Fix cancel-lambda for closed PRs

- Create a fallback function for some edge cases
- Process synchronized PRs as help for cancel-workflow action

tests/ci/cancel_and_rerun_workflow_lambda/app.py

@@ -92,28 +92,33 @@ WorkflowDescription = namedtuple(
 )
 
 
-def get_workflows_description_for_pull_request(pull_request_event):
+def get_workflows_description_for_pull_request(
+    pull_request_event,
+) -> List[WorkflowDescription]:
     head_repo = pull_request_event["head"]["repo"]["full_name"]
     head_branch = pull_request_event["head"]["ref"]
-    head_sha = pull_request_event["head"]["sha"]
     print("PR", pull_request_event["number"], "has head ref", head_branch)
-    workflows_data = []
-    workflows = _exec_get_with_retry(
-        API_URL + f"/actions/runs?branch={head_branch}&event=pull_request&page=1"
-    )
-    workflows_data += workflows["workflow_runs"]
-    i = 2
-    while len(workflows["workflow_runs"]) > 0:
-        workflows = _exec_get_with_retry(
-            API_URL + f"/actions/runs?branch={head_branch}&event=pull_request&page={i}"
-        )
-        workflows_data += workflows["workflow_runs"]
-        i += 1
-        if i > 30:
-            print("Too many workflows found")
-            break
 
-    DEBUG_INFO["workflows"] = []  # type: List[Dict[str, str]]
+    workflows_data = []
+    request_url = f"{API_URL}/actions/runs?per_page=100"
+    # Get all workflows for the current branch
+    for i in range(1, 11):
+        workflows = _exec_get_with_retry(
+            f"{request_url}&event=pull_request&branch={head_branch}&page={i}"
+        )
+        if not workflows["workflow_runs"]:
+            break
+        workflows_data += workflows["workflow_runs"]
+        if i == 10:
+            print("Too many workflows found")
+
+    if not workflows_data:
+        print("No workflows found by filter")
+        return []
+
+    print(f"Total workflows for the branch {head_branch} found: {len(workflows_data)}")
+
+    DEBUG_INFO["workflows"] = []
     workflow_descriptions = []
     for workflow in workflows_data:
         # Some time workflow["head_repository"]["full_name"] is None
@@ -123,13 +128,13 @@ def get_workflows_description_for_pull_request(pull_request_event):
             {
                 "full_name": workflow["head_repository"]["full_name"],
                 "name": workflow["name"],
+                "branch": workflow["head_branch"],
             }
         )
         # unfortunately we cannot filter workflows from forks in request to API
         # so doing it manually
         if (
-            workflow["head_sha"] == head_sha
+            workflow["head_repository"]["full_name"] == head_repo
-            and workflow["head_repository"]["full_name"] == head_repo
             and workflow["name"] in NEED_RERUN_OR_CANCELL_WORKFLOWS
         ):
             workflow_descriptions.append(
@@ -144,6 +149,60 @@ def get_workflows_description_for_pull_request(pull_request_event):
     return workflow_descriptions
 
 
+def get_workflow_description_fallback(event_data) -> List[WorkflowDescription]:
+    pull_request_event = event_data["pull_request"]
+    head_repo = pull_request_event["head"]["repo"]["full_name"]
+    head_branch = pull_request_event["head"]["ref"]
+    head_sha = pull_request_event["head"]["sha"]
+    print("Get last 500 workflows from API to search related there")
+    # Fallback for a case of an already deleted branch and no workflows received
+    request_url = f"{API_URL}/actions/runs?per_page=100"
+    workflows_data = []
+    i = 1
+    for i in range(1, 6):
+        workflows = _exec_get_with_retry(f"{request_url}&page={i}")
+        if not workflows["workflow_runs"]:
+            break
+        # Prefilter workflows
+        workflows_data += [
+            wf
+            for wf in workflows["workflow_runs"]
+            if wf["head_repository"] is not None
+            and wf["head_repository"]["full_name"] == head_repo
+            and wf["head_branch"] == head_branch
+            and wf["name"] in NEED_RERUN_OR_CANCELL_WORKFLOWS
+        ]
+
+    print(f"Total workflows in last 500 actions matches: {len(workflows_data)}")
+
+    DEBUG_INFO["workflows"] = [
+        {
+            "full_name": wf["head_repository"]["full_name"],
+            "name": wf["name"],
+            "branch": wf["head_branch"],
+        }
+        for wf in workflows_data
+    ]
+    if event_data["action"] == "synchronize":
+        print(f"Leave only workflows with SHA but {head_sha} for updated PR")
+        # Cancel all events with SHA different than current
+        workflows_data = list(
+            filter(lambda x: x["head_sha"] != head_sha, workflows_data)
+        )
+
+    workflow_descriptions = [
+        WorkflowDescription(
+            run_id=wf["id"],
+            status=wf["status"],
+            rerun_url=wf["rerun_url"],
+            cancel_url=wf["cancel_url"],
+        )
+        for wf in workflows_data
+    ]
+
+    return workflow_descriptions
+
+
 def get_workflow_description(workflow_id):
     workflow = _exec_get_with_retry(API_URL + f"/actions/runs/{workflow_id}")
     return WorkflowDescription(
@@ -189,6 +248,21 @@ def main(event):
     if action == "closed" or "do not test" in labels:
         print("PR merged/closed or manually labeled 'do not test' will kill workflows")
         workflow_descriptions = get_workflows_description_for_pull_request(pull_request)
+        workflow_descriptions = (
+            workflow_descriptions or get_workflow_description_fallback(event_data)
+        )
+        urls_to_cancel = []
+        for workflow_description in workflow_descriptions:
+            if workflow_description.status != "completed":
+                urls_to_cancel.append(workflow_description.cancel_url)
+        print(f"Found {len(urls_to_cancel)} workflows to cancel")
+        exec_workflow_url(urls_to_cancel, token)
+    elif action == "synchronize":
+        print("PR is synchronized, going to stop old actions")
+        workflow_descriptions = get_workflows_description_for_pull_request(pull_request)
+        workflow_descriptions = (
+            workflow_descriptions or get_workflow_description_fallback(event_data)
+        )
         urls_to_cancel = []
         for workflow_description in workflow_descriptions:
             if workflow_description.status != "completed":