Stop stripping sync from task kwargs and propagate it to sub-tasks

dojo_dispatch_task was popping sync from kwargs before running the
task, preventing tasks from knowing they should run their own
sub-tasks synchronously. Remove the pop so sync flows through to
task functions. Add explicit sync parameter to
post_process_findings_batch and forward it to calculate_grade so
product grading also runs inline when the caller requests sync
execution.

Author: valentijnscholten

dojo/celery_dispatch.py

@@ -84,8 +84,6 @@ def dojo_dispatch_task(task_or_sig: _SupportsSi | _SupportsApplyAsync | Signatur
     # Track foreground execution as a "created task" as well (matches historical dojo_async_task behavior)
     dojo_async_task_counter.incr(str(sig.task), args=sig.args, kwargs=sig_kwargs)
 
-    sig_kwargs.pop("sync", None)
-    sig = sig.clone(kwargs=sig_kwargs)
     eager = sig.apply()
     try:
         return eager.get(propagate=True)

dojo/finding/helper.py

@@ -468,6 +468,7 @@ def post_process_findings_batch(
     push_to_jira=False,
     jira_instance_id=None,
     user=None,
+    sync=False,
     **kwargs,
 ):
 
@@ -511,7 +512,7 @@ def post_process_findings_batch(
     if product_grading_option and system_settings.enable_product_grade:
         from dojo.celery_dispatch import dojo_dispatch_task  # noqa: PLC0415 circular import
 
-        dojo_dispatch_task(calculate_grade, findings[0].test.engagement.product.id)
+        dojo_dispatch_task(calculate_grade, findings[0].test.engagement.product.id, sync=sync)
 
     # If we received the ID of a jira instance, then we need to determine the keep in sync behavior
     jira_instance = None