Ensure QA run WACZs are deleted (#1715)

- When qa run is deleted
- When crawl is deleted

And adds tests for WACZ deletion.

Fixes #1713

backend/btrixcloud/basecrawls.py

@@ -23,6 +23,7 @@ from .models import (
     StorageRef,
     RUNNING_AND_STARTING_STATES,
     SUCCESSFUL_STATES,
+    QARun,
 )
 from .pagination import paginated_format, DEFAULT_PAGE_SIZE
 from .utils import dt_now
@@ -317,6 +318,7 @@ class BaseCrawlOps:
 
             if type_ == "crawl":
                 await self.page_ops.delete_crawl_pages(crawl_id, org.id)
+                await self.delete_all_crawl_qa_files(crawl_id, org)
 
             crawl_size = await self._delete_crawl_files(crawl, org)
             size += crawl_size
@@ -351,16 +353,20 @@ class BaseCrawlOps:
 
         return res.deleted_count, cids_to_update, quota_reached
 
-    async def _delete_crawl_files(self, crawl: BaseCrawl, org: Organization):
+    async def _delete_crawl_files(
+        self, crawl: Union[BaseCrawl, QARun], org: Organization
+    ):
         """Delete files associated with crawl from storage."""
         size = 0
         for file_ in crawl.files:
             size += file_.size
             if not await self.storage_ops.delete_crawl_file_object(org, file_):
                 raise HTTPException(status_code=400, detail="file_deletion_error")
-            await self.background_job_ops.create_delete_replica_jobs(
-                org, file_, crawl.id, crawl.type
-            )
+            # Not replicating QA run WACZs yet
+            if not isinstance(crawl, QARun):
+                await self.background_job_ops.create_delete_replica_jobs(
+                    org, file_, crawl.id, crawl.type
+                )
 
         return size
 
@@ -370,6 +376,14 @@ class BaseCrawlOps:
         org = await self.orgs.get_org_by_id(oid)
         return await self._delete_crawl_files(crawl, org)
 
+    async def delete_all_crawl_qa_files(self, crawl_id: str, org: Organization):
+        """Delete files for all qa runs in a crawl"""
+        crawl_raw = await self.get_crawl_raw(crawl_id)
+        qa_finished = crawl_raw.get("qaFinished", {})
+        for qa_run_raw in qa_finished.values():
+            qa_run = QARun(**qa_run_raw)
+            await self._delete_crawl_files(qa_run, org)
+
     async def _resolve_crawl_refs(
         self,
         crawl: Union[CrawlOut, CrawlOutWithResources],

backend/btrixcloud/crawls.py

@@ -828,11 +828,15 @@ class CrawlOps(BaseCrawlOps):
                 status_code=404, detail=f"crawl_not_found, (details: {exc})"
             )
 
-    async def delete_crawl_qa_runs(self, crawl_id: str, delete_list: DeleteQARunList):
+    async def delete_crawl_qa_runs(
+        self, crawl_id: str, delete_list: DeleteQARunList, org: Organization
+    ):
         """delete specified finished QA run"""
-
         count = 0
         for qa_run_id in delete_list.qa_run_ids:
+            await self.page_ops.delete_qa_run_from_pages(crawl_id, qa_run_id)
+            await self.delete_crawl_qa_run_files(crawl_id, qa_run_id, org)
+
             res = await self.crawls.find_one_and_update(
                 {"_id": crawl_id, "type": "crawl"},
                 {"$unset": {f"qaFinished.{qa_run_id}": ""}},
@@ -841,10 +845,21 @@ class CrawlOps(BaseCrawlOps):
             if res:
                 count += 1
 
-            await self.page_ops.delete_qa_run_from_pages(crawl_id, qa_run_id)
-
         return {"deleted": count}
 
+    async def delete_crawl_qa_run_files(
+        self, crawl_id: str, qa_run_id: str, org: Organization
+    ):
+        """delete crawl qa wacz files"""
+        qa_run = await self.get_qa_run(crawl_id, qa_run_id, org)
+        for file_ in qa_run.files:
+            if not await self.storage_ops.delete_crawl_file_object(org, file_):
+                raise HTTPException(status_code=400, detail="file_deletion_error")
+            # Not replicating QA run WACZs yet
+            # await self.background_job_ops.create_delete_replica_jobs(
+            #     org, file_, qa_run_id, "qa"
+            # )
+
     async def qa_run_finished(self, crawl_id: str):
         """clear active qa, add qa run to finished list, if successful"""
         crawl = await self.get_crawl(crawl_id)
@@ -900,10 +915,10 @@ class CrawlOps(BaseCrawlOps):
         qa = crawl_data.get("qa")
         return QARunOut(**qa) if qa else None
 
-    async def get_qa_run_for_replay(
+    async def get_qa_run(
         self, crawl_id: str, qa_run_id: str, org: Optional[Organization] = None
-    ) -> QARunWithResources:
-        """Fetch QA runs with resources for replay.json"""
+    ):
+        """Get QARun by id"""
         crawl = await self.get_crawl(crawl_id, org)
         qa_finished = crawl.qaFinished or {}
         qa_run = qa_finished.get(qa_run_id)
@@ -911,6 +926,15 @@ class CrawlOps(BaseCrawlOps):
         if not qa_run:
             raise HTTPException(status_code=404, detail="crawl_qa_not_found")
 
+        return qa_run
+
+    async def get_qa_run_for_replay(
+        self, crawl_id: str, qa_run_id: str, org: Optional[Organization] = None
+    ) -> QARunWithResources:
+        """Fetch QA runs with resources for replay.json"""
+        crawl = await self.get_crawl(crawl_id, org)
+        qa_run = await self.get_qa_run(crawl_id, qa_run_id, org)
+
         if not org:
             org = await self.orgs.get_org_by_id(crawl.oid)
             if not org:
@@ -1212,8 +1236,7 @@ def init_crawls_api(crawl_manager: CrawlManager, app, user_dep, *args):
         qa_run_ids: DeleteQARunList,
         org: Organization = Depends(org_crawl_dep),
     ):
-        # pylint: disable=unused-argument
-        return await ops.delete_crawl_qa_runs(crawl_id, qa_run_ids)
+        return await ops.delete_crawl_qa_runs(crawl_id, qa_run_ids, org)
 
     @app.get(
         "/orgs/{oid}/crawls/{crawl_id}/qa",

backend/test/test_qa.py

@@ -549,6 +549,16 @@ def test_delete_qa_runs(
     qa_run_pages_ready,
     failed_qa_run_id,
 ):
+    # Get download links for QA WACZs
+    r = requests.get(
+        f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawler_crawl_id}/qa/{qa_run_id}/replay.json",
+        headers=crawler_auth_headers,
+    )
+    data = r.json()
+    assert len(data["resources"]) == 1
+    qa_wacz_url = data["resources"][0]["path"]
+
+    # Delete QA runs
     r = requests.post(
         f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawler_crawl_id}/qa/delete",
         json={"qa_run_ids": [qa_run_id, failed_qa_run_id]},
@@ -575,6 +585,10 @@ def test_delete_qa_runs(
         time.sleep(5)
         count += 1
 
+    # Ensure QA WACZs was deleted
+    r = requests.get(f"http://localhost:30870{qa_wacz_url}")
+    assert r.status_code == 404
+
     # Ensure associated qa run information in pages is also deleted
     for qa_run in (qa_run_id, failed_qa_run_id):
         count = 0

backend/test/test_run_crawl.py

@@ -864,6 +864,18 @@ def test_delete_crawls_crawler(
     assert r.status_code == 200
     assert r.json()["total"] > 0
 
+    # Get WACZ presigned url for crawl about to delete
+    wacz_presigned_urls = []
+    r = requests.get(
+        f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawler_crawl_id}/replay.json",
+        headers=crawler_auth_headers,
+    )
+    assert r.status_code == 200
+    data = r.json()
+    assert len(data["resources"]) >= 1
+    for resource in data["resources"]:
+        wacz_presigned_urls.append(resource["path"])
+
     # Test that crawler user can delete own crawl
     r = requests.post(
         f"{API_PREFIX}/orgs/{default_org_id}/crawls/delete",
@@ -884,6 +896,11 @@ def test_delete_crawls_crawler(
     )
     assert r.status_code == 404
 
+    # Test that WACZs are deleted
+    for wacz_url in wacz_presigned_urls:
+        r = requests.get(f"http://localhost:30870{wacz_url}")
+        assert r.status_code == 404
+
     # Test that associated pages are also deleted
     r = requests.get(
         f"{API_PREFIX}/orgs/{default_org_id}/crawls/{crawler_crawl_id}/pages",