airbyte-cdk 6.33.2.dev1__tar.gz → 6.33.4__tar.gz

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 6.33.2.dev1
+Version: 6.33.4
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/auth/oauth.py

@@ -3,7 +3,7 @@
 #
 
 from dataclasses import InitVar, dataclass, field
-from datetime import timedelta
+from datetime import datetime, timedelta
 from typing import Any, List, Mapping, MutableMapping, Optional, Union
 
 from airbyte_cdk.sources.declarative.auth.declarative_authenticator import DeclarativeAuthenticator
@@ -232,8 +232,13 @@ class DeclarativeOauth2Authenticator(AbstractOauth2Authenticator, DeclarativeAut
         return self._refresh_request_headers.eval(self.config)
 
     def get_token_expiry_date(self) -> AirbyteDateTime:
+        if not self._has_access_token_been_initialized():
+            return AirbyteDateTime.from_datetime(datetime.min)
         return self._token_expiry_date  # type: ignore # _token_expiry_date is an AirbyteDateTime. It is never None despite what mypy thinks
 
+    def _has_access_token_been_initialized(self) -> bool:
+        return self._access_token is not None
+
     def set_token_expiry_date(self, value: Union[str, int]) -> None:
         self._token_expiry_date = self._parse_token_expiration_date(value)
 

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/concurrent_declarative_source.py

@@ -3,7 +3,7 @@
 #
 
 import logging
-from typing import Any, Generic, Iterator, List, Mapping, Optional, Tuple
+from typing import Any, Generic, Iterator, List, Mapping, MutableMapping, Optional, Tuple
 
 from airbyte_cdk.models import (
     AirbyteCatalog,
@@ -224,6 +224,7 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                     stream_state = self._connector_state_manager.get_stream_state(
                         stream_name=declarative_stream.name, namespace=declarative_stream.namespace
                     )
+                    stream_state = self._migrate_state(declarative_stream, stream_state)
 
                     retriever = self._get_retriever(declarative_stream, stream_state)
 
@@ -331,6 +332,8 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                     stream_state = self._connector_state_manager.get_stream_state(
                         stream_name=declarative_stream.name, namespace=declarative_stream.namespace
                     )
+                    stream_state = self._migrate_state(declarative_stream, stream_state)
+
                     partition_router = declarative_stream.retriever.stream_slicer._partition_router
 
                     perpartition_cursor = (
@@ -521,3 +524,14 @@ class ConcurrentDeclarativeSource(ManifestDeclarativeSource, Generic[TState]):
                 if stream.stream.name not in concurrent_stream_names
             ]
         )
+
+    @staticmethod
+    def _migrate_state(
+        declarative_stream: DeclarativeStream, stream_state: MutableMapping[str, Any]
+    ) -> MutableMapping[str, Any]:
+        for state_migration in declarative_stream.state_migrations:
+            if state_migration.should_migrate(stream_state):
+                # The state variable is expected to be mutable but the migrate method returns an immutable mapping.
+                stream_state = dict(state_migration.migrate(stream_state))
+
+        return stream_state

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/declarative_component_schema.yaml

