In recent years, the volume of generated RDF graphs streams from different fields of applications is very large and therefore difficult to process in an optimized manner. Indeed, processing such data in conventional triplestores can be costly in terms of execution time and memory consumption. Several works have examined data compression approach both on static and dynamic RDF data. In addition to those based on stored RDF data, two recent compression algorithms RDSZ and ERI were focused on RDF streams. Continuous compressed format requires less memory space but cannot be exploited through SPARQL queries. In this paper, we propose an approach for continuous querying RDSZ-based RDF streams without decompression phase. We add three algorithms from simple to aggregate query execution over RDSZ compressed items. Our experimentation use real datasets to demonstrate the effectiveness and efficiency of our proposition in term of query execution time and memory save.