Maybe I'm missing something, but if you only want to find documents with the exact path, then why use the path_hierarchy tokenizer? If you create the index with path mapped as type keyword (without an analyzer) instead, you will only find documents that match the exact path.