airbyte-cdk 6.7.0rc2__py3-none-any.whl → 6.7.1__py3-none-any.whl

airbyte_cdk/cli/__init__.py

@@ -0,0 +1 @@
+# Copyright (c) 2024 Airbyte, Inc., all rights reserved.

airbyte_cdk/cli/source_declarative_manifest/__init__.py

@@ -0,0 +1,6 @@
+from airbyte_cdk.cli.source_declarative_manifest._run import run
+
+
+__all__ = [
+    "run",
+]

airbyte_cdk/cli/source_declarative_manifest/_run.py

@@ -0,0 +1,224 @@
+# Copyright (c) 2024 Airbyte, Inc., all rights reserved.
+"""Defines the `source-declarative-manifest` connector, which installs alongside CDK.
+
+This file was originally imported from the dedicated connector directory, under the
+`airbyte` monorepo.
+
+Usage:
+
+```
+pipx install airbyte-cdk
+source-declarative-manifest --help
+source-declarative-manifest spec
+...
+```
+"""
+
+from __future__ import annotations
+
+import json
+import pkgutil
+import sys
+import traceback
+from collections.abc import Mapping
+from datetime import datetime
+from pathlib import Path
+from typing import Any, cast
+
+from orjson import orjson
+
+from airbyte_cdk.entrypoint import AirbyteEntrypoint, launch
+from airbyte_cdk.models import (
+    AirbyteErrorTraceMessage,
+    AirbyteMessage,
+    AirbyteMessageSerializer,
+    AirbyteStateMessage,
+    AirbyteTraceMessage,
+    ConfiguredAirbyteCatalog,
+    ConnectorSpecificationSerializer,
+    TraceType,
+    Type,
+)
+from airbyte_cdk.sources.declarative.concurrent_declarative_source import (
+    ConcurrentDeclarativeSource,
+)
+from airbyte_cdk.sources.declarative.yaml_declarative_source import YamlDeclarativeSource
+from airbyte_cdk.sources.source import TState
+
+
+class SourceLocalYaml(YamlDeclarativeSource):
+    """
+    Declarative source defined by a yaml file in the local filesystem
+    """
+
+    def __init__(
+        self,
+        catalog: ConfiguredAirbyteCatalog | None,
+        config: Mapping[str, Any] | None,
+        state: TState,
+        **kwargs: Any,
+    ) -> None:
+        """
+        HACK!
+            Problem: YamlDeclarativeSource relies on the calling module name/path to find the yaml file.
+            Implication: If you call YamlDeclarativeSource directly it will look for the yaml file in the wrong place. (e.g. the airbyte-cdk package)
+            Solution: Subclass YamlDeclarativeSource from the same location as the manifest to load.
+
+            When can we remove this?
+                When the airbyte-cdk is updated to not rely on the calling module name/path to find the yaml file.
+                When all manifest connectors are updated to use the new airbyte-cdk.
+                When all manifest connectors are updated to use the source-declarative-manifest as the base image.
+        """
+        super().__init__(
+            catalog=catalog,
+            config=config,
+            state=state,
+            path_to_yaml="manifest.yaml",
+        )
+
+
+def _is_local_manifest_command(args: list[str]) -> bool:
+    # Check for a local manifest.yaml file
+    return Path("/airbyte/integration_code/source_declarative_manifest/manifest.yaml").exists()
+
+
+def handle_command(args: list[str]) -> None:
+    if _is_local_manifest_command(args):
+        handle_local_manifest_command(args)
+    else:
+        handle_remote_manifest_command(args)
+
+
+def _get_local_yaml_source(args: list[str]) -> SourceLocalYaml:
+    try:
+        config, catalog, state = _parse_inputs_into_config_catalog_state(args)
+        return SourceLocalYaml(config=config, catalog=catalog, state=state)
+    except Exception as error:
+        print(
+            orjson.dumps(
+                AirbyteMessageSerializer.dump(
+                    AirbyteMessage(
+                        type=Type.TRACE,
+                        trace=AirbyteTraceMessage(
+                            type=TraceType.ERROR,
+                            emitted_at=int(datetime.now().timestamp() * 1000),
+                            error=AirbyteErrorTraceMessage(
+                                message=f"Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance. Error: {error}",
+                                stack_trace=traceback.format_exc(),
+                            ),
+                        ),
+                    )
+                )
+            ).decode()
+        )
+        raise error
+
+
+def handle_local_manifest_command(args: list[str]) -> None:
+    source = _get_local_yaml_source(args)
+    launch(
+        source=source,
+        args=args,
+    )
+
+
+def handle_remote_manifest_command(args: list[str]) -> None:
+    """Overrides the spec command to return the generalized spec for the declarative manifest source.
+
+    This is different from a typical low-code, but built and published separately source built as a ManifestDeclarativeSource,
+    because that will have a spec method that returns the spec for that specific source. Other than spec,
+    the generalized connector behaves the same as any other, since the manifest is provided in the config.
+    """
+    if args[0] == "spec":
+        json_spec = pkgutil.get_data(
+            "airbyte_cdk.cli.source_declarative_manifest",
+            "spec.json",
+        )
+        if json_spec is None:
+            raise FileNotFoundError(
+                "Could not find `spec.json` file for source-declarative-manifest"
+            )
+
+        spec_obj = json.loads(json_spec)
+        spec = ConnectorSpecificationSerializer.load(spec_obj)
+
+        message = AirbyteMessage(type=Type.SPEC, spec=spec)
+        print(AirbyteEntrypoint.airbyte_message_to_string(message))
+    else:
+        source = create_declarative_source(args)
+        launch(
+            source=source,
+            args=args,
+        )
+
+
+def create_declarative_source(args: list[str]) -> ConcurrentDeclarativeSource:
+    """Creates the source with the injected config.
+
+    This essentially does what other low-code sources do at build time, but at runtime,
+    with a user-provided manifest in the config. This better reflects what happens in the
+    connector builder.
+    """
+    try:
+        config, catalog, state = _parse_inputs_into_config_catalog_state(args)
+        if "__injected_declarative_manifest" not in config:
+            raise ValueError(
+                f"Invalid config: `__injected_declarative_manifest` should be provided at the root of the config but config only has keys {list(config.keys())}"
+            )
+        return ConcurrentDeclarativeSource(
+            config=config,
+            catalog=catalog,
+            state=state,
+            source_config=cast(dict[str, Any], config["__injected_declarative_manifest"]),
+        )
+    except Exception as error:
+        print(
+            orjson.dumps(
+                AirbyteMessageSerializer.dump(
+                    AirbyteMessage(
+                        type=Type.TRACE,
+                        trace=AirbyteTraceMessage(
+                            type=TraceType.ERROR,
+                            emitted_at=int(datetime.now().timestamp() * 1000),
+                            error=AirbyteErrorTraceMessage(
+                                message=f"Error starting the sync. This could be due to an invalid configuration or catalog. Please contact Support for assistance. Error: {error}",
+                                stack_trace=traceback.format_exc(),
+                            ),
+                        ),
+                    )
+                )
+            ).decode()
+        )
+        raise error
+
+
+def _parse_inputs_into_config_catalog_state(
+    args: list[str],
+) -> tuple[
+    Mapping[str, Any] | None,
+    ConfiguredAirbyteCatalog | None,
+    list[AirbyteStateMessage],
+]:
+    parsed_args = AirbyteEntrypoint.parse_args(args)
+    config = (
+        ConcurrentDeclarativeSource.read_config(parsed_args.config)
+        if hasattr(parsed_args, "config")
+        else None
+    )
+    catalog = (
+        ConcurrentDeclarativeSource.read_catalog(parsed_args.catalog)
+        if hasattr(parsed_args, "catalog")
+        else None
+    )
+    state = (
+        ConcurrentDeclarativeSource.read_state(parsed_args.state)
+        if hasattr(parsed_args, "state")
+        else []
+    )
+
+    return config, catalog, state
+
+
+def run() -> None:
+    args: list[str] = sys.argv[1:]
+    handle_command(args)

