airbyte-cdk 6.7.1rc3__py3-none-any.whl → 6.7.2__py3-none-any.whl

airbyte_cdk/sources/declarative/requesters/error_handlers/default_error_handler.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, List, Mapping, MutableMapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.requesters.error_handlers.default_http_response_filter import (
     DefaultHttpResponseFilter,
 )

airbyte_cdk/sources/declarative/requesters/error_handlers/default_http_response_filter.py

@@ -5,6 +5,7 @@
 from typing import Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.requesters.error_handlers.http_response_filter import (
     HttpResponseFilter,
 )

airbyte_cdk/sources/declarative/requesters/error_handlers/http_response_filter.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass
 from typing import Any, Mapping, Optional, Set, Union
 
 import requests
+
 from airbyte_cdk.models import FailureType
 from airbyte_cdk.sources.declarative.interpolation import InterpolatedString
 from airbyte_cdk.sources.declarative.interpolation.interpolated_boolean import InterpolatedBoolean

airbyte_cdk/sources/declarative/requesters/http_job_repository.py

@@ -6,6 +6,8 @@ from datetime import timedelta
 from typing import Any, Dict, Iterable, Mapping, Optional
 
 import requests
+from requests import Response
+
 from airbyte_cdk import AirbyteMessage
 from airbyte_cdk.logger import lazy_log
 from airbyte_cdk.models import FailureType, Type
@@ -23,7 +25,6 @@ from airbyte_cdk.sources.declarative.requesters.requester import Requester
 from airbyte_cdk.sources.declarative.retrievers.simple_retriever import SimpleRetriever
 from airbyte_cdk.sources.types import Record, StreamSlice
 from airbyte_cdk.utils import AirbyteTracedException
-from requests import Response
 
 LOGGER = logging.getLogger("airbyte")
 
@@ -41,7 +42,7 @@ class AsyncHttpJobRepository(AsyncJobRepository):
 
     job_timeout: Optional[timedelta] = None
     record_extractor: RecordExtractor = field(
-        init=False, repr=False, default_factory=lambda: ResponseToFileExtractor()
+        init=False, repr=False, default_factory=lambda: ResponseToFileExtractor({})
     )
 
     def __post_init__(self) -> None:

airbyte_cdk/sources/declarative/requesters/http_requester.py

@@ -9,6 +9,7 @@ from typing import Any, Callable, Mapping, MutableMapping, Optional, Union
 from urllib.parse import urljoin
 
 import requests
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import (
     DeclarativeAuthenticator,
     NoAuth,

airbyte_cdk/sources/declarative/requesters/paginators/default_paginator.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, Mapping, MutableMapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.decoders import (
     Decoder,
     JsonDecoder,

airbyte_cdk/sources/declarative/requesters/paginators/no_pagination.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass
 from typing import Any, Mapping, MutableMapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.requesters.paginators.paginator import Paginator
 from airbyte_cdk.sources.types import Record, StreamSlice, StreamState
 

airbyte_cdk/sources/declarative/requesters/paginators/paginator.py

@@ -7,6 +7,7 @@ from dataclasses import dataclass
 from typing import Any, Mapping, Optional
 
 import requests
+
 from airbyte_cdk.sources.declarative.requesters.request_options.request_options_provider import (
     RequestOptionsProvider,
 )

airbyte_cdk/sources/declarative/requesters/paginators/strategies/cursor_pagination_strategy.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, Dict, Mapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.decoders import (
     Decoder,
     JsonDecoder,

airbyte_cdk/sources/declarative/requesters/paginators/strategies/offset_increment.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, Mapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.decoders import (
     Decoder,
     JsonDecoder,

airbyte_cdk/sources/declarative/requesters/paginators/strategies/page_increment.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass
 from typing import Any, Mapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.interpolation import InterpolatedString
 from airbyte_cdk.sources.declarative.requesters.paginators.strategies.pagination_strategy import (
     PaginationStrategy,

airbyte_cdk/sources/declarative/requesters/paginators/strategies/pagination_strategy.py

@@ -7,6 +7,7 @@ from dataclasses import dataclass
 from typing import Any, Optional
 
 import requests
+
 from airbyte_cdk.sources.types import Record
 
 

airbyte_cdk/sources/declarative/requesters/paginators/strategies/stop_condition.py

