More ES robustness

whiletrue-industries · Feb 5, 2025 · 6ce70e9 · 6ce70e9
1 parent 6f4d88b
commit 6ce70e9
Show file tree

Hide file tree

Showing 2 changed files with 6 additions and 2 deletions.
diff --git a/odds/backend/scanner/worldbank/worldbank_catalog_scanner.py b/odds/backend/scanner/worldbank/worldbank_catalog_scanner.py
@@ -83,6 +83,7 @@ async def scan(self) -> AsyncIterator[Dataset]:
                     description='\n'.join(filter(None, [identification.get('title'), identification.get('subtitle'), identification.get('description')])),
                     publisher=identification.get('citation'),
                     publisher_description='',
-                    resources=resources
+                    resources=resources,
+                    link=f'https://datacatalog.worldbank.org/search/dataset/{dataset_unique_id}'
                 )
                 yield dataset
diff --git a/odds/common/qa_repo/es/es_qa_repo.py b/odds/common/qa_repo/es/es_qa_repo.py
@@ -67,7 +67,10 @@ async def storeQA(self, question: str, answer: str, success: bool, score: int, d
                 deployment_id=deployment_id,
                 embeddings=(await embedder.embed(question)).tolist()
             )
-            await client.create(index=ES_INDEX, id=id, body=body)
+            try:
+                await client.create(index=ES_INDEX, id=id, body=body)
+            except elasticsearch.ConflictError:
+                await client.update(index=ES_INDEX, id=id, body={'doc': body})
             return self.toQa(body)
 
     async def getQuestion(self, *, question: str=None, id: str=None, deployment_id: str=None) -> Optional[QA]: