings because it has been designed
+            # to be used by the SimpleRetriever and will be resolved from the provider from the slice directly
+            # instead of through jinja interpolation
+            if hasattr(model.requester, "request_parameters") and isinstance(
+                model.requester.request_parameters, Mapping
+            ):
+                model.requester.request_parameters = self._remove_query_properties(
+                    model.requester.request_parameters
+                )
         elif (
             hasattr(model.requester, "fetch_properties_from_endpoint")
             and model.requester.fetch_properties_from_endpoint
@@ -3333,7 +3346,7 @@ class ModelToComponentFactory:
             and model.pagination_reset
             and model.pagination_reset.limits
         ):
-            raise ValueError("PaginationResetLimits are not support while having record filter.")
+            raise ValueError("PaginationResetLimits are not supported while having record filter.")
 
         return SimpleRetriever(
             name=name,
@@ -3361,20 +3374,26 @@ class ModelToComponentFactory:
             return lambda: PaginationTracker()
 
         # Until we figure out a way to use any cursor for PaginationTracker, we will have to have this cursor selector logic
-        cursor_for_pagination_tracking = None
-        if isinstance(cursor, ConcurrentCursor):
-            cursor_for_pagination_tracking = cursor
-        elif isinstance(cursor, ConcurrentPerPartitionCursor):
-            cursor_for_pagination_tracking = cursor._cursor_factory.create(  # type: ignore  # if this becomes a problem, we would need to extract the cursor_factory instantiation logic and make it accessible here
-                {}, datetime.timedelta(0)
-            )
-        elif not isinstance(cursor, FinalStateCursor):
-            LOGGER.warning(
-                "Unknown cursor for PaginationTracker. Pagination resets might not work properly"
-            )
+        cursor_factory: Callable[[], Optional[ConcurrentCursor]] = lambda: None
+        if model.action == PaginationResetActionModel.RESET:
+            # in that case, we will let cursor_factory to return None even if the stream has a cursor
+            pass
+        elif model.action == PaginationResetActionModel.SPLIT_USING_CURSOR:
+            if isinstance(cursor, ConcurrentCursor):
+                cursor_factory = lambda: cursor.copy_without_state()  # type: ignore  # the if condition validates that it is a ConcurrentCursor
+            elif isinstance(cursor, ConcurrentPerPartitionCursor):
+                cursor_factory = lambda: cursor._cursor_factory.create(  # type: ignore  # if this becomes a problem, we would need to extract the cursor_factory instantiation logic and make it accessible here
+                    {}, datetime.timedelta(0)
+                )
+            elif not isinstance(cursor, FinalStateCursor):
+                LOGGER.warning(
+                    "Unknown cursor for PaginationTracker. Pagination resets might not work properly"
+                )
+        else:
+            raise ValueError(f"Unknown PaginationReset action: {model.action}")
 
         limit = model.limits.number_of_records if model and model.limits else None
-        return lambda: PaginationTracker(cursor_for_pagination_tracking, limit)
+        return lambda: PaginationTracker(cursor_factory(), limit)
 
     def _get_log_formatter(
         self, log_formatter: Callable[[Response], Any] | None, name: str
@@ -3402,7 +3421,7 @@ class ModelToComponentFactory:
         return bool(self._limit_slices_fetched or self._emit_connector_builder_messages)
 
     @staticmethod
-    def _query_properties_in_request_parameters(
+    def _has_query_properties_in_request_parameters(
         requester: Union[HttpRequesterModel, CustomRequesterModel],
     ) -> bool:
         if not hasattr(requester, "request_parameters"):
@@ -4216,3 +4235,26 @@ class ModelToComponentFactory:
             deduplicate=model.deduplicate if model.deduplicate is not None else True,
             config=config,
         )
+
+    def _ensure_query_properties_to_model(
+        self, requester: Union[HttpRequesterModel, CustomRequesterModel]
+    ) -> None:
+        """
+        For some reason, it seems like CustomRequesterModel request_parameters stays as dictionaries which means that
+        the other conditions relying on it being QueryPropertiesModel instead of a dict fail. Here, we migrate them to
+        proper model.
+        """
+        if not hasattr(requester, "request_parameters"):
+            return
+
+        request_parameters = requester.request_parameters
+        if request_parameters and isinstance(request_parameters, Dict):
+            for request_parameter_key in request_parameters.keys():
+                request_parameter = request_parameters[request_parameter_key]
+                if (
+                    isinstance(request_parameter, Dict)
+                    and request_parameter.get("type") == "QueryProperties"
+                ):
+                    request_parameters[request_parameter_key] = QueryPropertiesModel.parse_obj(
+                        request_parameter
+                    )

