fix: surgically add JsonItemsDecoder to generated model Union types

Instead of a full regeneration (which introduced conint/confloat drift that
broke MyPy), this takes the main branch model and makes only the minimal
edits needed:

1. Add JsonItemsDecoder class definition (after JsonDecoder)
2. Add JsonItemsDecoder to GzipDecoder.decoder Union
3. Add JsonItemsDecoder to ZipfileDecoder.decoder Union
4. Add JsonItemsDecoder to SimpleRetriever.decoder Union
5. Add JsonItemsDecoder to AsyncRetriever.decoder Union
6. Add JsonItemsDecoder to AsyncRetriever.download_decoder Union

Verified: MyPy clean (454 source files), ruff lint+format clean,
all decoder (60) and parser (165) tests pass.

Co-Authored-By: Devin AI <158243242+devin-ai-integration[bot]@users.noreply.github.com>

Author: ZaneHyattAB

airbyte_cdk/sources/declarative/models/declarative_component_schema.py

@@ -6,7 +6,7 @@
 from enum import Enum
 from typing import Any, Dict, List, Literal, Optional, Union
 
-from pydantic.v1 import BaseModel, Extra, Field, confloat, conint
+from pydantic.v1 import BaseModel, Extra, Field
 
 from airbyte_cdk.sources.declarative.models.base_model_with_deprecations import (
     BaseModelWithDeprecations,
@@ -18,6 +18,12 @@ class AuthFlowType(Enum):
     oauth1_0 = "oauth1.0"
 
 
+class ScopesJoinStrategy(Enum):
+    space = "space"
+    comma = "comma"
+    plus = "plus"
+
+
 class BasicHttpAuthenticator(BaseModel):
     type: Literal["BasicHttpAuthenticator"]
     username: str = Field(
@@ -51,9 +57,10 @@ class DynamicStreamCheckConfig(BaseModel):
     dynamic_stream_name: str = Field(
         ..., description="The dynamic stream name.", title="Dynamic Stream Name"
     )
-    stream_count: Optional[conint(ge=1)] = Field(
+    stream_count: Optional[int] = Field(
         None,
         description="The number of streams to attempt reading from during a check operation. If unset, all generated streams are checked. Must be a positive integer; if it exceeds the total number of available streams, all streams are checked.",
+        ge=1,
         title="Stream Count",
     )
 
@@ -91,16 +98,17 @@ class ConcurrencyLevel(BaseModel):
 
 class ConstantBackoffStrategy(BaseModel):
     type: Literal["ConstantBackoffStrategy"]
-    backoff_time_in_seconds: Union[confloat(ge=0.0), str] = Field(
+    backoff_time_in_seconds: Union[float, str] = Field(
         ...,
         description="Backoff time in seconds.",
         examples=[30, 30.5, "{{ config['backoff_time'] }}"],
         title="Backoff Time",
     )
-    jitter_range_in_seconds: Optional[confloat(ge=0.0)] = Field(
+    jitter_range_in_seconds: Optional[float] = Field(
         None,
         description="Optional additive jitter range in seconds. When set, the backoff time is uniformly distributed between backoff_time_in_seconds and backoff_time_in_seconds + (jitter_range_in_seconds * 2), so jitter only increases the base backoff.",
         examples=[15],
+        ge=0,
         title="Jitter Range",
     )
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
@@ -488,7 +496,7 @@ class Config:
     )
     weight: Optional[Union[int, str]] = Field(
         None,
-        description="The weight of a request matching this matcher when acquiring a call from the rate limiter. Different endpoints can consume different amounts from a shared budget by specifying different weights. If not set, each request counts as 1.\n",
+        description="The weight of a request matching this matcher when acquiring a call from the rate limiter. Different endpoints can consume different amounts from a shared budget by specifying different weights. If not set, each request counts as 1.",
         title="Weight",
     )
 
@@ -503,32 +511,6 @@ class OnNoRecords(Enum):
     emit_parent = "emit_parent"
 