@@ -6,10 +6,12 @@ from abc import ABC, abstractmethod
 from typing import Any, Optional
 
 import requests
+
 from airbyte_cdk.sources.declarative.incremental.declarative_cursor import DeclarativeCursor
 from airbyte_cdk.sources.declarative.requesters.paginators.strategies.pagination_strategy import (
     PaginationStrategy,
 )
+from airbyte_cdk.sources.streams.concurrent.cursor import ConcurrentCursor
 from airbyte_cdk.sources.types import Record
 
 
@@ -25,7 +27,11 @@ class PaginationStopCondition(ABC):
 
 
 class CursorStopCondition(PaginationStopCondition):
-    def __init__(self, cursor: DeclarativeCursor):
+    def __init__(
+        self,
+        cursor: DeclarativeCursor
+        | ConcurrentCursor,  # migrate to use both old and concurrent versions
+    ):
         self._cursor = cursor
 
     def is_met(self, record: Record) -> bool:
@@ -46,8 +52,8 @@ class StopConditionPaginationStrategyDecorator(PaginationStrategy):
             return None
         return self._delegate.next_page_token(response, last_page_size, last_record)
 
-    def reset(self) -> None:
-        self._delegate.reset()
+    def reset(self, reset_value: Optional[Any] = None) -> None:
+        self._delegate.reset(reset_value)
 
     def get_page_size(self) -> Optional[int]:
         return self._delegate.get_page_size()

airbyte_cdk/sources/declarative/requesters/request_options/interpolated_request_options_provider.py

@@ -5,6 +5,8 @@
 from dataclasses import InitVar, dataclass, field
 from typing import Any, Mapping, MutableMapping, Optional, Union
 
+from deprecated import deprecated
+
 from airbyte_cdk.sources.declarative.interpolation.interpolated_nested_mapping import NestedMapping
 from airbyte_cdk.sources.declarative.requesters.request_options.interpolated_nested_request_input_provider import (
     InterpolatedNestedRequestInputProvider,
@@ -17,7 +19,6 @@ from airbyte_cdk.sources.declarative.requesters.request_options.request_options_
 )
 from airbyte_cdk.sources.source import ExperimentalClassWarning
 from airbyte_cdk.sources.types import Config, StreamSlice, StreamState
-from deprecated import deprecated
 
 RequestInput = Union[str, Mapping[str, str]]
 ValidRequestTypes = (str, list)

airbyte_cdk/sources/declarative/requesters/requester.py

@@ -7,6 +7,7 @@ from enum import Enum
 from typing import Any, Callable, Mapping, MutableMapping, Optional, Union
 
 import requests
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
 from airbyte_cdk.sources.declarative.requesters.request_options.request_options_provider import (
     RequestOptionsProvider,

airbyte_cdk/sources/declarative/retrievers/async_retriever.py

@@ -4,6 +4,8 @@
 from dataclasses import InitVar, dataclass, field
 from typing import Any, Callable, Iterable, Mapping, Optional
 
+from deprecated.classic import deprecated
+
 from airbyte_cdk.models import FailureType
 from airbyte_cdk.sources.declarative.async_job.job_orchestrator import (
     AsyncJobOrchestrator,
@@ -17,7 +19,6 @@ from airbyte_cdk.sources.source import ExperimentalClassWarning
 from airbyte_cdk.sources.streams.core import StreamData
 from airbyte_cdk.sources.types import Config, StreamSlice, StreamState
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from deprecated.classic import deprecated
 
 
 @deprecated("This class is experimental. Use at your own risk.", category=ExperimentalClassWarning)

airbyte_cdk/sources/declarative/retrievers/simple_retriever.py

@@ -20,6 +20,7 @@ from typing import (
 )
 
 import requests
+
 from airbyte_cdk.models import AirbyteMessage
 from airbyte_cdk.sources.declarative.extractors.http_selector import HttpSelector
 from airbyte_cdk.sources.declarative.incremental import ResumableFullRefreshCursor
@@ -467,8 +468,9 @@ class SimpleRetriever(Retriever):
         else:
             return None
 
