airbyte-cdk 6.21.1.dev0__tar.gz → 6.22.0__tar.gz

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.3
 Name: airbyte-cdk
-Version: 6.21.1.dev0
+Version: 6.22.0
 Summary: A framework for writing Airbyte Connectors.
 License: MIT
 Keywords: airbyte,connector-development-kit,cdk
@@ -66,7 +66,7 @@ Requires-Dist: tiktoken (==0.8.0) ; extra == "vector-db-based"
 Requires-Dist: unstructured.pytesseract (>=0.3.12) ; extra == "file-based"
 Requires-Dist: unstructured[docx,pptx] (==0.10.27) ; extra == "file-based"
 Requires-Dist: wcmatch (==10.0)
-Requires-Dist: xmltodict (>=0.13.0,<0.14.0)
+Requires-Dist: xmltodict (>=0.13,<0.15)
 Project-URL: Documentation, https://docs.airbyte.io/
 Project-URL: Homepage, https://airbyte.com
 Project-URL: Repository, https://github.com/airbytehq/airbyte-python-cdk

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/airbyte_cdk/sources/declarative/declarative_component_schema.yaml

@@ -1514,6 +1514,7 @@ definitions:
         anyOf:
           - "$ref": "#/definitions/JsonDecoder"
           - "$ref": "#/definitions/XmlDecoder"
+          - "$ref": "#/definitions/CompositeRawDecoder"
       $parameters:
         type: object
         additionalProperties: true
@@ -2100,6 +2101,26 @@ definitions:
       $parameters:
         type: object
         additionalProperties: true
+  ZipfileDecoder:
+    title: Zipfile Decoder
+    description: Decoder for response data that is returned as zipfile(s).
+    type: object
+    additionalProperties: true
+    required:
+      - type
+      - parser
+    properties:
+      type:
+        type: string
+        enum: [ZipfileDecoder]
+      parser:
+        title: Parser
+        description: Parser to parse the decompressed data from the zipfile(s).
+        anyOf:
+          - "$ref": "#/definitions/GzipParser"
+          - "$ref": "#/definitions/JsonParser"
+          - "$ref": "#/definitions/JsonLineParser"
+          - "$ref": "#/definitions/CsvParser"
   ListPartitionRouter:
     title: List Partition Router
     description: A Partition router that specifies a list of attributes where each attribute describes a portion of the complete data set for a stream. During a sync, each value is iterated over and can be used as input to outbound API requests.
@@ -2928,6 +2949,7 @@ definitions:
           - "$ref": "#/definitions/XmlDecoder"
           - "$ref": "#/definitions/GzipJsonDecoder"
           - "$ref": "#/definitions/CompositeRawDecoder"
+          - "$ref": "#/definitions/ZipfileDecoder"
       $parameters:
         type: object
         additionalProperties: true
@@ -3126,6 +3148,8 @@ definitions:
           - "$ref": "#/definitions/IterableDecoder"
           - "$ref": "#/definitions/XmlDecoder"
           - "$ref": "#/definitions/GzipJsonDecoder"
+          - "$ref": "#/definitions/CompositeRawDecoder"
+          - "$ref": "#/definitions/ZipfileDecoder"
       download_decoder:
         title: Download Decoder
         description: Component decoding the download response so records can be extracted.
@@ -3136,6 +3160,8 @@ definitions:
           - "$ref": "#/definitions/IterableDecoder"
           - "$ref": "#/definitions/XmlDecoder"
           - "$ref": "#/definitions/GzipJsonDecoder"
+          - "$ref": "#/definitions/CompositeRawDecoder"
+          - "$ref": "#/definitions/ZipfileDecoder"
       $parameters:
         type: object
         additionalProperties: true

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/airbyte_cdk/sources/declarative/decoders/__init__.py