 
-class RecordExpander(BaseModel):
-    type: Literal["RecordExpander"]
-    expand_records_from_field: List[str] = Field(
-        ...,
-        description="Path to a nested array field within each record. Items from this array will be extracted and emitted as separate records. Supports wildcards (*) for matching multiple arrays.",
-        examples=[
-            ["lines", "data"],
-            ["items"],
-            ["nested", "array"],
-            ["sections", "*", "items"],
-        ],
-        title="Expand Records From Field",
-    )
-    remain_original_record: Optional[bool] = Field(
-        False,
-        description='If true, each expanded record will include the original parent record in an "original_record" field. Defaults to false.',
-        title="Remain Original Record",
-    )
-    on_no_records: Optional[OnNoRecords] = Field(
-        OnNoRecords.skip,
-        description='Behavior when the expansion path is missing, not a list, or an empty list. "skip" (default) emits nothing. "emit_parent" emits the original parent record unchanged.',
-        title="On No Records",
-    )
-    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
-
-
 class ExponentialBackoffStrategy(BaseModel):
     type: Literal["ExponentialBackoffStrategy"]
     factor: Optional[Union[float, str]] = Field(
@@ -537,10 +519,11 @@ class ExponentialBackoffStrategy(BaseModel):
         examples=[5, 5.5, "10"],
         title="Factor",
     )
-    jitter_range_in_seconds: Optional[confloat(ge=0.0)] = Field(
+    jitter_range_in_seconds: Optional[float] = Field(
         None,
         description="Optional additive jitter range in seconds. When set, the backoff time is uniformly distributed between computed_backoff and computed_backoff + (jitter_range_in_seconds * 2), so jitter only increases the computed backoff.",
         examples=[2],
+        ge=0,
         title="Jitter Range",
     )
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
@@ -695,13 +678,12 @@ class JsonItemsDecoder(BaseModel):
     type: Literal["JsonItemsDecoder"]
     items_path: str = Field(
         ...,
-        description="Dot-separated path to the JSON array whose elements should be yielded as records. Uses `ijson` path syntax (e.g. `data.users`), not JSONPath syntax — do not include leading `$.` or trailing `[*]`.",
-        examples=["dataByDepartmentAndSearchTerm", "dataByAsin", "data.users"],
+        description="Dot-separated path to the JSON array whose elements should be yielded as records. Uses `ijson` path syntax (e.g. `data.users`), not JSONPath syntax \u2014 do not include leading `$.` or trailing `[*]`.",
         title="Items Path",
     )
     encoding: Optional[str] = Field(
         "utf-8",
-        description="Text encoding used to decode the streamed bytes before JSON parsing.",
+        description="The character encoding of the JSON data. Defaults to UTF-8.",
         title="Encoding",
     )
 
@@ -864,32 +846,22 @@ class NoPagination(BaseModel):
     type: Literal["NoPagination"]
 
 
-class Scope(BaseModel):
-    class Config:
-        extra = Extra.allow
-
-    scope: str = Field(..., description="The OAuth scope string to request from the provider.")
-
-
-class OptionalScope(BaseModel):
+class State(BaseModel):
     class Config:
         extra = Extra.allow
 
-    scope: str = Field(..., description="The OAuth scope string to request from the provider.")
-
-
-class ScopesJoinStrategy(Enum):
-    space = "space"
-    comma = "comma"
-    plus = "plus"
+    min: int
+    max: int
 
 
-class State(BaseModel):
+class OAuthScope(BaseModel):
     class Config:
         extra = Extra.allow
 
-    min: int
-    max: int
+    scope: str = Field(
+        ...,
+        description="The OAuth scope string to request from the provider.",
+    )
 
 
 class OauthConnectorInputSpecification(BaseModel):
@@ -911,13 +883,17 @@ class Config:
         examples=["user:read user:read_orders workspaces:read"],
         title="Scopes",
     )
-    scopes: Optional[List[Scope]] = Field(
+    # NOTE: scopes, optional_scopes, and scopes_join_strategy are processed by the
+    # platform OAuth handler (DeclarativeOAuthSpecHandler.kt), not by the CDK runtime.
+    # The CDK schema defines the manifest contract; the platform reads these fields
+    # during the OAuth consent flow to build the authorization URL.
+    scopes: Optional[List[OAuthScope]] = Field(
         None,
         description="List of OAuth scope objects. When present, takes precedence over the `scope` string property.\nThe scope values are joined using the `scopes_join_strategy` (default: space) before being\nsent to the OAuth provider.",
         examples=[[{"scope": "user:read"}, {"scope": "user:write"}]],
         title="Scopes",
     )
-    optional_scopes: Optional[List[OptionalScope]] = Field(
+    optional_scopes: Optional[List[OAuthScope]] = Field(
         None,
         description="Optional OAuth scope objects that may or may not be granted.",
         examples=[[{"scope": "admin:read"}]],
@@ -1290,14 +1266,7 @@ class AsyncJobStatusMap(BaseModel):
     completed: List[str]
     failed: List[str]
     timeout: List[str]
-    skipped: Optional[List[str]] = Field(
-        None,
-        description="Statuses that indicate the job was skipped because there is no data to return. Jobs with these statuses will not be retried and no records will be fetched.",
-    )
-
-
-class BlockSimultaneousSyncsAction(BaseModel):
-    type: Literal["BlockSimultaneousSyncsAction"]
+    skipped: Optional[List[str]] = None
 
 
 class ValueType(Enum):
@@ -2138,23 +2107,28 @@ class DefaultPaginator(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
-class DpathExtractor(BaseModel):
-    type: Literal["DpathExtractor"]
-    field_path: List[str] = Field(
+class RecordExpander(BaseModel):
+    type: Literal["RecordExpander"]
+    expand_records_from_field: List[str] = Field(
         ...,
-        description='List of potentially nested fields describing the full path of the field to extract. Use "*" to extract all values from an array. See more info in the [docs](https://docs.airbyte.com/connector-development/config-based/understanding-the-yaml-file/record-selector).',
+        description="Path to a nested array field within each record. Items from this array will be extracted and emitted as separate records. Supports wildcards (*) for matching multiple arrays.",
         examples=[
-            ["data"],
-            ["data", "records"],
-            ["data", "{{ parameters.name }}"],
-            ["data", "*", "record"],
+            ["lines", "data"],
+            ["items"],
+            ["nested", "array"],
+            ["sections", "*", "items"],
         ],
-        title="Field Path",
+        title="Expand Records From Field",
     )
-    record_expander: Optional[RecordExpander] = Field(
-        None,
-        description="Optional component to expand records by extracting items from nested array fields.",
-        title="Record Expander",
+    remain_original_record: Optional[bool] = Field(
+        False,
+        description='If true, each expanded record will include the original parent record in an "original_record" field. Defaults to false.',
+        title="Remain Original Record",
+    )
+    on_no_records: Optional[OnNoRecords] = Field(
+        OnNoRecords.skip,
+        description='Behavior when the expansion path is missing, not a list, or an empty list. "skip" (default) emits nothing. "emit_parent" emits the original parent record unchanged.',
+        title="On No Records",
     )
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
@@ -2227,27 +2201,6 @@ class ListPartitionRouter(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
-class RecordSelector(BaseModel):
-    type: Literal["RecordSelector"]
-    extractor: Union[DpathExtractor, CustomRecordExtractor]
-    record_filter: Optional[Union[RecordFilter, CustomRecordFilter]] = Field(
-        None,
-        description="Responsible for filtering records to be emitted by the Source.",
-        title="Record Filter",
-    )
-    schema_normalization: Optional[Union[SchemaNormalization, CustomSchemaNormalization]] = Field(
-        None,
-        description="Responsible for normalization according to the schema.",
-        title="Schema Normalization",
-    )
-    transform_before_filtering: Optional[bool] = Field(
-        None,
-        description="If true, transformation will be applied before record filtering.",
-        title="Transform Before Filtering",
-    )
-    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
-
-
 class PaginationReset(BaseModel):
     type: Literal["PaginationReset"]
     action: Action1
@@ -2368,6 +2321,27 @@ class Config:
     )
 
 
+class DpathExtractor(BaseModel):
+    type: Literal["DpathExtractor"]
+    field_path: List[str] = Field(
+        ...,
+        description='List of potentially nested fields describing the full path of the field to extract. Use "*" to extract all values from an array. See more info in the [docs](https://docs.airbyte.com/connector-development/config-based/understanding-the-yaml-file/record-selector).',
+        examples=[
+            ["data"],
+            ["data", "records"],
+            ["data", "{{ parameters.name }}"],
+            ["data", "*", "record"],
+        ],
+        title="Field Path",
+    )
+    record_expander: Optional[RecordExpander] = Field(
+        None,
+        description="Optional component to expand records by extracting items from nested array fields.",
+        title="Record Expander",
+    )
+    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
+
+
 class ZipfileDecoder(BaseModel):
     class Config:
         extra = Extra.allow
@@ -2380,6 +2354,27 @@ class Config:
     )
 
