airbyte-cdk 7.3.7.post3.dev18576192478__py3-none-any.whl → 7.3.9__py3-none-any.whl

airbyte_cdk/sources/declarative/extractors/response_to_file_extractor.py

@@ -7,11 +7,11 @@ import uuid
 import zlib
 from contextlib import closing
 from dataclasses import InitVar, dataclass
+from math import nan
 from typing import Any, Dict, Iterable, Mapping, Optional, Tuple
 
 import pandas as pd
 import requests
-from numpy import nan
 
 from airbyte_cdk.sources.declarative.extractors.record_extractor import RecordExtractor
 

airbyte_cdk/sources/file_based/exceptions.py

@@ -157,3 +157,7 @@ class CustomFileBasedException(AirbyteTracedException):
 
 class FileSizeLimitError(CustomFileBasedException):
     pass
+
+
+class EmptyFileSchemaInferenceError(AirbyteTracedException):
+    pass

airbyte_cdk/sources/file_based/file_types/csv_parser.py

@@ -22,7 +22,11 @@ from airbyte_cdk.sources.file_based.config.csv_format import (
     InferenceType,
 )
 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.exceptions import (
+    EmptyFileSchemaInferenceError,
+    FileBasedSourceError,
+    RecordParseError,
+)
 from airbyte_cdk.sources.file_based.file_based_stream_reader import (
     AbstractFileBasedStreamReader,
     FileReadMode,
@@ -203,7 +207,7 @@ class CsvParser(FileTypeParser):
                 break
 
         if not type_inferrer_by_field:
-            raise AirbyteTracedException(
+            raise EmptyFileSchemaInferenceError(
                 message=f"Could not infer schema as there are no rows in {file.uri}. If having an empty CSV file is expected, ignore this. "
                 f"Else, please contact Airbyte.",
                 failure_type=FailureType.config_error,

airbyte_cdk/sources/file_based/file_types/excel_parser.py

@@ -9,8 +9,6 @@ from typing import Any, Dict, Iterable, Mapping, Optional, Tuple, Union
 
 import orjson
 import pandas as pd
-from numpy import datetime64, issubdtype
-from numpy import dtype as dtype_
 from pydantic.v1 import BaseModel
 
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import (
@@ -141,7 +139,7 @@ class ExcelParser(FileTypeParser):
     @staticmethod
     def dtype_to_json_type(
         current_type: Optional[str],
-        dtype: dtype_,  # type: ignore [type-arg]
+        dtype: Any,  # Type object from pandas DataFrame
     ) -> str:
         """
         Convert Pandas DataFrame types to Airbyte Types.
@@ -163,7 +161,7 @@ class ExcelParser(FileTypeParser):
             return "number"
         if dtype == "bool" and (not current_type or current_type == "boolean"):
             return "boolean"
-        if issubdtype(dtype, datetime64):
+        if pd.api.types.is_datetime64_any_dtype(dtype):
             return "date-time"
         return "string"
 

airbyte_cdk/sources/file_based/stream/default_file_based_stream.py

@@ -9,13 +9,26 @@ from collections import defaultdict
 from copy import deepcopy
 from functools import cache
 from os import path
-from typing import Any, Dict, Iterable, List, Mapping, MutableMapping, Optional, Set, Tuple, Union
+from typing import (
+    Any,
+    Dict,
+    Iterable,
+    List,
+    Mapping,
+    MutableMapping,
+    NoReturn,
+    Optional,
+    Set,
+    Tuple,
+    Union,
+)
 
 from airbyte_cdk.models import AirbyteLogMessage, AirbyteMessage, AirbyteStream, FailureType, Level
 from airbyte_cdk.models import Type as MessageType
 from airbyte_cdk.sources.file_based.config.file_based_stream_config import PrimaryKeyType
 from airbyte_cdk.sources.file_based.exceptions import (
     DuplicatedFilesError,
+    EmptyFileSchemaInferenceError,
     FileBasedSourceError,
     InvalidSchemaError,
     MissingSchemaError,
@@ -230,7 +243,7 @@ class DefaultFileBasedStream(AbstractFileBasedStream, IncrementalMixin):
         return self.ab_last_mod_col
 
     @cache
-    def get_json_schema(self) -> JsonSchema:
+    def get_json_schema(self) -> JsonSchema:  # type: ignore
         if self.use_file_transfer:
             return file_transfer_schema
         extra_fields = {
@@ -246,12 +259,12 @@ class DefaultFileBasedStream(AbstractFileBasedStream, IncrementalMixin):
                 exception=AirbyteTracedException(exception=config_exception),
                 failure_type=FailureType.config_error,
             )
+        except EmptyFileSchemaInferenceError as exc:
+            self._raise_schema_inference_error(exc)
         except AirbyteTracedException as ate:
             raise ate
         except Exception as exc:
-            raise SchemaInferenceError(
-                FileBasedSourceError.SCHEMA_INFERENCE_ERROR, stream=self.name
-            ) from exc
+            self._raise_schema_inference_error(exc)
         else:
             return {"type": "object", "properties": {**extra_fields, **schema["properties"]}}
 
@@ -380,17 +393,24 @@ class DefaultFileBasedStream(AbstractFileBasedStream, IncrementalMixin):
 
         return base_schema
 
-    async def _infer_file_schema(self, file: RemoteFile) -> SchemaType:
+    async def _infer_file_schema(self, file: RemoteFile) -> SchemaType:  # type: ignore
         try:
             return await self.get_parser().infer_schema(
                 self.config, file, self.stream_reader, self.logger
             )
+        except EmptyFileSchemaInferenceError as exc:
+            self._raise_schema_inference_error(exc, file)
         except AirbyteTracedException as ate:
             raise ate
         except Exception as exc:
-            raise SchemaInferenceError(
-                FileBasedSourceError.SCHEMA_INFERENCE_ERROR,
-                file=file.uri,
-                format=str(self.config.format),
-                stream=self.name,
-            ) from exc
+            self._raise_schema_inference_error(exc, file)
+
+    def _raise_schema_inference_error(
+        self, exc: Exception, file: Optional[RemoteFile] = None
+    ) -> NoReturn:
+        raise SchemaInferenceError(
+            FileBasedSourceError.SCHEMA_INFERENCE_ERROR,
+            file=file.uri if file else None,
+            format=str(self.config.format) if self.config.format else None,
+            stream=self.name,
+        ) from exc

{airbyte_cdk-7.3.7.post3.dev18576192478.dist-info → airbyte_cdk-7.3.9.dist-info}/METADATA

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 7.3.7.post3.dev18576192478
+Version: 7.3.9
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT
@@ -35,7 +35,7 @@ Requires-Dist: click (>=8.1.8,<9.0.0)
 Requires-Dist: cohere (>=4.21,<6.0.0) ; extra == "vector-db-based"
 Requires-Dist: cryptography (>=44.0.0,<45.0.0)
 Requires-Dist: dateparser (>=1.2.2,<2.0.0)
-Requires-Dist: ddtrace (>=3.12.0,<4.0.0) ; extra == "manifest-server"
+Requires-Dist: ddtrace (>=3,<4) ; extra == "manifest-server"
 Requires-Dist: dpath (>=2.1.6,<3.0.0)
 Requires-Dist: dunamai (>=1.22.0,<2.0.0)
 Requires-Dist: fastapi (>=0.116.1) ; extra == "manifest-server"
@@ -45,11 +45,10 @@ Requires-Dist: google-cloud-secret-manager (>=2.17.0,<3.0.0)
 Requires-Dist: isodate (>=0.6.1,<0.7.0)
 Requires-Dist: jsonref (>=1,<2)
 Requires-Dist: jsonschema (>=4.17.3,<5.0)
-Requires-Dist: langchain (>=0.1.16,<1.0.0) ; extra == "vector-db-based"
-Requires-Dist: langchain_core (>=0.1.42,<1.0.0)
+Requires-Dist: langchain (==0.1.16) ; extra == "vector-db-based"
+Requires-Dist: langchain_core (==0.1.42)
 Requires-Dist: markdown ; extra == "file-based"
 Requires-Dist: nltk (==3.9.1)
-Requires-Dist: numpy (>=2.0.0,<3.0.0)
 Requires-Dist: openai[embeddings] (==0.27.9) ; extra == "vector-db-based"
 Requires-Dist: orjson (>=3.10.7,<4.0.0)
 Requires-Dist: packaging

{airbyte_cdk-7.3.7.post3.dev18576192478.dist-info → airbyte_cdk-7.3.9.dist-info}/RECORD