airbyte_cdk/cli/source_declarative_manifest/spec.json

@@ -0,0 +1,17 @@
+{
+  "documentationUrl": "https://docs.airbyte.com/integrations/sources/low-code",
+  "connectionSpecification": {
+    "$schema": "http://json-schema.org/draft-07/schema#",
+    "title": "Low-code source spec",
+    "type": "object",
+    "required": ["__injected_declarative_manifest"],
+    "additionalProperties": true,
+    "properties": {
+      "__injected_declarative_manifest": {
+        "title": "Low-code manifest",
+        "type": "object",
+        "description": "The low-code manifest that defines the components of the source."
+      }
+    }
+  }
+}

airbyte_cdk/config_observation.py

@@ -10,6 +10,8 @@ import time
 from copy import copy
 from typing import Any, List, MutableMapping
 
+from orjson import orjson
+
 from airbyte_cdk.models import (
     AirbyteControlConnectorConfigMessage,
     AirbyteControlMessage,
@@ -18,7 +20,6 @@ from airbyte_cdk.models import (
     OrchestratorType,
     Type,
 )
-from orjson import orjson
 
 
 class ObservedDict(dict):  # type: ignore # disallow_any_generics is set to True, and dict is equivalent to dict[Any]

airbyte_cdk/connector.py

@@ -11,6 +11,7 @@ from abc import ABC, abstractmethod
 from typing import Any, Generic, Mapping, Optional, Protocol, TypeVar
 
 import yaml
+
 from airbyte_cdk.models import (
     AirbyteConnectionStatus,
     ConnectorSpecification,

airbyte_cdk/connector_builder/connector_builder_handler.py

@@ -12,8 +12,8 @@ from airbyte_cdk.models import (
     AirbyteRecordMessage,
     AirbyteStateMessage,
     ConfiguredAirbyteCatalog,
+    Type,
 )
-from airbyte_cdk.models import Type
 from airbyte_cdk.models import Type as MessageType
 from airbyte_cdk.sources.declarative.declarative_source import DeclarativeSource
 from airbyte_cdk.sources.declarative.manifest_declarative_source import ManifestDeclarativeSource

airbyte_cdk/connector_builder/main.py

@@ -6,6 +6,8 @@
 import sys
 from typing import Any, List, Mapping, Optional, Tuple
 
+from orjson import orjson
+
 from airbyte_cdk.connector import BaseConnector
 from airbyte_cdk.connector_builder.connector_builder_handler import (
     TestReadLimits,
@@ -25,7 +27,6 @@ from airbyte_cdk.models import (
 from airbyte_cdk.sources.declarative.manifest_declarative_source import ManifestDeclarativeSource
 from airbyte_cdk.sources.source import Source
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from orjson import orjson
 
 
 def get_config_and_catalog_from_args(

airbyte_cdk/destinations/destination.py

@@ -9,6 +9,8 @@ import sys
 from abc import ABC, abstractmethod
 from typing import Any, Iterable, List, Mapping
 
+from orjson import orjson
+
 from airbyte_cdk.connector import Connector
 from airbyte_cdk.exception_handler import init_uncaught_exception_handler
 from airbyte_cdk.models import (
@@ -20,7 +22,6 @@ from airbyte_cdk.models import (
 )
 from airbyte_cdk.sources.utils.schema_helpers import check_config_against_spec_or_exit
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from orjson import orjson
 
 logger = logging.getLogger("airbyte")
 

airbyte_cdk/destinations/vector_db_based/config.py

@@ -5,9 +5,10 @@
 from typing import Any, Dict, List, Literal, Optional, Union
 
 import dpath
+from pydantic.v1 import BaseModel, Field
+
 from airbyte_cdk.utils.oneof_option_config import OneOfOptionConfig
 from airbyte_cdk.utils.spec_schema_transformations import resolve_refs
-from pydantic.v1 import BaseModel, Field
 
 
 class SeparatorSplitterConfigModel(BaseModel):

airbyte_cdk/destinations/vector_db_based/document_processor.py

@@ -8,6 +8,10 @@ from dataclasses import dataclass
 from typing import Any, Dict, List, Mapping, Optional, Tuple
 
 import dpath
+from langchain.text_splitter import Language, RecursiveCharacterTextSplitter
+from langchain.utils import stringify_dict
+from langchain_core.documents.base import Document
+
 from airbyte_cdk.destinations.vector_db_based.config import (
     ProcessingConfigModel,
     SeparatorSplitterConfigModel,
@@ -21,9 +25,6 @@ from airbyte_cdk.models import (
     DestinationSyncMode,
 )
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException, FailureType
-from langchain.text_splitter import Language, RecursiveCharacterTextSplitter
-from langchain.utils import stringify_dict
-from langchain_core.documents.base import Document
 
 METADATA_STREAM_FIELD = "_ab_stream"
 METADATA_RECORD_ID_FIELD = "_ab_record_id"

airbyte_cdk/destinations/vector_db_based/embedder.py

@@ -7,6 +7,11 @@ from abc import ABC, abstractmethod
 from dataclasses import dataclass
 from typing import List, Optional, Union, cast
 