@@ -2,7 +2,12 @@
 # Copyright (c) 2023 Airbyte, Inc., all rights reserved.
 #
 
-from airbyte_cdk.sources.declarative.decoders.composite_raw_decoder import CompositeRawDecoder
+from airbyte_cdk.sources.declarative.decoders.composite_raw_decoder import (
+    CompositeRawDecoder,
+    GzipParser,
+    JsonParser,
+    Parser,
+)
 from airbyte_cdk.sources.declarative.decoders.decoder import Decoder
 from airbyte_cdk.sources.declarative.decoders.json_decoder import (
     GzipJsonDecoder,
@@ -15,15 +20,18 @@ from airbyte_cdk.sources.declarative.decoders.pagination_decoder_decorator impor
     PaginationDecoderDecorator,
 )
 from airbyte_cdk.sources.declarative.decoders.xml_decoder import XmlDecoder
+from airbyte_cdk.sources.declarative.decoders.zipfile_decoder import ZipfileDecoder
 
 __all__ = [
     "Decoder",
     "CompositeRawDecoder",
     "JsonDecoder",
+    "JsonParser",
     "JsonlDecoder",
     "IterableDecoder",
     "GzipJsonDecoder",
     "NoopDecoder",
     "PaginationDecoderDecorator",
     "XmlDecoder",
+    "ZipfileDecoder",
 ]

airbyte_cdk-6.22.0/airbyte_cdk/sources/declarative/decoders/zipfile_decoder.py

@@ -0,0 +1,59 @@
+#
+# Copyright (c) 2024 Airbyte, Inc., all rights reserved.
+#
+
+import logging
+import zipfile
+from dataclasses import dataclass
+from io import BytesIO
+from typing import Any, Generator, MutableMapping
+
+import orjson
+import requests
+
+from airbyte_cdk.models import FailureType
+from airbyte_cdk.sources.declarative.decoders import Decoder
+from airbyte_cdk.sources.declarative.decoders.composite_raw_decoder import (
+    Parser,
+)
+from airbyte_cdk.utils import AirbyteTracedException
+
+logger = logging.getLogger("airbyte")
+
+
+@dataclass
+class ZipfileDecoder(Decoder):
+    parser: Parser
+
+    def is_stream_response(self) -> bool:
+        return False
+
+    def decode(
+        self, response: requests.Response
+    ) -> Generator[MutableMapping[str, Any], None, None]:
+        try:
+            with zipfile.ZipFile(BytesIO(response.content)) as zip_file:
+                for file_name in zip_file.namelist():
+                    unzipped_content = zip_file.read(file_name)
+                    buffered_content = BytesIO(unzipped_content)
+                    try:
+                        yield from self.parser.parse(buffered_content)
+                    except Exception as e:
+                        logger.error(
+                            f"Failed to parse file: {file_name} from zip file: {response.request.url} with exception {e}."
+                        )
+                        raise AirbyteTracedException(
+                            message=f"Failed to parse file: {file_name} from zip file.",
+                            internal_message=f"Failed to parse file: {file_name} from zip file: {response.request.url}.",
+                            failure_type=FailureType.system_error,
+                        ) from e
+        except zipfile.BadZipFile as e:
+            logger.error(
+                f"Received an invalid zip file in response to URL: {response.request.url}. "
+                f"The size of the response body is: {len(response.content)}"
+            )
+            raise AirbyteTracedException(
+                message="Received an invalid zip file in response.",
+                internal_message=f"Received an invalid zip file in response to URL: {response.request.url}.",
+                failure_type=FailureType.system_error,
+            ) from e

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/airbyte_cdk/sources/declarative/models/declarative_component_schema.py

@@ -1242,9 +1242,6 @@ class LegacySessionTokenAuthenticator(BaseModel):
 
 
 class JsonParser(BaseModel):
-    class Config:
-        extra = Extra.allow
-
     type: Literal["JsonParser"]
     encoding: Optional[str] = "utf-8"
 
