optimize BackupFinalizer.finalize_files_and_backup blob querying

Author: Fallen-Breath

prime_backup/action/helpers/backup_finalizer.py

@@ -18,16 +18,16 @@ def __init__(self, session: DbSession):
 	def finalize_files_and_backup(self, backup: schema.Backup, files: List[schema.File]):
 		self.session.flush()  # ensure all blobs has their blob.id allocated
 
-		file_blobs = self.session.get_blobs_by_hashes_opt([
+		hash_to_blob_id = self.session.get_blob_ids_by_hashes_opt([
 			file.blob_hash for file in files
 			if file.blob_hash is not None
 		])
 		for file in files:
 			if file.blob_hash is not None:
-				file_blob = file_blobs[file.blob_hash]
-				if file_blob is None:
+				blob_id = hash_to_blob_id[file.blob_hash]
+				if blob_id is None:
 					raise AssertionError('blob of file does not exists: {}'.format(file))
-				file.blob_id = file_blob.id
+				file.blob_id = blob_id
 
 		allocate_args = FilesetAllocateArgs.from_config(self.config)
 		allocate_result = FilesetAllocator(self.session, files).allocate(allocate_args)

prime_backup/db/session.py

@@ -347,6 +347,19 @@ def get_blobs_by_hashes_opt(self, hashes: List[str]) -> Dict[str, Optional[schem
 				result[blob.hash] = blob
 		return result
 
+	def get_blob_ids_by_hashes_opt(self, hashes: List[str]) -> Dict[str, Optional[int]]:
+		"""
+		:return: a dict, hash -> optional blob id. All given hashes are in the dict
+		"""
+		result: Dict[str, Optional[int]] = {h: None for h in hashes}
+		for view in collection_utils.slicing_iterate(hashes, self.__safe_var_limit):
+			for blob_hash, blob_id in self.session.execute(
+					select(schema.Blob.hash, schema.Blob.id).
+					where(schema.Blob.hash.in_(view))
+			).all():
+				result[blob_hash] = blob_id
+		return result
+
 	def get_blobs_by_hashes(self, hashes: List[str]) -> Dict[str, schema.Blob]:
 		"""
 		:return: a dict, hash -> blob. All given hashes are in the dict