As we use newly introduced object lru cache to manage the condition cache entry in duckdb v1.5, we can no longer scan through the cache entries, instead, we need a non-evictable inverted index for efficient lookup.
but it is just a workaround for now, after we introduce the internal extension table for persistence entry, we can use a unify cache store and different cache key prefix to manage.
relate to: #23
As we use newly introduced object lru cache to manage the condition cache entry in duckdb v1.5, we can no longer scan through the cache entries, instead, we need a non-evictable inverted index for efficient lookup.
but it is just a workaround for now, after we introduce the internal extension table for persistence entry, we can use a unify cache store and different cache key prefix to manage.
relate to: #23