Merge pull request #156 from wri/develop

Batman fixes to master

src/datapump/clients/data_api.py

@@ -129,14 +129,24 @@ def create_vector_version(
     def create_version(
         self, dataset: str, version: str, payload: Dict[str, Any]
     ) -> Dict[str, Any]:
-
-        uri = f"{GLOBALS.data_api_uri}/dataset/{dataset}/{version}"
-        return self._send_request(ValidMethods.put, uri, payload)["data"]
+        try:
+            uri = f"{GLOBALS.data_api_uri}/dataset/{dataset}/{version}"
+            return self._send_request(ValidMethods.put, uri, payload)["data"]
+        except DataApiResponseError as e:
+            # Workaround for GTC-2986
+            # Getting a 500 response when creating version, but version is still created
+            # causing a 400 response on subsequent retries since we're trying to PUT
+            # an already existing version.
+            # For now, let's just return the version if it exists.
+            # Otherwise, propagate original exception.
+            try:
+                return self.get_version(dataset, version)
+            except DataApiResponseError:
+                raise e
 
     def create_aux_asset(
         self, dataset: str, version: str, payload: Dict[str, Any]
     ) -> Dict[str, Any]:
-
         uri = f"{GLOBALS.data_api_uri}/dataset/{dataset}/{version}/assets"
         return self._send_request(ValidMethods.post, uri, payload)["data"]
 

src/datapump/sync/rw_areas.py

@@ -3,6 +3,7 @@
 import os
 import traceback
 from contextlib import contextmanager
+from datetime import datetime, timedelta
 from typing import Any, Dict, Iterator, List, Optional, Set, Tuple
 
 import requests
@@ -82,8 +83,10 @@ def get_pending_areas() -> List[Any]:
     LOGGER.info(f"Using token {token()} for {api_prefix()} API")
     headers: Dict[str, str] = {"Authorization": f"Bearer {token()}"}
 
-    # Area sync
-    sync_url: str = f"https://{api_prefix()}-api.globalforestwatch.org/v2/area/sync"
+    # For some reason we are the only place calling this RW API to sync
+    # new subscriptions with areas. See GTC-2987 to fix this workflow.
+    yesterday = (datetime.now() - timedelta(1)).strftime("%Y-%m-%d")
+    sync_url: str = f"https://{api_prefix()}-api.globalforestwatch.org/v2/area/sync?startDate={yesterday}"
     sync_resp = requests.post(sync_url, headers=headers)
 
     if sync_resp.status_code != 200: