Robust and scalable content-and-structure indexing

Abstract Frequent queries on semi-structured hierarchical data are Content-and-Structure (CAS) that filter items based their location in the structure and value for some attribute. We propose Robust Scalable (RSCAS) index to efficiently answer CAS big data. To get an is robust against with varying selectivities, we introduce a novel dynamic interleaving merges path dimensions of composite keys balanced manner. store interleaved our trie-based RSCAS index, which supports wide range queries, including wildcards descendant axes. implement as log-structured merge tree scale it data-intensive applications high insertion rate. illustrate RSCAS’s robustness scalability by indexing from Software Heritage (SWH) archive, world’s largest, publicly available source code archive.