Automatically detect if non-lazy logging interpolation is used (#24910)

We used to have pylint check that was preventing it but since
we have no pylint, we need to do some intelligent guessing
based on AST of the python code.

.pre-commit-config.yaml

@@ -798,6 +798,14 @@ repos:
         pass_filenames: true
         files: ^docs/apache-airflow-providers-[^/]*/index\.rst$
         additional_dependencies: ['rich>=12.4.4', 'pyyaml']
+      - id: check-lazy-logging
+        name: Check that all logging methods are lazy
+        entry: ./scripts/ci/pre_commit/pre_commit_check_lazy_logging.py
+        language: python
+        pass_filenames: true
+        files: \.py$
+        exclude: ^airflow/_vendor/
+        additional_dependencies: ['rich>=12.4.4', 'astor']
       - id: create-missing-init-py-files-tests
         name: Create missing init.py files in tests
         entry: ./scripts/ci/pre_commit/pre_commit_check_init_in_tests.py

STATIC_CODE_CHECKS.rst

@@ -175,6 +175,8 @@ require Breeze Docker image to be build locally.
 +--------------------------------------------------------+------------------------------------------------------------------+---------+
 | check-integrations-are-consistent                      | Check if integration list is consistent in various places        |         |
 +--------------------------------------------------------+------------------------------------------------------------------+---------+
+| check-lazy-logging                                     | Check that all logging methods are lazy                          |         |
++--------------------------------------------------------+------------------------------------------------------------------+---------+
 | check-merge-conflict                                   | Check that merge conflicts are not being committed               |         |
 +--------------------------------------------------------+------------------------------------------------------------------+---------+
 | check-newsfragments-are-valid                          | Check newsfragments are valid                                    |         |

airflow/providers/databricks/operators/databricks.py

@@ -85,12 +85,12 @@ def _handle_deferrable_databricks_operator_execution(operator, hook, log, contex
     """
     if operator.do_xcom_push and context is not None:
         context['ti'].xcom_push(key=XCOM_RUN_ID_KEY, value=operator.run_id)
-    log.info(f'Run submitted with run_id: {operator.run_id}')
+    log.info('Run submitted with run_id: %s', operator.run_id)
 
     run_page_url = hook.get_run_page_url(operator.run_id)
     if operator.do_xcom_push and context is not None:
         context['ti'].xcom_push(key=XCOM_RUN_PAGE_URL_KEY, value=run_page_url)
-    log.info(f'View run status, Spark UI, and logs at {run_page_url}')
+    log.info('View run status, Spark UI, and logs at %s', run_page_url)
 
     if operator.wait_for_termination:
         operator.defer(
@@ -107,7 +107,7 @@ def _handle_deferrable_databricks_operator_completion(event: dict, log: Logger)
     validate_trigger_event(event)
     run_state = RunState.from_json(event['run_state'])
     run_page_url = event['run_page_url']
-    log.info(f'View run status, Spark UI, and logs at {run_page_url}')
+    log.info('View run status, Spark UI, and logs at %s', run_page_url)
 
     if run_state.is_successful:
         log.info('Job run completed successfully.')

airflow/providers/google/cloud/operators/bigquery.py

@@ -2152,7 +2152,7 @@ def execute(self, context: Any):
         )
 
         try:
-            self.log.info(f"Executing: {self.configuration}")
+            self.log.info("Executing: %s'", self.configuration)
             job = self._submit_job(hook, job_id)
             self._handle_job_error(job)
         except Conflict:

airflow/providers/sftp/hooks/sftp.py

@@ -202,7 +202,7 @@ def create_directory(self, path: str, mode: int = 777) -> None:
         """
         conn = self.get_conn()
         if self.isdir(path):
-            self.log.info(f"{path} already exists")
+            self.log.info("%s already exists", path)
             return
         elif self.isfile(path):
             raise AirflowException(f"{path} already exists and is a file")
@@ -211,7 +211,7 @@ def create_directory(self, path: str, mode: int = 777) -> None:
             if dirname and not self.isdir(dirname):
                 self.create_directory(dirname, mode)
             if basename:
-                self.log.info(f"Creating {path}")
+                self.log.info("Creating %s", path)
                 conn.mkdir(path, mode=mode)
 
     def delete_directory(self, path: str) -> None:

airflow/providers/sftp/sensors/sftp.py

@@ -62,7 +62,7 @@ def __init__(
 
     def poke(self, context: 'Context') -> bool:
         self.hook = SFTPHook(self.sftp_conn_id)
-        self.log.info(f"Poking for {self.path}, with pattern {self.file_pattern}")
+        self.log.info("Poking for %s, with pattern %s", self.path, self.file_pattern)
 
         if self.file_pattern:
             file_from_pattern = self.hook.get_file_by_pattern(self.path, self.file_pattern)

dev/breeze/src/airflow_breeze/pre_commit_ids.py

@@ -44,6 +44,7 @@
     'check-hooks-apply',
     'check-incorrect-use-of-LoggingMixin',
     'check-integrations-are-consistent',
+    'check-lazy-logging',
     'check-merge-conflict',
     'check-newsfragments-are-valid',
     'check-no-providers-in-core-examples',

images/breeze/output-commands-hash.txt

@@ -28,7 +28,7 @@ self-upgrade:b5437c0a1a91533a11ee9d0a9692369c
 setup-autocomplete:355b72dee171c2fcba46fc90ac7c97b0
 shell:4680295fdd8a276d51518d29360c365c
 start-airflow:92cf775a952439a32d409cd2536da507
-static-checks:bc6b9a121ea38404ac5f28e727146b90
+static-checks:8b4c0a3891e1d65ddf832ec4f5b71491
 stop:8ebd8a42f1003495d37b884de5ac7ce6
 tests:e39111ecbd33a65ababb3cbfbac2b069
 verify-image:a6b3c70957aea96a5d4d261f23359a2d