airbyte-cdk 6.45.2__tar.gz → 6.45.2.dev0__tar.gz

{airbyte_cdk-6.45.2 → airbyte_cdk-6.45.2.dev0}/PKG-INFO

@@ -1,6 +1,6 @@
 Metadata-Version: 2.1
 Name: airbyte-cdk
-Version: 6.45.2
+Version: 6.45.2.dev0
 Summary: A framework for writing Airbyte Connectors.
 Home-page: https://airbyte.com
 License: MIT

{airbyte_cdk-6.45.2 → airbyte_cdk-6.45.2.dev0}/airbyte_cdk/sources/declarative/parsers/model_to_component_factory.py

@@ -2054,6 +2054,7 @@ class ModelToComponentFactory:
         config: Config,
         *,
         url_base: str,
+        extractor_model: Optional[Union[CustomRecordExtractorModel, DpathExtractorModel]] = None,
         decoder: Optional[Decoder] = None,
         cursor_used_for_stop_condition: Optional[DeclarativeCursor] = None,
     ) -> Union[DefaultPaginator, PaginatorTestReadDecorator]:
@@ -2075,7 +2076,10 @@ class ModelToComponentFactory:
             else None
         )
         pagination_strategy = self._create_component_from_model(
-            model=model.pagination_strategy, config=config, decoder=decoder_to_use
+            model=model.pagination_strategy,
+            config=config,
+            decoder=decoder_to_use,
+            extractor_model=extractor_model,
         )
         if cursor_used_for_stop_condition:
             pagination_strategy = StopConditionPaginationStrategyDecorator(
@@ -2572,7 +2576,12 @@ class ModelToComponentFactory:
         )
 
     def create_offset_increment(
-        self, model: OffsetIncrementModel, config: Config, decoder: Decoder, **kwargs: Any
+        self,
+        model: OffsetIncrementModel,
+        config: Config,
+        decoder: Decoder,
+        extractor_model: Optional[Union[CustomRecordExtractorModel, DpathExtractorModel]] = None,
+        **kwargs: Any,
     ) -> OffsetIncrement:
         if isinstance(decoder, PaginationDecoderDecorator):
             inner_decoder = decoder.decoder
@@ -2587,10 +2596,24 @@ class ModelToComponentFactory:
                 self._UNSUPPORTED_DECODER_ERROR.format(decoder_type=type(inner_decoder))
             )
 
+        # Ideally we would instantiate the runtime extractor from highest most level (in this case the SimpleRetriever)
+        # so that it can be shared by OffSetIncrement and RecordSelector. However, due to how we instantiate the
+        # decoder with various decorators here, but not in create_record_selector, it is simpler to retain existing
+        # behavior by having two separate extractors with identical behavior since they use the same extractor model.
+        # When we have more time to investigate we can look into reusing the same component.
+        extractor = (
+            self._create_component_from_model(
+                model=extractor_model, config=config, decoder=decoder_to_use
+            )
+            if extractor_model
+            else None
+        )
+
         return OffsetIncrement(
             page_size=model.page_size,
             config=config,
             decoder=decoder_to_use,
+            extractor=extractor,
             inject_on_first_request=model.inject_on_first_request or False,
             parameters=model.parameters or {},
         )
@@ -2954,6 +2977,7 @@ class ModelToComponentFactory:
                 model=model.paginator,
                 config=config,
                 url_base=url_base,
+                extractor_model=model.record_selector.extractor,
                 decoder=decoder,
                 cursor_used_for_stop_condition=cursor_used_for_stop_condition,
             )

{airbyte_cdk-6.45.2 → airbyte_cdk-6.45.2.dev0}/airbyte_cdk/sources/declarative/requesters/paginators/strategies/offset_increment.py

@@ -12,6 +12,7 @@ from airbyte_cdk.sources.declarative.decoders import (
     JsonDecoder,
     PaginationDecoderDecorator,
 )
+from airbyte_cdk.sources.declarative.extractors.dpath_extractor import RecordExtractor
 from airbyte_cdk.sources.declarative.interpolation import InterpolatedString
 from airbyte_cdk.sources.declarative.requesters.paginators.strategies.pagination_strategy import (
     PaginationStrategy,
@@ -46,6 +47,7 @@ class OffsetIncrement(PaginationStrategy):
     config: Config
     page_size: Optional[Union[str, int]]
     parameters: InitVar[Mapping[str, Any]]
+    extractor: Optional[RecordExtractor]
     decoder: Decoder = field(
         default_factory=lambda: PaginationDecoderDecorator(decoder=JsonDecoder(parameters={}))
     )
@@ -75,6 +77,14 @@ class OffsetIncrement(PaginationStrategy):
     ) -> Optional[Any]:
         decoded_response = next(self.decoder.decode(response))
 
+        if self.extractor:
+            page_size_from_response = len(list(self.extractor.extract_records(response=response)))
+            # The extractor could return 0 records which is valid, but evaluates to False. Our fallback in other
+            # cases as the best effort option is to use the incoming last_page_size
+            last_page_size = (
+                page_size_from_response if page_size_from_response is not None else last_page_size
+            )
+
         # Stop paginating when there are fewer records than the page size or the current page has no records
         if (
             self._page_size

{airbyte_cdk-6.45.2 → airbyte_cdk-6.45.2.dev0}/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 = "6.45.2"  # Version will be calculated dynamically.
+version = "6.45.2dev0"  # Version will be calculated dynamically.
 
 [tool.poetry-dynamic-versioning]
 enable = false