-    @staticmethod
-    def _extract_record(stream_data: StreamData, stream_slice: StreamSlice) -> Optional[Record]:
+    def _extract_record(
+        self, stream_data: StreamData, stream_slice: StreamSlice
+    ) -> Optional[Record]:
         """
         As we allow the output of _read_pages to be StreamData, it can be multiple things. Therefore, we need to filter out and normalize
         to data to streamline the rest of the process.
@@ -477,9 +479,15 @@ class SimpleRetriever(Retriever):
             # Record is not part of `StreamData` but is the most common implementation of `Mapping[str, Any]` which is part of `StreamData`
             return stream_data
         elif isinstance(stream_data, (dict, Mapping)):
-            return Record(dict(stream_data), stream_slice)
+            return Record(
+                data=dict(stream_data), associated_slice=stream_slice, stream_name=self.name
+            )
         elif isinstance(stream_data, AirbyteMessage) and stream_data.record:
-            return Record(stream_data.record.data, stream_slice)
+            return Record(
+                data=stream_data.record.data,  # type:ignore # AirbyteMessage always has record.data
+                associated_slice=stream_slice,
+                stream_name=self.name,
+            )
         return None
 
     # stream_slices is defined with arguments on http stream and fixing this has a long tail of dependencies. Will be resolved by the decoupling of http stream and simple retriever

airbyte_cdk/sources/declarative/stream_slicers/declarative_partition_generator.py

@@ -1,14 +1,13 @@
 # Copyright (c) 2024 Airbyte, Inc., all rights reserved.
 
-from typing import Iterable, Optional, Mapping, Any, Callable
+from typing import Any, Callable, Iterable, Mapping, Optional
 
 from airbyte_cdk.sources.declarative.retrievers import Retriever
 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
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
 from airbyte_cdk.sources.streams.concurrent.partitions.stream_slicer import StreamSlicer
-from airbyte_cdk.sources.types import StreamSlice
+from airbyte_cdk.sources.types import Record, StreamSlice
 from airbyte_cdk.utils.slice_hasher import SliceHasher
 
 
@@ -59,7 +58,11 @@ class DeclarativePartition(Partition):
     def read(self) -> Iterable[Record]:
         for stream_data in self._retriever.read_records(self._json_schema, self._stream_slice):
             if isinstance(stream_data, Mapping):
-                yield Record(stream_data, self)
+                yield Record(
+                    data=stream_data,
+                    stream_name=self.stream_name(),
+                    associated_slice=self._stream_slice,
+                )
             else:
                 self._message_repository.emit_message(stream_data)
 

airbyte_cdk/sources/declarative/transformations/add_fields.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, Dict, List, Mapping, Optional, Type, Union
 
 import dpath
+
 from airbyte_cdk.sources.declarative.interpolation.interpolated_string import InterpolatedString
 from airbyte_cdk.sources.declarative.transformations import RecordTransformation
 from airbyte_cdk.sources.types import Config, FieldPointer, StreamSlice, StreamState

airbyte_cdk/sources/declarative/transformations/remove_fields.py

@@ -7,6 +7,7 @@ from typing import Any, Dict, List, Mapping, Optional
 
 import dpath
 import dpath.exceptions
+
 from airbyte_cdk.sources.declarative.interpolation.interpolated_boolean import InterpolatedBoolean
 from airbyte_cdk.sources.declarative.transformations import RecordTransformation
 from airbyte_cdk.sources.types import Config, FieldPointer, StreamSlice, StreamState

airbyte_cdk/sources/declarative/yaml_declarative_source.py

@@ -6,6 +6,7 @@ import pkgutil
 from typing import Any, List, Mapping, Optional
 
 import yaml
+
 from airbyte_cdk.models import AirbyteStateMessage, ConfiguredAirbyteCatalog
 from airbyte_cdk.sources.declarative.concurrent_declarative_source import (
     ConcurrentDeclarativeSource,

airbyte_cdk/sources/embedded/tools.py

@@ -5,6 +5,7 @@
 from typing import Any, Callable, Dict, Iterable, Optional
 
 import dpath
+
 from airbyte_cdk.models import AirbyteStream
 
 

airbyte_cdk/sources/file_based/config/abstract_file_based_spec.py

@@ -7,10 +7,11 @@ from abc import abstractmethod
 from typing import Any, Dict, List, Literal, Optional, Union
 
 import dpath
+from pydantic.v1 import AnyUrl, BaseModel, Field
+
 from airbyte_cdk import OneOfOptionConfig
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import FileBasedStreamConfig
 from airbyte_cdk.sources.utils import schema_helpers
-from pydantic.v1 import AnyUrl, BaseModel, Field
 