+from langchain.embeddings.cohere import CohereEmbeddings
+from langchain.embeddings.fake import FakeEmbeddings
+from langchain.embeddings.localai import LocalAIEmbeddings
+from langchain.embeddings.openai import OpenAIEmbeddings
+
 from airbyte_cdk.destinations.vector_db_based.config import (
     AzureOpenAIEmbeddingConfigModel,
     CohereEmbeddingConfigModel,
@@ -19,10 +24,6 @@ from airbyte_cdk.destinations.vector_db_based.config import (
 from airbyte_cdk.destinations.vector_db_based.utils import create_chunks, format_exception
 from airbyte_cdk.models import AirbyteRecordMessage
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException, FailureType
-from langchain.embeddings.cohere import CohereEmbeddings
-from langchain.embeddings.fake import FakeEmbeddings
-from langchain.embeddings.localai import LocalAIEmbeddings
-from langchain.embeddings.openai import OpenAIEmbeddings
 
 
 @dataclass

airbyte_cdk/entrypoint.py

@@ -16,6 +16,9 @@ from typing import Any, DefaultDict, Iterable, List, Mapping, Optional
 from urllib.parse import urlparse
 
 import requests
+from orjson import orjson
+from requests import PreparedRequest, Response, Session
+
 from airbyte_cdk.connector import TConfig
 from airbyte_cdk.exception_handler import init_uncaught_exception_handler
 from airbyte_cdk.logger import init_logger
@@ -38,8 +41,6 @@ from airbyte_cdk.utils import is_cloud_environment, message_utils
 from airbyte_cdk.utils.airbyte_secrets_utils import get_secrets, update_secrets
 from airbyte_cdk.utils.constants import ENV_REQUEST_CACHE_PATH
 from airbyte_cdk.utils.traced_exception import AirbyteTracedException
-from orjson import orjson
-from requests import PreparedRequest, Response, Session
 
 logger = init_logger("airbyte")
 

airbyte_cdk/logger.py

@@ -7,6 +7,8 @@ import logging
 import logging.config
 from typing import Any, Callable, Mapping, Optional, Tuple
 
+from orjson import orjson
+
 from airbyte_cdk.models import (
     AirbyteLogMessage,
     AirbyteMessage,
@@ -15,7 +17,6 @@ from airbyte_cdk.models import (
     Type,
 )
 from airbyte_cdk.utils.airbyte_secrets_utils import filter_secrets
-from orjson import orjson
 
 LOGGING_CONFIG = {
     "version": 1,

airbyte_cdk/models/__init__.py

@@ -39,7 +39,9 @@ from .airbyte_protocol import (
     FailureType,
     Level,
     OAuthConfigSpecification,
+    OauthConnectorInputSpecification,
     OrchestratorType,
+    State,
     Status,
     StreamDescriptor,
     SyncMode,

airbyte_cdk/models/airbyte_protocol.py

@@ -5,10 +5,11 @@
 from dataclasses import InitVar, dataclass
 from typing import Annotated, Any, Dict, List, Mapping, Optional, Union
 
-from airbyte_cdk.models.file_transfer_record_message import AirbyteFileTransferRecordMessage
 from airbyte_protocol_dataclasses.models import *  # noqa: F403  # Allow '*'
 from serpyco_rs.metadata import Alias
 
+from airbyte_cdk.models.file_transfer_record_message import AirbyteFileTransferRecordMessage
+
 # ruff: noqa: F405  # ignore fuzzy import issues with 'import *'
 
 

airbyte_cdk/sources/concurrent_source/concurrent_read_processor.py

@@ -17,8 +17,8 @@ from airbyte_cdk.sources.streams.concurrent.abstract_stream import AbstractStrea
 from airbyte_cdk.sources.streams.concurrent.partition_enqueuer import PartitionEnqueuer
 from airbyte_cdk.sources.streams.concurrent.partition_reader import PartitionReader
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
 from airbyte_cdk.sources.streams.concurrent.partitions.types import PartitionCompleteSentinel
+from airbyte_cdk.sources.types import Record
 from airbyte_cdk.sources.utils.record_helper import stream_data_to_airbyte_message
 from airbyte_cdk.sources.utils.slice_logger import SliceLogger
 from airbyte_cdk.utils import AirbyteTracedException
@@ -147,11 +147,11 @@ class ConcurrentReadProcessor:
         # AbstractStreams are expected to return data as they are expected.
         # Any transformation on the data should be done before reaching this point
         message = stream_data_to_airbyte_message(
-            stream_name=record.partition.stream_name(),
+            stream_name=record.stream_name,
             data_or_message=record.data,
             is_file_transfer_message=record.is_file_transfer_message,
         )
-        stream = self._stream_name_to_instance[record.partition.stream_name()]
+        stream = self._stream_name_to_instance[record.stream_name]
 