@@ -41,11 +41,7 @@ properties:
   concurrency_level:
     "$ref": "#/definitions/ConcurrencyLevel"
   api_budget:
-    title: API Budget
-    description: Defines how many requests can be made to the API in a given time frame. This field accepts either a generic APIBudget or an HTTP-specific configuration (HTTPAPIBudget) to be applied across all streams.
-    anyOf:
-      - "$ref": "#/definitions/APIBudget"
-      - "$ref": "#/definitions/HTTPAPIBudget"
+    "$ref": "#/definitions/HTTPAPIBudget"
   metadata:
     type: object
     description: For internal Airbyte use only - DO NOT modify manually. Used by consumers of declarative manifests for storing related metadata.
@@ -1371,40 +1367,13 @@ definitions:
       $parameters:
         type: object
         additional_properties: true
-  APIBudget:
-    title: API Budget
-    description: >
-      A generic API budget configuration that defines the policies (rate limiting rules)
-      and the maximum number of attempts to acquire a call credit. This budget does not automatically
-      update itself based on HTTP response headers.
-    type: object
-    required:
-      - type
-      - policies
-    properties:
-      type:
-        type: string
-        enum: [APIBudget]
-      policies:
-        title: Policies
-        description: List of call rate policies that define how many calls are allowed.
-        type: array
-        items:
-          anyOf:
-            - "$ref": "#/definitions/FixedWindowCallRatePolicy"
-            - "$ref": "#/definitions/MovingWindowCallRatePolicy"
-            - "$ref": "#/definitions/UnlimitedCallRatePolicy"
-      maximum_attempts_to_acquire:
-        title: Maximum Attempts to Acquire
-        description: The maximum number of attempts to acquire a call before giving up.
-        type: integer
-        default: 100000
-    additionalProperties: true
   HTTPAPIBudget:
     title: HTTP API Budget
     description: >
-      An HTTP-specific API budget that extends APIBudget by updating rate limiting information based
-      on HTTP response headers. It extracts available calls and the next reset timestamp from the HTTP responses.
+      Defines how many requests can be made to the API in a given time frame. `HTTPAPIBudget` extracts the remaining
+      call count and the reset time from HTTP response headers using the header names provided by
+      `ratelimit_remaining_header` and `ratelimit_reset_header`. Only requests using `HttpRequester`
+      are rate-limited; custom components that bypass `HttpRequester` are not covered by this budget.
     type: object
     required:
       - type
@@ -1439,11 +1408,6 @@ definitions:
         items:
           type: integer
         default: [429]
-      maximum_attempts_to_acquire:
-        title: Maximum Attempts to Acquire
-        description: The maximum number of attempts to acquire a call before giving up.
-        type: integer
-        default: 100000
     additionalProperties: true
   FixedWindowCallRatePolicy:
     title: Fixed Window Call Rate Policy
@@ -1451,7 +1415,6 @@ definitions:
     type: object
     required:
       - type
-      - next_reset_ts
       - period
       - call_limit
       - matchers
@@ -1459,16 +1422,10 @@ definitions:
       type:
         type: string
         enum: [FixedWindowCallRatePolicy]
-      next_reset_ts:
-        title: Next Reset Timestamp
-        description: The timestamp when the rate limit will reset.
-        type: string
-        format: date-time
       period:
         title: Period
         description: The time interval for the rate limit window.
         type: string
-        format: duration
       call_limit:
         title: Call Limit
         description: The maximum number of calls allowed within the period.
@@ -1539,7 +1496,9 @@ definitions:
         title: Interval
         description: The time interval for the rate limit.
         type: string
-        format: duration
+        examples:
+          - "PT1H"
+          - "P1D"
     additionalProperties: true
   HttpRequestRegexMatcher:
     title: HTTP Request Matcher
@@ -1756,7 +1715,6 @@ definitions:
         anyOf:
           - "$ref": "#/definitions/JsonDecoder"
           - "$ref": "#/definitions/XmlDecoder"
-          - "$ref": "#/definitions/CompositeRawDecoder"
       $parameters:
         type: object
         additionalProperties: true
@@ -2340,23 +2298,6 @@ definitions:
       $parameters:
         type: object
         additionalProperties: true
-  GzipJsonDecoder:
-    title: GzipJson Decoder
-    description: Use this if the response is Gzip compressed Json.
-    type: object
-    additionalProperties: true
-    required:
-      - type
-    properties:
-      type:
-        type: string
-        enum: [GzipJsonDecoder]
-      encoding:
-        type: string
-        default: utf-8
-      $parameters:
-        type: object
-        additionalProperties: true
   ZipfileDecoder:
     title: Zipfile Decoder
     description: Decoder for response data that is returned as zipfile(s).
@@ -2364,19 +2305,19 @@ definitions:
     additionalProperties: true
     required:
       - type
-      - parser
+      - decoder
     properties:
       type:
         type: string
         enum: [ZipfileDecoder]
-      parser:
+      decoder:
         title: Parser
         description: Parser to parse the decompressed data from the zipfile(s).
         anyOf:
-          - "$ref": "#/definitions/GzipParser"
-          - "$ref": "#/definitions/JsonParser"
-          - "$ref": "#/definitions/JsonLineParser"
-          - "$ref": "#/definitions/CsvParser"
+          - "$ref": "#/definitions/CsvDecoder"
+          - "$ref": "#/definitions/GzipDecoder"
+          - "$ref": "#/definitions/JsonDecoder"
+          - "$ref": "#/definitions/JsonlDecoder"
   ListPartitionRouter:
     title: List Partition Router
     description: A Partition router that specifies a list of attributes where each attribute describes a portion of the complete data set for a stream. During a sync, each value is iterated over and can be used as input to outbound API requests.
@@ -3209,79 +3150,39 @@ definitions:
         description: Component decoding the response so records can be extracted.
         anyOf:
           - "$ref": "#/definitions/CustomDecoder"
+          - "$ref": "#/definitions/CsvDecoder"
+          - "$ref": "#/definitions/GzipDecoder"
           - "$ref": "#/definitions/JsonDecoder"
           - "$ref": "#/definitions/JsonlDecoder"
           - "$ref": "#/definitions/IterableDecoder"
           - "$ref": "#/definitions/XmlDecoder"
-          - "$ref": "#/definitions/GzipJsonDecoder"
-          - "$ref": "#/definitions/CompositeRawDecoder"
           - "$ref": "#/definitions/ZipfileDecoder"
       $parameters:
         type: object
         additionalProperties: true
-  CompositeRawDecoder:
-    description: "(This is experimental, use at your own risk)"
+  GzipDecoder:
     type: object
     required:
       - type
-      - parser
+      - decoder
     properties:
       type:
         type: string
-        enum: [CompositeRawDecoder]
-      parser:
-        anyOf:
-          - "$ref": "#/definitions/GzipParser"
-          - "$ref": "#/definitions/JsonParser"
-          - "$ref": "#/definitions/JsonLineParser"
-          - "$ref": "#/definitions/CsvParser"
-  #    PARSERS
-  GzipParser:
-    type: object
-    required:
-      - type
-      - inner_parser
-    properties:
-      type:
-        type: string
-        enum: [GzipParser]
-      inner_parser:
+        enum: [GzipDecoder]
+      decoder:
         anyOf:
-          - "$ref": "#/definitions/JsonLineParser"
-          - "$ref": "#/definitions/CsvParser"
-          - "$ref": "#/definitions/JsonParser"
-  JsonParser:
-    title: JsonParser
-    description: Parser used for parsing str, bytes, or bytearray data and returning data in a dictionary format.
-    type: object
-    required:
-      - type
-    properties:
-      type:
-        type: string
-        enum: [JsonParser]
-      encoding:
-        type: string
-        default: utf-8
-  JsonLineParser:
-    type: object
-    required:
-      - type
-    properties:
-      type:
-        type: string
-        enum: [JsonLineParser]
-      encoding:
-        type: string
-        default: utf-8
-  CsvParser:
+          - "$ref": "#/definitions/CsvDecoder"
+          - "$ref": "#/definitions/GzipDecoder"
+          - "$ref": "#/definitions/JsonDecoder"
+          - "$ref": "#/definitions/JsonlDecoder"
+  CsvDecoder:
     type: object
     required:
       - type
     properties:
       type:
         type: string
-        enum: [CsvParser]
+        enum: [CsvDecoder]
       encoding:
         type: string
         default: utf-8
@@ -3409,24 +3310,24 @@ definitions:
         description: Component decoding the response so records can be extracted.
         anyOf:
           - "$ref": "#/definitions/CustomDecoder"
+          - "$ref": "#/definitions/CsvDecoder"
+          - "$ref": "#/definitions/GzipDecoder"
           - "$ref": "#/definitions/JsonDecoder"
           - "$ref": "#/definitions/JsonlDecoder"
           - "$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.
         anyOf:
           - "$ref": "#/definitions/CustomDecoder"
+          - "$ref": "#/definitions/CsvDecoder"
+          - "$ref": "#/definitions/GzipDecoder"
           - "$ref": "#/definitions/JsonDecoder"
           - "$ref": "#/definitions/JsonlDecoder"
           - "$ref": "#/definitions/IterableDecoder"
           - "$ref": "#/definitions/XmlDecoder"
-          - "$ref": "#/definitions/GzipJsonDecoder"
-          - "$ref": "#/definitions/CompositeRawDecoder"
           - "$ref": "#/definitions/ZipfileDecoder"
       $parameters:
         type: object

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/decoders/__init__.py

@@ -10,10 +10,8 @@ from airbyte_cdk.sources.declarative.decoders.composite_raw_decoder import (
 )
 from airbyte_cdk.sources.declarative.decoders.decoder import Decoder
 from airbyte_cdk.sources.declarative.decoders.json_decoder import (
-    GzipJsonDecoder,
     IterableDecoder,
     JsonDecoder,
-    JsonlDecoder,
 )
 from airbyte_cdk.sources.declarative.decoders.noop_decoder import NoopDecoder
 from airbyte_cdk.sources.declarative.decoders.pagination_decoder_decorator import (
@@ -27,9 +25,7 @@ __all__ = [
     "CompositeRawDecoder",
     "JsonDecoder",
     "JsonParser",
-    "JsonlDecoder",
     "IterableDecoder",
-    "GzipJsonDecoder",
     "NoopDecoder",
     "PaginationDecoderDecorator",
     "XmlDecoder",

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/decoders/composite_raw_decoder.py

@@ -1,5 +1,6 @@
 import csv
 import gzip
+import io
 import json
 import logging
 from abc import ABC, abstractmethod
@@ -130,11 +131,15 @@ class CompositeRawDecoder(Decoder):
     """
 
     parser: Parser
+    stream_response: bool = True
 
     def is_stream_response(self) -> bool:
-        return True
+        return self.stream_response
 
     def decode(
         self, response: requests.Response
     ) -> Generator[MutableMapping[str, Any], None, None]:
-        yield from self.parser.parse(data=response.raw)  # type: ignore[arg-type]
+        if self.is_stream_response():
+            yield from self.parser.parse(data=response.raw)  # type: ignore[arg-type]
+        else:
+            yield from self.parser.parse(data=io.BytesIO(response.content))

airbyte_cdk-6.33.4/airbyte_cdk/sources/declarative/decoders/json_decoder.py

@@ -0,0 +1,65 @@
+#
+# Copyright (c) 2023 Airbyte, Inc., all rights reserved.
+#
+import codecs
+import logging
+from dataclasses import InitVar, dataclass
+from gzip import decompress
+from typing import Any, Generator, List, Mapping, MutableMapping, Optional
+
+import orjson
+import requests
+
+from airbyte_cdk.sources.declarative.decoders import CompositeRawDecoder, JsonParser
+from airbyte_cdk.sources.declarative.decoders.decoder import Decoder
+
+logger = logging.getLogger("airbyte")
+
+
+class JsonDecoder(Decoder):
+    """
+    Decoder strategy that returns the json-encoded content of a response, if any.
+
+    Usually, we would try to instantiate the equivalent `CompositeRawDecoder(parser=JsonParser(), stream_response=False)` but there were specific historical behaviors related to the JsonDecoder that we didn't know if we could remove like the fallback on {} in case of errors.
+    """
+
+    def __init__(self, parameters: Mapping[str, Any]):
+        self._decoder = CompositeRawDecoder(parser=JsonParser(), stream_response=False)
+
+    def is_stream_response(self) -> bool:
+        return self._decoder.is_stream_response()
+
+    def decode(
+        self, response: requests.Response
+    ) -> Generator[MutableMapping[str, Any], None, None]:
+        """
+        Given the response is an empty string or an emtpy list, the function will return a generator with an empty mapping.
+        """
+        has_yielded = False
+        try:
+            for element in self._decoder.decode(response):
+                yield element
+                has_yielded = True
+        except Exception:
+            yield {}
+
+        if not has_yielded:
+            yield {}
+
+
+@dataclass
+class IterableDecoder(Decoder):
+    """
+    Decoder strategy that returns the string content of the response, if any.
+    """
+
+    parameters: InitVar[Mapping[str, Any]]
+
+    def is_stream_response(self) -> bool:
+        return True
+
+    def decode(
+        self, response: requests.Response
+    ) -> Generator[MutableMapping[str, Any], None, None]:
+        for line in response.iter_lines():
+            yield {"record": line.decode()}

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/extractors/record_selector.py

@@ -41,6 +41,7 @@ class RecordSelector(HttpSelector):
     _name: Union[InterpolatedString, str] = field(init=False, repr=False, default="")
     record_filter: Optional[RecordFilter] = None
     transformations: List[RecordTransformation] = field(default_factory=lambda: [])
+    transform_before_filtering: bool = False
 