 
 class DeliverRecords(BaseModel):

airbyte_cdk/sources/file_based/config/avro_format.py

@@ -3,9 +3,10 @@
 #
 
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class AvroFormat(BaseModel):
     class Config(OneOfOptionConfig):

airbyte_cdk/sources/file_based/config/csv_format.py

@@ -6,10 +6,11 @@ import codecs
 from enum import Enum
 from typing import Any, Dict, List, Optional, Set, Union
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field, root_validator, validator
 from pydantic.v1.error_wrappers import ValidationError
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class InferenceType(Enum):
     NONE = "None"

airbyte_cdk/sources/file_based/config/excel_format.py

@@ -2,9 +2,10 @@
 # Copyright (c) 2024 Airbyte, Inc., all rights reserved.
 #
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class ExcelFormat(BaseModel):
     class Config(OneOfOptionConfig):

airbyte_cdk/sources/file_based/config/file_based_stream_config.py

@@ -5,6 +5,8 @@
 from enum import Enum
 from typing import Any, List, Mapping, Optional, Union
 
+from pydantic.v1 import BaseModel, Field, validator
+
 from airbyte_cdk.sources.file_based.config.avro_format import AvroFormat
 from airbyte_cdk.sources.file_based.config.csv_format import CsvFormat
 from airbyte_cdk.sources.file_based.config.excel_format import ExcelFormat
@@ -13,7 +15,6 @@ from airbyte_cdk.sources.file_based.config.parquet_format import ParquetFormat
 from airbyte_cdk.sources.file_based.config.unstructured_format import UnstructuredFormat
 from airbyte_cdk.sources.file_based.exceptions import ConfigValidationError, FileBasedSourceError
 from airbyte_cdk.sources.file_based.schema_helpers import type_mapping_to_jsonschema
-from pydantic.v1 import BaseModel, Field, validator
 
 PrimaryKeyType = Optional[Union[str, List[str]]]
 

airbyte_cdk/sources/file_based/config/jsonl_format.py

@@ -2,9 +2,10 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class JsonlFormat(BaseModel):
     class Config(OneOfOptionConfig):

airbyte_cdk/sources/file_based/config/parquet_format.py

@@ -3,9 +3,10 @@
 #
 
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class ParquetFormat(BaseModel):
     class Config(OneOfOptionConfig):

airbyte_cdk/sources/file_based/config/unstructured_format.py

@@ -4,9 +4,10 @@
 
 from typing import List, Literal, Optional, Union
 
-from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from pydantic.v1 import BaseModel, Field
 
+from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
+
 
 class LocalProcessingConfigModel(BaseModel):
     mode: Literal["local"] = Field("local", const=True)

airbyte_cdk/sources/file_based/file_based_source.py

@@ -8,6 +8,8 @@ from abc import ABC
 from collections import Counter
 from typing import Any, Iterator, List, Mapping, Optional, Tuple, Type, Union
 
+from pydantic.v1.error_wrappers import ValidationError
+
 from airbyte_cdk.logger import AirbyteLogFormatter, init_logger
 from airbyte_cdk.models import (
     AirbyteMessage,
@@ -60,7 +62,6 @@ from airbyte_cdk.sources.streams import Stream
 from airbyte_cdk.sources.streams.concurrent.cursor import CursorField
 from airbyte_cdk.utils.analytics_message import create_analytics_message
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from pydantic.v1.error_wrappers import ValidationError
 
 DEFAULT_CONCURRENCY = 100
 MAX_CONCURRENCY = 100

airbyte_cdk/sources/file_based/file_based_stream_reader.py

@@ -10,9 +10,10 @@ from io import IOBase
 from os import makedirs, path
 from typing import Any, Dict, Iterable, List, Optional, Set
 
+from wcmatch.glob import GLOBSTAR, globmatch
+
 from airbyte_cdk.sources.file_based.config.abstract_file_based_spec import AbstractFileBasedSpec
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
-from wcmatch.glob import GLOBSTAR, globmatch
 
 
 class FileReadMode(Enum):

airbyte_cdk/sources/file_based/file_types/avro_parser.py

@@ -6,6 +6,7 @@ import logging
 from typing import Any, Dict, Iterable, Mapping, Optional, Tuple
 
 import fastavro
+
 from airbyte_cdk.sources.file_based.config.avro_format import AvroFormat
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import FileBasedStreamConfig
 from airbyte_cdk.sources.file_based.exceptions import FileBasedSourceError, RecordParseError

airbyte_cdk/sources/file_based/file_types/csv_parser.py

@@ -12,6 +12,8 @@ from io import IOBase
 from typing import Any, Callable, Dict, Generator, Iterable, List, Mapping, Optional, Set, Tuple
 from uuid import uuid4
 
+from orjson import orjson
+
 from airbyte_cdk.models import FailureType
 from airbyte_cdk.sources.file_based.config.csv_format import (
     CsvFormat,
@@ -29,7 +31,6 @@ from airbyte_cdk.sources.file_based.file_types.file_type_parser import FileTypeP
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
 from airbyte_cdk.sources.file_based.schema_helpers import TYPE_PYTHON_MAPPING, SchemaType
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from orjson import orjson
 
 DIALECT_NAME = "_config_dialect"
 

airbyte_cdk/sources/file_based/file_types/excel_parser.py

@@ -8,6 +8,11 @@ from pathlib import Path
 from typing import Any, Dict, Iterable, Mapping, Optional, Tuple, Union
 
 import pandas as pd
+from numpy import datetime64, issubdtype
+from numpy import dtype as dtype_
+from orjson import orjson
+from pydantic.v1 import BaseModel
+
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import (
     ExcelFormat,
     FileBasedStreamConfig,
@@ -24,11 +29,6 @@ from airbyte_cdk.sources.file_based.file_based_stream_reader import (
 from airbyte_cdk.sources.file_based.file_types.file_type_parser import FileTypeParser
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
 from airbyte_cdk.sources.file_based.schema_helpers import SchemaType
-from numpy import datetime64
-from numpy import dtype as dtype_
-from numpy import issubdtype
-from orjson import orjson
-from pydantic.v1 import BaseModel
 
 
 class ExcelParser(FileTypeParser):

airbyte_cdk/sources/file_based/file_types/jsonl_parser.py

@@ -6,6 +6,8 @@ import json
 import logging
 from typing import Any, Dict, Iterable, Mapping, Optional, Tuple, Union
 
+from orjson import orjson
+
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import FileBasedStreamConfig
 from airbyte_cdk.sources.file_based.exceptions import FileBasedSourceError, RecordParseError
 from airbyte_cdk.sources.file_based.file_based_stream_reader import (
@@ -19,7 +21,6 @@ from airbyte_cdk.sources.file_based.schema_helpers import (
     SchemaType,
     merge_schemas,
 )
-from orjson import orjson
 
 
 class JsonlParser(FileTypeParser):

airbyte_cdk/sources/file_based/file_types/parquet_parser.py

@@ -10,6 +10,8 @@ from urllib.parse import unquote
 
 import pyarrow as pa
 import pyarrow.parquet as pq
+from pyarrow import DictionaryArray, Scalar
+
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import (
     FileBasedStreamConfig,
     ParquetFormat,
@@ -26,7 +28,6 @@ from airbyte_cdk.sources.file_based.file_based_stream_reader import (
 from airbyte_cdk.sources.file_based.file_types.file_type_parser import FileTypeParser
 from airbyte_cdk.sources.file_based.remote_file import RemoteFile
 from airbyte_cdk.sources.file_based.schema_helpers import SchemaType
-from pyarrow import DictionaryArray, Scalar
 
 
 class ParquetParser(FileTypeParser):

airbyte_cdk/sources/file_based/file_types/unstructured_parser.py

@@ -9,7 +9,16 @@ from typing import Any, Dict, Iterable, List, Mapping, Optional, Tuple, Union
 
 import backoff
 import dpath
+import nltk
 import requests
+from unstructured.file_utils.filetype import (
+    EXT_TO_FILETYPE,
+    FILETYPE_TO_MIMETYPE,
+    STR_TO_FILETYPE,
+    FileType,
+    detect_filetype,
+)
+
 from airbyte_cdk.models import FailureType
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import FileBasedStreamConfig
 from airbyte_cdk.sources.file_based.config.unstructured_format import (
@@ -28,14 +37,6 @@ from airbyte_cdk.sources.file_based.remote_file import RemoteFile
 from airbyte_cdk.sources.file_based.schema_helpers import SchemaType
 from airbyte_cdk.utils import is_cloud_environment
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from unstructured.file_utils.filetype import (
-    EXT_TO_FILETYPE,
-    FILETYPE_TO_MIMETYPE,
-    STR_TO_FILETYPE,
-    FileType,
-    detect_filetype,
-)
-import nltk
 
 unstructured_partition_pdf = None
 unstructured_partition_docx = None

airbyte_cdk/sources/file_based/stream/abstract_file_based_stream.py

@@ -6,6 +6,8 @@ from abc import abstractmethod
 from functools import cache, cached_property, lru_cache
 from typing import Any, Dict, Iterable, List, Mapping, Optional, Type
 
+from deprecated import deprecated
+
 from airbyte_cdk import AirbyteMessage
 from airbyte_cdk.models import SyncMode
 from airbyte_cdk.sources.file_based.availability_strategy import (
@@ -30,7 +32,6 @@ from airbyte_cdk.sources.file_based.stream.cursor import AbstractFileBasedCursor
 from airbyte_cdk.sources.file_based.types import StreamSlice
 from airbyte_cdk.sources.streams import Stream
 from airbyte_cdk.sources.streams.checkpoint import Cursor
-from deprecated import deprecated
 
 
 class AbstractFileBasedStream(Stream):

airbyte_cdk/sources/file_based/stream/concurrent/adapters.py

@@ -7,6 +7,8 @@ import logging
 from functools import cache, lru_cache
 from typing import TYPE_CHECKING, Any, Iterable, List, Mapping, MutableMapping, Optional, Union
 
+from deprecated.classic import deprecated
+
 from airbyte_cdk.models import (
     AirbyteLogMessage,
     AirbyteMessage,
@@ -39,11 +41,10 @@ from airbyte_cdk.sources.streams.concurrent.helpers import (
 )
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
 from airbyte_cdk.sources.streams.concurrent.partitions.partition_generator import PartitionGenerator
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
 from airbyte_cdk.sources.streams.core import StreamData
+from airbyte_cdk.sources.types import Record
 from airbyte_cdk.sources.utils.schema_helpers import InternalConfig
 from airbyte_cdk.sources.utils.slice_logger import SliceLogger
-from deprecated.classic import deprecated
 
 if TYPE_CHECKING:
     from airbyte_cdk.sources.file_based.stream.concurrent.cursor import (
@@ -247,7 +248,7 @@ class FileBasedStreamPartition(Partition):
                     self._stream.transformer.transform(
                         data_to_return, self._stream.get_json_schema()
                     )
-                    yield Record(data_to_return, self)
+                    yield Record(data=data_to_return, stream_name=self.stream_name())
                 elif (
                     isinstance(record_data, AirbyteMessage)
                     and record_data.type == Type.RECORD
@@ -265,7 +266,7 @@ class FileBasedStreamPartition(Partition):
                     else:
                         yield Record(
                             data=record_message_data,
-                            partition=self,
+                            stream_name=self.stream_name(),
                             is_file_transfer_message=self._use_file_transfer(),
                         )
                 else:

airbyte_cdk/sources/file_based/stream/concurrent/cursor/abstract_concurrent_file_based_cursor.py

@@ -12,7 +12,7 @@ from airbyte_cdk.sources.file_based.stream.cursor import AbstractFileBasedCursor
 from airbyte_cdk.sources.file_based.types import StreamState
 from airbyte_cdk.sources.streams.concurrent.cursor import Cursor
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
+from airbyte_cdk.sources.types import Record
 
 if TYPE_CHECKING:
     from airbyte_cdk.sources.file_based.stream.concurrent.adapters import FileBasedStreamPartition

airbyte_cdk/sources/file_based/stream/concurrent/cursor/file_based_concurrent_cursor.py

@@ -19,7 +19,7 @@ from airbyte_cdk.sources.file_based.types import StreamState
 from airbyte_cdk.sources.message.repository import MessageRepository
 from airbyte_cdk.sources.streams.concurrent.cursor import CursorField
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
+from airbyte_cdk.sources.types import Record
 
 if TYPE_CHECKING:
     from airbyte_cdk.sources.file_based.stream.concurrent.adapters import FileBasedStreamPartition