         if message.type == MessageType.RECORD:
             if self._record_counter[stream.name] == 0:

airbyte_cdk/sources/concurrent_source/concurrent_source.py

@@ -18,11 +18,11 @@ from airbyte_cdk.sources.streams.concurrent.abstract_stream import AbstractStrea
 from airbyte_cdk.sources.streams.concurrent.partition_enqueuer import PartitionEnqueuer
 from airbyte_cdk.sources.streams.concurrent.partition_reader import PartitionReader
 from airbyte_cdk.sources.streams.concurrent.partitions.partition import Partition
-from airbyte_cdk.sources.streams.concurrent.partitions.record import Record
 from airbyte_cdk.sources.streams.concurrent.partitions.types import (
     PartitionCompleteSentinel,
     QueueItem,
 )
+from airbyte_cdk.sources.types import Record
 from airbyte_cdk.sources.utils.slice_logger import DebugSliceLogger, SliceLogger
 
 

airbyte_cdk/sources/config.py

@@ -4,9 +4,10 @@
 
 from typing import Any, Dict
 
-from airbyte_cdk.sources.utils.schema_helpers import expand_refs, rename_key
 from pydantic.v1 import BaseModel
 
+from airbyte_cdk.sources.utils.schema_helpers import expand_refs, rename_key
+
 
 class BaseConfig(BaseModel):
     """Base class for connector spec, adds the following behaviour:

airbyte_cdk/sources/declarative/auth/jwt.py

@@ -8,6 +8,7 @@ from datetime import datetime
 from typing import Any, Mapping, Optional, Union
 
 import jwt
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
 from airbyte_cdk.sources.declarative.interpolation.interpolated_boolean import InterpolatedBoolean
 from airbyte_cdk.sources.declarative.interpolation.interpolated_mapping import InterpolatedMapping

airbyte_cdk/sources/declarative/auth/oauth.py

@@ -6,6 +6,7 @@ from dataclasses import InitVar, dataclass, field
 from typing import Any, List, Mapping, Optional, Union
 
 import pendulum
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
 from airbyte_cdk.sources.declarative.interpolation.interpolated_mapping import InterpolatedMapping
 from airbyte_cdk.sources.declarative.interpolation.interpolated_string import InterpolatedString

airbyte_cdk/sources/declarative/auth/selective_authenticator.py

@@ -6,6 +6,7 @@ from dataclasses import dataclass
 from typing import Any, List, Mapping
 
 import dpath
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
 
 

airbyte_cdk/sources/declarative/auth/token.py

@@ -8,6 +8,8 @@ from dataclasses import InitVar, dataclass
 from typing import Any, Mapping, Union
 
 import requests
+from cachetools import TTLCache, cached
+
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
 from airbyte_cdk.sources.declarative.auth.token_provider import TokenProvider
 from airbyte_cdk.sources.declarative.interpolation.interpolated_string import InterpolatedString
@@ -16,7 +18,6 @@ from airbyte_cdk.sources.declarative.requesters.request_option import (
     RequestOptionType,
 )
 from airbyte_cdk.sources.types import Config
-from cachetools import TTLCache, cached
 
 
 @dataclass

airbyte_cdk/sources/declarative/auth/token_provider.py

@@ -10,6 +10,9 @@ from typing import Any, List, Mapping, Optional, Union
 
 import dpath
 import pendulum
+from isodate import Duration
+from pendulum import DateTime
+
 from airbyte_cdk.sources.declarative.decoders.decoder import Decoder
 from airbyte_cdk.sources.declarative.decoders.json_decoder import JsonDecoder
 from airbyte_cdk.sources.declarative.exceptions import ReadException
@@ -18,8 +21,6 @@ from airbyte_cdk.sources.declarative.requesters.requester import Requester
 from airbyte_cdk.sources.http_logger import format_http_message
 from airbyte_cdk.sources.message import MessageRepository, NoopMessageRepository
 from airbyte_cdk.sources.types import Config
-from isodate import Duration
-from pendulum import DateTime
 
 
 class TokenProvider: