More fixes for Pyright.

Author: jonathan-cantu-icf

apps/bfd-model-idr/augment_sample_resources.py

@@ -261,7 +261,8 @@ def create_rendering_line_provider(npi_num):
     matching_rows = cond_sk_df[cond_sk_df["CLM_RLT_COND_SGNTR_SK"] == str(cond_sk)]
     for _, row in matching_rows.iterrows():
         cond_cd = row.get("CLM_RLT_COND_CD")
-        if pd.notna(cond_cd) and str(cond_cd) != "~":
+        # Use manual NaN check instead of pd.isna()
+        if cond_cd is not None and cond_cd == cond_cd and str(cond_cd) != "~":
             supporting_info_components.append({"CLM_RLT_COND_CD": str(cond_cd)})
 
 fac_type = cur_sample_data.get("CLM_BILL_FAC_TYPE_CD")

apps/bfd-pipeline-idr/extractor.py

@@ -32,12 +32,14 @@
     IDR_WAREHOUSE,
     MIN_BATCH_COMPLETION_DATE,
 )
-from timer import Timer
+from .timer import Timer
+from typing import Iterator, Sequence
+from model.base_model import IdrBaseModel
+from abc import ABC, abstractmethod
 
 logger = logging.getLogger(__name__)
 
-
-class Extractor[T](ABC):
+class Extractor[T: IdrBaseModel](ABC):
     def __init__(self, cls: type[T], partition: LoadPartition) -> None:
         self.cls = cls
         self.type_adapter = TypeAdapter(list[self.cls])
@@ -66,10 +68,7 @@ def _greatest_col(self, cols: list[str]) -> str:
 
     def _get_batch_size(self) -> int:
         if ENABLE_DATE_PARTITIONS:
-            # Larger tables take up more memory, so we'll try to normalize
-            # the total memory used here based on the number of columns
             return round(BATCH_MULTIPLIER / len(self.cls.columns_raw()))
-        # If date partitioning is not enabled, the number of concurrent jobs will be small
         return 100_000
 
     def get_query(self, start_time: datetime, load_mode: LoadMode) -> str:
@@ -83,15 +82,15 @@ def extract_idr_data(
     ) -> Iterator[Sequence[T]]:
         is_historical = progress is None or progress.is_historical()
         fetch_query = self.get_query(start_time, load_mode)
-        # GREATEST doesn't work with nulls so we need to coalesce here
         batch_timestamp_cols = self._coalesce_dates(
             self.cls.batch_timestamp_col_alias(is_historical)
         )
         update_timestamp_cols = self._coalesce_dates(self.cls.update_timestamp_col_alias())
-        # We need to create batches using the most recent timestamp from all of the
-        # insert/update timestamps
         batch_timestamp_clause = self._greatest_col([*batch_timestamp_cols, *update_timestamp_cols])
-        min_transaction_date = self.cls.model_type().min_transaction_date
+        
+        # Fix: defer type checking for dynamic model_type()
+        model_instance = self.cls.model_type()  # type: ignore[call-arg]
+        min_transaction_date = model_instance.min_transaction_date
 
         batch_id_order = ""
         batch_id_clause = ""
@@ -101,7 +100,6 @@ def extract_idr_data(
         logger.info("extracting %s", self.cls.table())
         order_by = f"ORDER BY {batch_timestamp_clause} {batch_id_order}"
         if progress is None:
-            # No saved progress, process the whole table from the beginning
             return self.extract_many(
                 fetch_query.replace(
                     "{WHERE_CLAUSE}",

apps/bfd-pipeline-idr/loader.py

@@ -17,7 +17,7 @@
     bfd_db_username,
     force_load_progress,
 )
-from timer import Timer
+from .timer import Timer
 
 logger = logging.getLogger(__name__)
 

apps/utils/locust_tests/common/bfd_user_base.py

@@ -16,7 +16,6 @@
 from common.validation import ValidationResult
 from locust import FastHttpUser, events
 from locust.argument_parser import LocustArgumentParser
-from locust.contrib.fasthttp import ResponseContextManager
 from locust.env import Environment
 
 _COMPARISONS_METADATA_PATH = None
@@ -188,11 +187,7 @@ def get_by_url(
             catch_response=True,
         ) as response:
             if response.status_code != 200:
-                if isinstance(response, ResponseContextManager):
-                    # pylint: disable=E1121
-                    response.failure(f"Status Code: {response.status_code}")
-                else:
-                    response.failure()
+                response.failure(f"Status Code: {response.status_code}")
             elif response.text:
                 # Check for valid "next" URLs that we can add to a URL pool.
                 next_url = BFDUserBase.__get_next_url(response.text)
@@ -226,11 +221,7 @@ def post_by_url(
             catch_response=True,
         ) as response:
             if response.status_code != 200:
-                if isinstance(response, ResponseContextManager):
-                    # pylint: disable=E1121
-                    response.failure(f"Status Code: {response.status_code}")
-                else:
-                    response.failure()
+                response.failure(f"Status Code: {response.status_code}")
             elif response.text:
                 # Check for valid "next" URLs that we can add to a URL pool.
                 next_url = BFDUserBase.__get_next_url(response.text)

apps/utils/locust_tests/services/server-load/controller/controller.py

@@ -93,20 +93,19 @@ def _start_node(
 
     return response
 
-
 def _main() -> None:
     try:
         cluster_id = get_ssm_parameter(
-            ssm_client=ssm_client,
+            ssm_client=ssm_client,  # type: ignore[arg-type]
             name=f"/bfd/{environment}/common/nonsensitive/rds_cluster_identifier",
         )
         username = get_ssm_parameter(
-            ssm_client=ssm_client,
+            ssm_client=ssm_client,  # type: ignore[arg-type]
             name=f"/bfd/{environment}/server/sensitive/db/username",
             with_decrypt=True,
         )
         raw_password = get_ssm_parameter(
-            ssm_client=ssm_client,
+            ssm_client=ssm_client,  # type: ignore[arg-type]
             name=f"/bfd/{environment}/server/sensitive/db/password",
             with_decrypt=True,
         )

pyproject.toml

@@ -5,12 +5,14 @@ requires-python="==3.14.3"
 description = "BFD Python workspace for unified dependency management"
 
 dependencies = [
+    "tqdm",
     "psycopg[binary,pool]>=3.2.13",
     "pydantic==2.13.3",
     # sftp-outbound-transfer dependencies
     "aws-lambda-powertools[all]",
     "paramiko",
     "boto3",
+    "types-boto3>=1.43.2",
     "aiohttp>=3.13.0",
     "asyncclick>=8.3.0.5",
     "anyio>=4.11.0",
@@ -46,7 +48,6 @@ members = [
 
 [dependency-groups]
 dev = [
-    "tqdm",
     "ruff",
     "pyright",
     # bfd-pipeline-idr dev dependencies
@@ -123,6 +124,7 @@ docstring-code-format = true
 typeCheckingMode = "strict"
 venvPath = "."
 venv = ".venv"
+# TODO: Move these downstream for lambdas with powertools only:
 reportMissingImports = false
 reportMissingTypeStubs = "none"
 # AWS Lambda Powertools compatibility - relaxed type checking for ops services
@@ -138,18 +140,15 @@ exclude = [
     "**/.venv",
     "**/node_modules",
     "**/__pycache__",
-#    "**/bfd-model-idr/**",
+    "apps/utils/locust_tests/v1/**",
+    "apps/utils/locust_tests/v2/**",
     "**/bfd-model-rif/**",
-    "**/bfd-pipeline-idr/**",
-#    "**/locust_tests/**",
-#    "**/synthetic-load-converter/**",
-#    "**/sftp_outbound_transfer/**",
     "apps/bfd-model/**",
     "apps/bfd-pipeline/**",
     "apps/bfd-server/**",
-    "apps/utils/**",
+    "ops/utils/synthetic-data/**",
+    "ops/services/02-insights-bb2/**",
     "docs/**",
-    "ops/**",
     "insights/**",
     "src/**",
     "static-site/**",