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/declarative/requesters/error_handlers/composite_error_handler.py

@@ -66,14 +66,14 @@ class CompositeErrorHandler(ErrorHandler):
             if not isinstance(matched_error_resolution, ErrorResolution):
                 continue
 
-            if matched_error_resolution.response_action == ResponseAction.SUCCESS:
+            if matched_error_resolution.response_action in [
+                ResponseAction.SUCCESS,
+                ResponseAction.RETRY,
+                ResponseAction.IGNORE,
+                ResponseAction.RESET_PAGINATION,
+            ]:
                 return matched_error_resolution
 
-            if (
-                matched_error_resolution.response_action == ResponseAction.RETRY
-                or matched_error_resolution.response_action == ResponseAction.IGNORE
-            ):
-                return matched_error_resolution
         if matched_error_resolution:
             return matched_error_resolution
 

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/declarative/retrievers/pagination_tracker.py

@@ -22,7 +22,7 @@ class PaginationTracker:
         """
         self._cursor = cursor
         self._limit = max_number_of_records
-        self.reset()
+        self._reset()
 
         """
         Given we have a cursor, we do not allow for the same slice to be processed twice because we assume we will
@@ -41,9 +41,8 @@ class PaginationTracker:
     def has_reached_limit(self) -> bool:
         return self._limit is not None and self._record_count >= self._limit
 
-    def reset(self) -> None:
+    def _reset(self) -> None:
         self._record_count = 0
-        self._number_of_attempt_with_same_slice = 0
 
     def reduce_slice_range_if_possible(self, stream_slice: StreamSlice) -> StreamSlice:
         new_slice = self._cursor.reduce_slice_range(stream_slice) if self._cursor else stream_slice
@@ -61,4 +60,5 @@ class PaginationTracker:
         else:
             self._number_of_attempt_with_same_slice = 0
 
+        self._reset()
         return new_slice

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/declarative/retrievers/simple_retriever.py

@@ -437,7 +437,6 @@ class SimpleRetriever(Retriever):
                     break
 
             if reset_pagination or pagination_tracker.has_reached_limit():
-                pagination_tracker.reset()
                 next_page_token = self._get_initial_next_page_token()
                 previous_slice = stream_slice
                 stream_slice = pagination_tracker.reduce_slice_range_if_possible(stream_slice)

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/file_based/file_based_stream_reader.py

@@ -3,13 +3,15 @@
 #
 
 import logging
+import time
 from abc import ABC, abstractmethod
 from datetime import datetime
 from enum import Enum
 from io import IOBase
 from os import makedirs, path
-from typing import Any, Callable, Iterable, List, MutableMapping, Optional, Set, Tuple
+from typing import Any, Iterable, List, MutableMapping, Optional, Set, Tuple
 
+from airbyte_protocol_dataclasses.models import FailureType
 from wcmatch.glob import GLOBSTAR, globmatch
 
 from airbyte_cdk.models import AirbyteRecordMessageFileReference
@@ -19,8 +21,9 @@ from airbyte_cdk.sources.file_based.config.validate_config_transfer_modes import
     preserve_directory_structure,
     use_file_transfer,
 )
+from airbyte_cdk.sources.file_based.exceptions import FileSizeLimitError
 from airbyte_cdk.sources.file_based.file_record_data import FileRecordData
-from airbyte_cdk.sources.file_based.remote_file import RemoteFile
+from airbyte_cdk.sources.file_based.remote_file import RemoteFile, UploadableRemoteFile
 
 
 class FileReadMode(Enum):
@@ -34,6 +37,7 @@ class AbstractFileBasedStreamReader(ABC):
     FILE_NAME = "file_name"
     LOCAL_FILE_PATH = "local_file_path"
     FILE_FOLDER = "file_folder"
+    FILE_SIZE_LIMIT = 1_500_000_000
 
     def __init__(self) -> None:
         self._config = None
@@ -113,16 +117,6 @@ class AbstractFileBasedStreamReader(ABC):
                     seen.add(file.uri)
                     yield file
 
-    @abstractmethod
-    def file_size(self, file: RemoteFile) -> int:
-        """Utility method to get size of the remote file.
-
-        This is required for connectors that will support writing to
-        files. If the connector does not support writing files, then the
-        subclass can simply `return 0`.
-        """
-        ...
-
     @staticmethod
     def file_matches_globs(file: RemoteFile, globs: List[str]) -> bool:
         # Use the GLOBSTAR flag to enable recursive ** matching
@@ -153,9 +147,8 @@ class AbstractFileBasedStreamReader(ABC):
             return include_identities_stream(self.config)
         return False
 