     def __post_init__(self, parameters: Mapping[str, Any]) -> None:
         self._parameters = parameters
@@ -104,9 +105,17 @@ class RecordSelector(HttpSelector):
         Until we decide to move this logic away from the selector, we made this method public so that users like AsyncJobRetriever could
         share the logic of doing transformations on a set of records.
         """
-        filtered_data = self._filter(all_data, stream_state, stream_slice, next_page_token)
-        transformed_data = self._transform(filtered_data, stream_state, stream_slice)
-        normalized_data = self._normalize_by_schema(transformed_data, schema=records_schema)
+        if self.transform_before_filtering:
+            transformed_data = self._transform(all_data, stream_state, stream_slice)
+            transformed_filtered_data = self._filter(
+                transformed_data, stream_state, stream_slice, next_page_token
+            )
+        else:
+            filtered_data = self._filter(all_data, stream_state, stream_slice, next_page_token)
+            transformed_filtered_data = self._transform(filtered_data, stream_state, stream_slice)
+        normalized_data = self._normalize_by_schema(
+            transformed_filtered_data, schema=records_schema
+        )
         for data in normalized_data:
             yield Record(data=data, stream_name=self.name, associated_slice=stream_slice)
 

{airbyte_cdk-6.33.2.dev1 → airbyte_cdk-6.33.4}/airbyte_cdk/sources/declarative/incremental/concurrent_partition_cursor.py

@@ -58,8 +58,7 @@ class ConcurrentPerPartitionCursor(Cursor):
     CurrentPerPartitionCursor expects the state of the ConcurrentCursor to follow the format {cursor_field: cursor_value}.
     """
 
-    DEFAULT_MAX_PARTITIONS_NUMBER = 25_000
-    SWITCH_TO_GLOBAL_LIMIT = 1000
+    DEFAULT_MAX_PARTITIONS_NUMBER = 10000
     _NO_STATE: Mapping[str, Any] = {}
     _NO_CURSOR_STATE: Mapping[str, Any] = {}
     _GLOBAL_STATE_KEY = "state"
@@ -100,7 +99,7 @@ class ConcurrentPerPartitionCursor(Cursor):
         self._new_global_cursor: Optional[StreamState] = None
         self._lookback_window: int = 0
         self._parent_state: Optional[StreamState] = None
-        self._number_of_partitions: int = 0
+        self._over_limit: int = 0
         self._use_global_cursor: bool = False
         self._partition_serializer = PerPartitionKeySerializer()
 
@@ -234,8 +233,8 @@ class ConcurrentPerPartitionCursor(Cursor):
           or removed due to being the oldest.
         """
         with self._lock:
-            self._number_of_partitions += 1
             while len(self._cursor_per_partition) > self.DEFAULT_MAX_PARTITIONS_NUMBER - 1:
+                self._over_limit += 1
                 # Try removing finished partitions first
                 for partition_key in list(self._cursor_per_partition.keys()):
                     if (
@@ -246,7 +245,7 @@ class ConcurrentPerPartitionCursor(Cursor):
                             partition_key
                         )  # Remove the oldest partition
                         logger.warning(
-                            f"The maximum number of partitions has been reached. Dropping the oldest finished partition: {oldest_partition}. Over limit: {self._number_of_partitions}."
+                            f"The maximum number of partitions has been reached. Dropping the oldest finished partition: {oldest_partition}. Over limit: {self._over_limit}."
                         )
                         break
                 else:
@@ -255,7 +254,7 @@ class ConcurrentPerPartitionCursor(Cursor):
                         1
                     ]  # Remove the oldest partition
                     logger.warning(
-                        f"The maximum number of partitions has been reached. Dropping the oldest partition: {oldest_partition}. Over limit: {self._number_of_partitions}."
+                        f"The maximum number of partitions has been reached. Dropping the oldest partition: {oldest_partition}. Over limit: {self._over_limit}."
                     )
 
     def _set_initial_state(self, stream_state: StreamState) -> None:
@@ -356,10 +355,6 @@ class ConcurrentPerPartitionCursor(Cursor):
 
     def observe(self, record: Record) -> None:
         if not self._use_global_cursor and self.limit_reached():
-            logger.info(
-                f"Exceeded the 'SWITCH_TO_GLOBAL_LIMIT' of {self.SWITCH_TO_GLOBAL_LIMIT}. "
-                f"Switching to global cursor for {self._stream_name}."
-            )
             self._use_global_cursor = True
 
         if not record.associated_slice:
@@ -402,4 +397,4 @@ class ConcurrentPerPartitionCursor(Cursor):
         return cursor
 
     def limit_reached(self) -> bool:
-        return self._number_of_partitions > self.SWITCH_TO_GLOBAL_LIMIT
+        return self._over_limit > self.DEFAULT_MAX_PARTITIONS_NUMBER