@@ -147,7 +147,7 @@ airbyte_cdk/sources/declarative/extractors/http_selector.py,sha256=2zWZ4ewTqQC8V
 airbyte_cdk/sources/declarative/extractors/record_extractor.py,sha256=XJELMjahAsaomlvQgN2zrNO0DJX0G0fr9r682gUz7Pg,691
 airbyte_cdk/sources/declarative/extractors/record_filter.py,sha256=sNLGjFX0fnqO_p1F5JO7-tBwjX83wZdz6W-WJTQpGps,3188
 airbyte_cdk/sources/declarative/extractors/record_selector.py,sha256=vCpwX1PVRFPYKMzm0DHHP3YEZ0Gmd3bBzggOsRha038,7192
-airbyte_cdk/sources/declarative/extractors/response_to_file_extractor.py,sha256=WJyA2OYIEgFpVP5Y3o0tIj69AV6IKkn9B16MeXaEItI,6513
+airbyte_cdk/sources/declarative/extractors/response_to_file_extractor.py,sha256=JI_O74KkwOn1Ex0dIr1aJ22eLvq_lUF4VeoOtgEfn1U,6512
 airbyte_cdk/sources/declarative/extractors/type_transformer.py,sha256=d6Y2Rfg8pMVEEnHllfVksWZdNVOU55yk34O03dP9muY,1626
 airbyte_cdk/sources/declarative/incremental/__init__.py,sha256=_y8H65KgdmVNpwQAzXtXzi-t9mY6bmIIAWtRAbpHfEo,295
 airbyte_cdk/sources/declarative/incremental/concurrent_partition_cursor.py,sha256=ldroIGnz1rie9cNZF-Jsl3J6yAqNe7KYS7PNE342Eqs,27995
@@ -296,15 +296,15 @@ airbyte_cdk/sources/file_based/config/validate_config_transfer_modes.py,sha256=O
 airbyte_cdk/sources/file_based/discovery_policy/__init__.py,sha256=gl3ey6mZbyfraB9P3pFhf9UJp2JeTZ1SUFAopy2iBvY,301
 airbyte_cdk/sources/file_based/discovery_policy/abstract_discovery_policy.py,sha256=dCfXX529Rd5rtopg4VeEgTPJjFtqjtjzPq6LCw18Wt0,605
 airbyte_cdk/sources/file_based/discovery_policy/default_discovery_policy.py,sha256=-xujTidtrq6HC00WKbjQh1CZdT5LMuzkp5BLjqDmfTY,1007
-airbyte_cdk/sources/file_based/exceptions.py,sha256=WP0qkG6fpWoBpOyyicgp5YNE393VWyegq5qSy0v4QtM,7362
+airbyte_cdk/sources/file_based/exceptions.py,sha256=pD_5avYUTVgDTDKedSHkivlohxlM4bfIAVmsl16LCAo,7434
 airbyte_cdk/sources/file_based/file_based_source.py,sha256=Xg8OYWnGc-OcVBglvS08uwAWGWHBhEqsBnyODIkOK-4,20051
 airbyte_cdk/sources/file_based/file_based_stream_permissions_reader.py,sha256=4e7FXqQ9hueacexC0SyrZyjF8oREYHza8pKF9CgKbD8,5050
 airbyte_cdk/sources/file_based/file_based_stream_reader.py,sha256=Yg9KRXpyAtElBrUOO8oX4WHQH6k6Lk7keklrZmB5Klg,9614
 airbyte_cdk/sources/file_based/file_record_data.py,sha256=Vkr5AyZzlsOezjVCLhFrm_WpymlQdolWCnFAwqLJ9Iw,453
 airbyte_cdk/sources/file_based/file_types/__init__.py,sha256=blCLn0-2LC-ZdgcNyDEhqM2RiUvEjEBh-G4-t32ZtuM,1268
 airbyte_cdk/sources/file_based/file_types/avro_parser.py,sha256=USEYqiICXBWpDV443VtNOCmUA-GINzY_Zah74_5w3qQ,10860
-airbyte_cdk/sources/file_based/file_types/csv_parser.py,sha256=QlCXB-ry3np67Q_VerQEPoWDOTcPTB6Go4ydZxY9ae4,20445
-airbyte_cdk/sources/file_based/file_types/excel_parser.py,sha256=BeplCq0hmojELU6bZCvvpRLpQ9us81TqbGYwrhd3INo,7188
+airbyte_cdk/sources/file_based/file_types/csv_parser.py,sha256=7DmwH1sHx08j7s58zfRydpU9OGl8V2wOEK_IUKC0Wy8,20500
+airbyte_cdk/sources/file_based/file_types/excel_parser.py,sha256=eN47x7TnRSMXXy2NOBHffOnNZLwkyQuHNUR0K_sXw0Y,7134
 airbyte_cdk/sources/file_based/file_types/file_transfer.py,sha256=rFxWaqItBux9tPf4xU03LT6b-wDZf1QolM92mP8Diuk,1120
 airbyte_cdk/sources/file_based/file_types/file_type_parser.py,sha256=JgpH21PrbRqwK92BJklZWvh2TndA6xZ-eP1LPMo44oQ,2832
 airbyte_cdk/sources/file_based/file_types/jsonl_parser.py,sha256=GwyNyxmST4RX-XpXy7xVH0D-znYWWBmGv_pVAu95oHQ,5886
@@ -326,7 +326,7 @@ airbyte_cdk/sources/file_based/stream/concurrent/cursor/file_based_final_state_c
 airbyte_cdk/sources/file_based/stream/cursor/__init__.py,sha256=MhFB5hOo8sjwvCh8gangaymdg3EJWYt_72brFOZt068,191
 airbyte_cdk/sources/file_based/stream/cursor/abstract_file_based_cursor.py,sha256=om-x3gZFPgWDpi15S9RxZmR36VHnk8sytgN6LlBQhAw,1934
 airbyte_cdk/sources/file_based/stream/cursor/default_file_based_cursor.py,sha256=VGV7xLyBribuBMVrXtO1xqkWJD86bl7yhXtjnwLMohM,7051
-airbyte_cdk/sources/file_based/stream/default_file_based_stream.py,sha256=gzSN4cLywwzo_U1OdFS3Of_-4DRkUcX_j7Mv30MrxQs,17154
+airbyte_cdk/sources/file_based/stream/default_file_based_stream.py,sha256=LaVegOgqMnsCihtWJxRpo6wB0w3bwS27UONnOhtm2QA,17614
 airbyte_cdk/sources/file_based/stream/identities_stream.py,sha256=FZH83Geoy3K3nwUk2VVNJERFcXUTnl-4XljjucUM23s,1893
 airbyte_cdk/sources/file_based/stream/permissions_file_based_stream.py,sha256=6KxqdD3-VvwxDTk7TtZ0M32fga4CI3qZ9IKdAkySpx0,3844
 airbyte_cdk/sources/file_based/types.py,sha256=INxG7OPnkdUP69oYNKMAbwhvV1AGvLRHs1J6pIia2FI,218
@@ -459,9 +459,9 @@ airbyte_cdk/utils/slice_hasher.py,sha256=EDxgROHDbfG-QKQb59m7h_7crN1tRiawdf5uU7G
 airbyte_cdk/utils/spec_schema_transformations.py,sha256=9YDJmnIGFsT51CVQf2tSSvTapGimITjEFGbUTSZAGTI,963
 airbyte_cdk/utils/stream_status_utils.py,sha256=ZmBoiy5HVbUEHAMrUONxZvxnvfV9CesmQJLDTAIWnWw,1171
 airbyte_cdk/utils/traced_exception.py,sha256=C8uIBuCL_E4WnBAOPSxBicD06JAldoN9fGsQDp463OY,6292
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/LICENSE_SHORT,sha256=aqF6D1NcESmpn-cqsxBtszTEnHKnlsp8L4x9wAh3Nxg,55
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/METADATA,sha256=h7yvpzPpkvihOLwh4XpAkBGqI3M6ISWeiitMiQshjnQ,6841
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/entry_points.txt,sha256=eLZ2UYvJZGm1s07Pplcs--1Gim60YhZWTb53j_dghwU,195
-airbyte_cdk-7.3.7.post3.dev18576192478.dist-info/RECORD,,
+airbyte_cdk-7.3.9.dist-info/LICENSE.txt,sha256=Wfe61S4BaGPj404v8lrAbvhjYR68SHlkzeYrg3_bbuM,1051
+airbyte_cdk-7.3.9.dist-info/LICENSE_SHORT,sha256=aqF6D1NcESmpn-cqsxBtszTEnHKnlsp8L4x9wAh3Nxg,55
+airbyte_cdk-7.3.9.dist-info/METADATA,sha256=R63LO-GP_SINpssfiNjKaM6cYClGSxQv3juKRp0ZZM0,6759
+airbyte_cdk-7.3.9.dist-info/WHEEL,sha256=Nq82e9rUAnEjt98J6MlVmMCZb-t9cYE2Ir1kpBmnWfs,88
+airbyte_cdk-7.3.9.dist-info/entry_points.txt,sha256=eLZ2UYvJZGm1s07Pplcs--1Gim60YhZWTb53j_dghwU,195
+airbyte_cdk-7.3.9.dist-info/RECORD,,