 
+class RecordSelector(BaseModel):
+    type: Literal["RecordSelector"]
+    extractor: Union[DpathExtractor, CustomRecordExtractor]
+    record_filter: Optional[Union[RecordFilter, CustomRecordFilter]] = Field(
+        None,
+        description="Responsible for filtering records to be emitted by the Source.",
+        title="Record Filter",
+    )
+    schema_normalization: Optional[Union[SchemaNormalization, CustomSchemaNormalization]] = Field(
+        None,
+        description="Responsible for normalization according to the schema.",
+        title="Schema Normalization",
+    )
+    transform_before_filtering: Optional[bool] = Field(
+        None,
+        description="If true, transformation will be applied before record filtering.",
+        title="Transform Before Filtering",
+    )
+    parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
+
+
 class ConfigMigration(BaseModel):
     type: Literal["ConfigMigration"]
     description: Optional[str] = Field(
@@ -2472,7 +2467,7 @@ class Config:
     api_budget: Optional[HTTPAPIBudget] = None
     stream_groups: Optional[Dict[str, StreamGroup]] = Field(
         None,
-        description="Groups of streams that share a common resource and should not be read simultaneously. Each group defines a set of stream references and an action that controls how concurrent reads are managed. Only applies to ConcurrentDeclarativeSource.\n",
+        description="Groups of streams that share a common resource and should not be read simultaneously. Each group defines a set of stream references and an action that controls how concurrent reads are managed. Only applies to ConcurrentDeclarativeSource.",
         title="Stream Groups",
     )
     max_concurrent_async_job_count: Optional[Union[int, str]] = Field(
@@ -2512,7 +2507,7 @@ class Config:
     api_budget: Optional[HTTPAPIBudget] = None
     stream_groups: Optional[Dict[str, StreamGroup]] = Field(
         None,
-        description="Groups of streams that share a common resource and should not be read simultaneously. Each group defines a set of stream references and an action that controls how concurrent reads are managed. Only applies to ConcurrentDeclarativeSource.\n",
+        description="Groups of streams that share a common resource and should not be read simultaneously. Each group defines a set of stream references and an action that controls how concurrent reads are managed. Only applies to ConcurrentDeclarativeSource.",
         title="Stream Groups",
     )
     max_concurrent_async_job_count: Optional[Union[int, str]] = Field(
@@ -3008,7 +3003,7 @@ class StateDelegatingStream(BaseModel):
     )
     api_retention_period: Optional[str] = Field(
         None,
-        description="The data retention period of the incremental API (ISO8601 duration). If the cursor value is older than this retention period, the connector will automatically fall back to a full refresh to avoid data loss.\nThis is useful for APIs like Stripe Events API which only retain data for 30 days.\n* **PT1H**: 1 hour\n* **P1D**: 1 day\n* **P1W**: 1 week\n* **P1M**: 1 month\n* **P1Y**: 1 year\n* **P30D**: 30 days\n",
+        description="The data retention period of the incremental API (ISO8601 duration). If the cursor value is older than this retention period, the connector will automatically fall back to a full refresh to avoid data loss.\nThis is useful for APIs like Stripe Events API which only retain data for 30 days.\n  * **PT1H**: 1 hour\n  * **P1D**: 1 day\n  * **P1W**: 1 week\n  * **P1M**: 1 month\n  * **P1Y**: 1 year\n  * **P30D**: 30 days\n",
         examples=["P30D", "P90D", "P1Y"],
         title="API Retention Period",
     )
@@ -3108,9 +3103,10 @@ class AsyncRetriever(BaseModel):
         None,
         description="The time in minutes after which the single Async Job should be considered as Timed Out.",
     )
-    failed_retry_wait_time_in_seconds: Optional[Union[conint(ge=1), str]] = Field(
+    failed_retry_wait_time_in_seconds: Optional[Union[int, str]] = Field(
         None,
         description="Time in seconds to wait before retrying a failed async job. Only applies to jobs that ran on the API side and reported a FAILED status (e.g. report generation failed due to a cooldown). Creation failures (HTTP errors when starting a job, such as 429s) and TIMED_OUT jobs are retried immediately and are not affected by this setting. When set, the orchestrator defers retry of real failed jobs until the wait time has elapsed, without blocking other jobs.",
+        ge=1,
     )
     download_target_requester: Optional[Union[HttpRequester, CustomRequester]] = Field(
         None,
@@ -3189,14 +3185,20 @@ class AsyncRetriever(BaseModel):
     parameters: Optional[Dict[str, Any]] = Field(None, alias="$parameters")
 
 
+class BlockSimultaneousSyncsAction(BaseModel):
+    type: Literal["BlockSimultaneousSyncsAction"]
+
+
 class StreamGroup(BaseModel):
-    streams: List[DeclarativeStream] = Field(
+    streams: List[str] = Field(
         ...,
-        description="List of references to streams that belong to this group.\n",
+        description='List of references to streams that belong to this group. Use JSON references to stream definitions (e.g., "#/definitions/my_stream").',
         title="Streams",
     )
     action: BlockSimultaneousSyncsAction = Field(
-        ..., description="The action to apply to streams in this group.", title="Action"
+        ...,
+        description="The action to apply to streams in this group.",
+        title="Action",
     )