-    @abstractmethod
     def upload(
-        self, file: RemoteFile, local_directory: str, logger: logging.Logger
+        self, file: UploadableRemoteFile, local_directory: str, logger: logging.Logger
     ) -> Tuple[FileRecordData, AirbyteRecordMessageFileReference]:
         """
         This is required for connectors that will support writing to
@@ -173,7 +166,53 @@ class AbstractFileBasedStreamReader(ABC):
                    - file_size_bytes (int): The size of the referenced file in bytes.
                    - source_file_relative_path (str): The relative path to the referenced file in source.
         """
-        ...
+        if not isinstance(file, UploadableRemoteFile):
+            raise TypeError(f"Expected UploadableRemoteFile, got {type(file)}")
+
+        file_size = file.size
+
+        if file_size > self.FILE_SIZE_LIMIT:
+            message = f"File size exceeds the {self.FILE_SIZE_LIMIT / 1e9} GB limit."
+            raise FileSizeLimitError(
+                message=message, internal_message=message, failure_type=FailureType.config_error
+            )
+
+        file_paths = self._get_file_transfer_paths(
+            source_file_relative_path=file.source_file_relative_path,
+            staging_directory=local_directory,
+        )
+        local_file_path = file_paths[self.LOCAL_FILE_PATH]
+        file_relative_path = file_paths[self.FILE_RELATIVE_PATH]
+        file_name = file_paths[self.FILE_NAME]
+
+        logger.info(
+            f"Starting to download the file {file.file_uri_for_logging} with size: {file_size / (1024 * 1024):,.2f} MB ({file_size / (1024 * 1024 * 1024):.2f} GB)"
+        )
+        start_download_time = time.time()
+
+        file.download_to_local_directory(local_file_path)
+
+        write_duration = time.time() - start_download_time
+        logger.info(
+            f"Finished downloading the file {file.file_uri_for_logging} and saved to {local_file_path} in {write_duration:,.2f} seconds."
+        )
+
+        file_record_data = FileRecordData(
+            folder=file_paths[self.FILE_FOLDER],
+            file_name=file_name,
+            bytes=file_size,
+            id=file.id,
+            mime_type=file.mime_type,
+            created_at=file.created_at,
+            updated_at=file.updated_at,
+            source_uri=file.uri,
+        )
+        file_reference = AirbyteRecordMessageFileReference(
+            staging_file_url=local_file_path,
+            source_file_relative_path=file_relative_path,
+            file_size_bytes=file_size,
+        )
+        return file_record_data, file_reference
 
     def _get_file_transfer_paths(
         self, source_file_relative_path: str, staging_directory: str

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/file_based/file_types/file_transfer.py

@@ -7,7 +7,7 @@ from typing import Iterable, Tuple
 from airbyte_cdk.models import AirbyteRecordMessageFileReference
 from airbyte_cdk.sources.file_based.file_based_stream_reader import AbstractFileBasedStreamReader
 from airbyte_cdk.sources.file_based.file_record_data import FileRecordData
-from airbyte_cdk.sources.file_based.remote_file import RemoteFile
+from airbyte_cdk.sources.file_based.remote_file import UploadableRemoteFile
 from airbyte_cdk.sources.utils.files_directory import get_files_directory
 
 
@@ -17,7 +17,7 @@ class FileTransfer:
 
     def upload(
         self,
-        file: RemoteFile,
+        file: UploadableRemoteFile,
         stream_reader: AbstractFileBasedStreamReader,
         logger: logging.Logger,
     ) -> Iterable[Tuple[FileRecordData, AirbyteRecordMessageFileReference]]:

airbyte_cdk-7.3.3/airbyte_cdk/sources/file_based/remote_file.py

@@ -0,0 +1,57 @@
+#
+# Copyright (c) 2023 Airbyte, Inc., all rights reserved.
+#
+from abc import ABC, abstractmethod
+from datetime import datetime
+from typing import Optional
+
+from pydantic.v1 import BaseModel
+
+
+class RemoteFile(BaseModel):
+    """
+    A file in a file-based stream.
+    """
+
+    uri: str
+    last_modified: datetime
+    mime_type: Optional[str] = None
+
+
+class UploadableRemoteFile(RemoteFile, ABC):
+    """
+    A file in a file-based stream that supports uploading(file transferring).
+    """
+
+    id: Optional[str] = None
+    created_at: Optional[str] = None
+    updated_at: Optional[str] = None
+
+    @property
+    @abstractmethod
+    def size(self) -> int:
+        """
+        Returns the file size in bytes.
+        """
+        ...
+
+    @abstractmethod
+    def download_to_local_directory(self, local_file_path: str) -> None:
+        """
+        Download the file from remote source to local storage.
+        """
+        ...
+
+    @property
+    def source_file_relative_path(self) -> str:
+        """
+        Returns the relative path of the source file.
+        """
+        return self.uri
+
+    @property
+    def file_uri_for_logging(self) -> str:
+        """
+        Returns the URI for the file being logged.
+        """
+        return self.uri

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/sources/streams/concurrent/cursor.py

@@ -19,7 +19,7 @@ from typing import (
 )
 
 from airbyte_cdk.sources.connector_state_manager import ConnectorStateManager
-from airbyte_cdk.sources.message import MessageRepository
+from airbyte_cdk.sources.message import MessageRepository, NoopMessageRepository
 from airbyte_cdk.sources.streams import NO_CURSOR_STATE_KEY
 from airbyte_cdk.sources.streams.concurrent.clamping import ClampingStrategy, NoClamping
 from airbyte_cdk.sources.streams.concurrent.cursor_types import CursorValueType, GapType
@@ -136,6 +136,24 @@ class ConcurrentCursor(Cursor):
     _START_BOUNDARY = 0
     _END_BOUNDARY = 1
 
+    def copy_without_state(self) -> "ConcurrentCursor":
+        return self.__class__(
+            stream_name=self._stream_name,
+            stream_namespace=self._stream_namespace,
+            stream_state={},
+            message_repository=NoopMessageRepository(),
+            connector_state_manager=ConnectorStateManager(),
+            connector_state_converter=self._connector_state_converter,
+            cursor_field=self._cursor_field,
+            slice_boundary_fields=self._slice_boundary_fields,
+            start=self._start,
+            end_provider=self._end_provider,
+            lookback_window=self._lookback_window,
+            slice_range=self._slice_range,
+            cursor_granularity=self._cursor_granularity,
+            clamping_strategy=self._clamping_strategy,
+        )
+
     def __init__(
         self,
         stream_name: str,
@@ -528,17 +546,22 @@ class ConcurrentCursor(Cursor):
                 "Attempting to reduce slice while records are not returned in incremental order might lead to missing records"
             )
 
-        return StreamSlice(
-            partition=stream_slice.partition,
-            cursor_slice={
-                self._slice_boundary_fields_wrapper[
-                    self._START_BOUNDARY
-                ]: self._connector_state_converter.output_format(
-                    self._most_recent_cursor_value_per_partition[stream_slice]
-                ),
-                self._slice_boundary_fields_wrapper[self._END_BOUNDARY]: stream_slice.cursor_slice[
-                    self._slice_boundary_fields_wrapper[self._END_BOUNDARY]
-                ],
-            },
-            extra_fields=stream_slice.extra_fields,
-        )
+        if stream_slice in self._most_recent_cursor_value_per_partition:
+            return StreamSlice(
+                partition=stream_slice.partition,
+                cursor_slice={
+                    self._slice_boundary_fields_wrapper[
+                        self._START_BOUNDARY
+                    ]: self._connector_state_converter.output_format(
+                        self._most_recent_cursor_value_per_partition[stream_slice]
+                    ),
+                    self._slice_boundary_fields_wrapper[
+                        self._END_BOUNDARY
+                    ]: stream_slice.cursor_slice[
+                        self._slice_boundary_fields_wrapper[self._END_BOUNDARY]
+                    ],
+                },
+                extra_fields=stream_slice.extra_fields,
+            )
+        else:
+            return stream_slice

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/airbyte_cdk/test/mock_http/response_builder.py

@@ -75,6 +75,25 @@ class NestedPath(Path):
         return f"NestedPath(path={self._path})"
 
 
+class RootPath:
+    """
+    Path to use when the root of the response is an array.
+    """
+
+    def write(self, template: List[Dict[str, Any]], value: List[Dict[str, Any]]) -> None:
+        template.extend(value)
+
+    def update(self, template: List[Dict[str, Any]], value: List[Any]) -> None:
+        template.clear()
+        template.extend(value)
+
+    def extract(self, template: List[Dict[str, Any]]) -> Any:
+        return template
+
+    def __str__(self) -> str:
+        return f"RootPath"
+
+
 class PaginationStrategy(ABC):
     @abstractmethod
     def update(self, response: Dict[str, Any]) -> None:
@@ -149,12 +168,14 @@ class RecordBuilder:
 class HttpResponseBuilder:
     def __init__(
         self,
-        template: Dict[str, Any],
-        records_path: Union[FieldPath, NestedPath],
+        template: Union[Dict[str, Any], List[Dict[str, Any]]],
+        records_path: Union[FieldPath, NestedPath, RootPath],
         pagination_strategy: Optional[PaginationStrategy],
     ):
-        self._response = template
+        _validate_path_with_response(records_path, template)
+
         self._records: List[RecordBuilder] = []
+        self._response = template
         self._records_path = records_path
         self._pagination_strategy = pagination_strategy
         self._status_code = 200
@@ -169,6 +190,9 @@ class HttpResponseBuilder:
                 "`pagination_strategy` was not provided and hence, fields related to the pagination can't be modified. Please provide "
                 "`pagination_strategy` while instantiating ResponseBuilder to leverage this capability"
             )
+        elif isinstance(self._response, List):
+            raise ValueError("pagination_strategy requires the response to be a dict but was list")
+
         self._pagination_strategy.update(self._response)
         return self
 
@@ -177,7 +201,7 @@ class HttpResponseBuilder:
         return self
 
     def build(self) -> HttpResponse:
-        self._records_path.update(self._response, [record.build() for record in self._records])
+        self._records_path.update(self._response, [record.build() for record in self._records])  # type: ignore  # validated using _validate_path_with_response
         return HttpResponse(json.dumps(self._response), self._status_code)
 
 
@@ -208,15 +232,16 @@ def find_binary_response(resource: str, execution_folder: str) -> bytes:
 
 def create_record_builder(
     response_template: Dict[str, Any],
-    records_path: Union[FieldPath, NestedPath],
+    records_path: Union[FieldPath, NestedPath, RootPath],
     record_id_path: Optional[Path] = None,
     record_cursor_path: Optional[Union[FieldPath, NestedPath]] = None,
 ) -> RecordBuilder:
     """
     This will use the first record define at `records_path` as a template for the records. If more records are defined, they will be ignored
     """
+    _validate_path_with_response(records_path, response_template)
     try:
-        record_template = records_path.extract(response_template)[0]
+        record_template = records_path.extract(response_template)[0]  # type: ignore  # validated using _validate_path_with_response
         if not record_template:
             raise ValueError(
                 f"Could not extract any record from template at path `{records_path}`. "
@@ -230,8 +255,20 @@ def create_record_builder(
 
 
 def create_response_builder(
-    response_template: Dict[str, Any],
-    records_path: Union[FieldPath, NestedPath],
+    response_template: Union[Dict[str, Any], List[Dict[str, Any]]],
+    records_path: Union[FieldPath, NestedPath, RootPath],
     pagination_strategy: Optional[PaginationStrategy] = None,
 ) -> HttpResponseBuilder:
     return HttpResponseBuilder(response_template, records_path, pagination_strategy)
+
+
+def _validate_path_with_response(
+    records_path: Union[FieldPath, NestedPath, RootPath],
+    response_template: Union[Dict[str, Any], List[Dict[str, Any]]],
+) -> None:
+    if isinstance(response_template, List) and not isinstance(records_path, RootPath):
+        raise ValueError("templates of type lists require RootPath")
+    elif isinstance(response_template, Dict) and not isinstance(
+        records_path, (FieldPath, NestedPath)
+    ):
+        raise ValueError("templates of type dict either require FieldPath or NestedPath")

{airbyte_cdk-7.3.2.dev0 → airbyte_cdk-7.3.3}/pyproject.toml

@@ -23,7 +23,7 @@ classifiers = [
 keywords = ["airbyte", "connector-development-kit", "cdk"]
 
 # Python CDK uses dynamic versioning: https://github.com/mtkennerly/poetry-dynamic-versioning
-version = "7.3.2dev0"  # Version will be calculated dynamically.
+version = "7.3.3"  # Version will be calculated dynamically.
 
 [tool.poetry-dynamic-versioning]
 enable = false

airbyte_cdk-7.3.2.dev0/airbyte_cdk/sources/file_based/remote_file.py

@@ -1,18 +0,0 @@
-#
-# Copyright (c) 2023 Airbyte, Inc., all rights reserved.
-#
-
-from datetime import datetime
-from typing import Optional
-
-from pydantic.v1 import BaseModel
-
-
-class RemoteFile(BaseModel):
-    """
-    A file in a file-based stream.
-    """
-
-    uri: str
-    last_modified: datetime
-    mime_type: Optional[str] = None