Cache busting for wot

Author: mephenor

src/ghga_connector/core/downloading/api_calls.py

@@ -31,7 +31,7 @@
 
 
 async def _get_authorization(
-    file_id: str, work_package_accessor: WorkPackageAccessor
+    file_id: str, work_package_accessor: WorkPackageAccessor, bust_cache: bool = False
 ) -> httpx.Headers:
     """
     Fetch work order token using accessor and prepare DCS endpoint URL and headers for a
@@ -41,7 +41,9 @@ async def _get_authorization(
     for at least another `CACHE_MIN_FRESH` seconds.
     """
     # fetch a work order token
-    decrypted_token = await work_package_accessor.get_work_order_token(file_id=file_id)
+    decrypted_token = await work_package_accessor.get_work_order_token(
+        file_id=file_id, bust_cache=bust_cache
+    )
     # build headers
     headers = httpx.Headers(
         {
@@ -71,7 +73,7 @@ async def get_envelope_authorization(
 
 
 async def get_file_authorization(
-    file_id: str, work_package_accessor: WorkPackageAccessor
+    file_id: str, work_package_accessor: WorkPackageAccessor, bust_cache: bool = False
 ) -> UrlAndHeaders:
     """
     Fetch work order token using accessor and prepare DCS endpoint URL and headers to get
@@ -80,7 +82,9 @@ async def get_file_authorization(
     # build URL
     url = f"{work_package_accessor.dcs_api_url}/objects/{file_id}"
     headers = await _get_authorization(
-        file_id=file_id, work_package_accessor=work_package_accessor
+        file_id=file_id,
+        work_package_accessor=work_package_accessor,
+        bust_cache=bust_cache,
     )
     return UrlAndHeaders(url, headers)
 

src/ghga_connector/core/downloading/batch_processing.py

@@ -218,6 +218,7 @@ async def _check_file(self, file_id: str) -> None:
                 url_and_headers = await get_file_authorization(
                     file_id=file_id,
                     work_package_accessor=self.work_package_accessor,
+                    bust_cache=True,
                 )
                 response = await get_download_url(
                     client=self.client, url_and_headers=url_and_headers, bust_cache=True

src/ghga_connector/core/downloading/downloader.py

@@ -204,6 +204,7 @@ async def fetch_download_url(self, bust_cache: bool = False) -> URLResponse:
                 url_and_headers = await get_file_authorization(
                     file_id=self._file_id,
                     work_package_accessor=self._work_package_accessor,
+                    bust_cache=True,
                 )
                 response = await get_download_url(
                     client=self._client,

src/ghga_connector/core/work_package.py

@@ -92,7 +92,9 @@ async def get_package_files(self) -> dict[str, str]:
         work_package = response.json()
         return work_package["files"]
 
-    async def get_work_order_token(self, *, file_id: str) -> str:
+    async def get_work_order_token(
+        self, *, file_id: str, bust_cache: bool = False
+    ) -> str:
         """Call WPS endpoint to retrieve and decrypt work order token."""
         url = f"{self.api_url}/work-packages/{self.package_id}/files/{file_id}/work-order-tokens"
 
@@ -103,6 +105,12 @@ async def get_work_order_token(self, *, file_id: str) -> str:
                 "Cache-Control": f"min-fresh={CACHE_MIN_FRESH}",
             }
         )
+        if bust_cache:
+            # update cache-control headers to get fresh response from source
+            cache_control_headers = headers.get("Cache-Control")
+            cache_control_headers = [cache_control_headers, "max-age=0"]
+            headers["Cache-Control"] = ",".join(cache_control_headers)
+
         response = await self._call_url(fn=self.client.post, headers=headers, url=url)
 
         status_code = response.status_code