@@ -1680,6 +1677,18 @@ class CompositeErrorHandler(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
+class ZipfileDecoder(BaseModel):
+    class Config:
+        extra = Extra.allow
+
+    type: Literal["ZipfileDecoder"]
+    parser: Union[GzipParser, JsonParser, JsonLineParser, CsvParser] = Field(
+        ...,
+        description="Parser to parse the decompressed data from the zipfile(s).",
+        title="Parser",
+    )
+
+
 class CompositeRawDecoder(BaseModel):
     type: Literal["CompositeRawDecoder"]
     parser: Union[GzipParser, JsonParser, JsonLineParser, CsvParser]
@@ -1886,7 +1895,7 @@ class SessionTokenAuthenticator(BaseModel):
         description="Authentication method to use for requests sent to the API, specifying how to inject the session token.",
         title="Data Request Authentication",
     )
-    decoder: Optional[Union[JsonDecoder, XmlDecoder]] = Field(
+    decoder: Optional[Union[JsonDecoder, XmlDecoder, CompositeRawDecoder]] = Field(
         None, description="Component used to decode the response.", title="Decoder"
     )
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
@@ -2092,6 +2101,7 @@ class SimpleRetriever(BaseModel):
             XmlDecoder,
             GzipJsonDecoder,
             CompositeRawDecoder,
+            ZipfileDecoder,
         ]
     ] = Field(
         None,
@@ -2168,6 +2178,8 @@ class AsyncRetriever(BaseModel):
             IterableDecoder,
             XmlDecoder,
             GzipJsonDecoder,
+            CompositeRawDecoder,
+            ZipfileDecoder,
         ]
     ] = Field(
         None,
@@ -2182,6 +2194,8 @@ class AsyncRetriever(BaseModel):
             IterableDecoder,
             XmlDecoder,
             GzipJsonDecoder,
+            CompositeRawDecoder,
+            ZipfileDecoder,
         ]
     ] = Field(
         None,

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py

@@ -66,6 +66,7 @@ from airbyte_cdk.sources.declarative.decoders import (
     JsonlDecoder,
     PaginationDecoderDecorator,
     XmlDecoder,
+    ZipfileDecoder,
 )
 from airbyte_cdk.sources.declarative.decoders.composite_raw_decoder import (
     CompositeRawDecoder,
@@ -359,6 +360,9 @@ from airbyte_cdk.sources.declarative.models.declarative_component_schema import
 from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
     XmlDecoder as XmlDecoderModel,
 )
+from airbyte_cdk.sources.declarative.models.declarative_component_schema import (
+    ZipfileDecoder as ZipfileDecoderModel,
+)
 from airbyte_cdk.sources.declarative.partition_routers import (
     CartesianProductStreamSlicer,
     ListPartitionRouter,
@@ -578,6 +582,7 @@ class ModelToComponentFactory:
             ConfigComponentsResolverModel: self.create_config_components_resolver,
             StreamConfigModel: self.create_stream_config,
             ComponentMappingDefinitionModel: self.create_components_mapping_definition,
+            ZipfileDecoderModel: self.create_zipfile_decoder,
         }
 
         # Needed for the case where we need to perform a second parse on the fields of a custom component
@@ -1820,6 +1825,12 @@ class ModelToComponentFactory:
     ) -> GzipJsonDecoder:
         return GzipJsonDecoder(parameters={}, encoding=model.encoding)
 
+    def create_zipfile_decoder(
+        self, model: ZipfileDecoderModel, config: Config, **kwargs: Any
+    ) -> ZipfileDecoder:
+        parser = self._create_component_from_model(model=model.parser, config=config)
+        return ZipfileDecoder(parser=parser)
+
     def create_gzip_parser(
         self, model: GzipParserModel, config: Config, **kwargs: Any
     ) -> GzipParser:

{airbyte_cdk-6.21.1.dev0 → airbyte_cdk-6.22.0}/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 = "6.21.1dev0"  # Version will be calculated dynamically.
+version = "6.22.0"  # Version will be calculated dynamically.
 
 [tool.poetry-dynamic-versioning]
 enable = false
@@ -79,7 +79,7 @@ pytz = "2024.2"
 orjson = "^3.10.7"
 serpyco-rs = "^1.10.2"
 sqlalchemy = {version = "^2.0,!=2.0.36", optional = true }
-xmltodict = "^0.13.0"
+xmltodict = ">=0.13,<0.15"
 Unidecode = "^1.3"
 
 [tool.poetry.group.dev.dependencies]