NAME edu.berkeley.cs.db.yfilter.filter.EXfilter - Processes queries on a set of documents, and outputs the query results and the processing performance. SYNOPSIS java edu.berkeley.cs.db.yfilter.filter.EXfilter (DOC_FILE | DOC_DIR) QUERY_FILE [options] PARAMETERS DOC_FILE | DOC_DIR XML document or Directory containing a set of XML documents. QUERY_FILE File containing the queries. OPTIONS --num_docs=NUM Number of documents to be processed. --num_queries=NUM Number of queries to be processed. --profile Enables profiling. NOTE: The profiling results are output to "profile.txt" in the current working directory. --result=LEVEL Result level. LEVEL can be one of the following: NONE - no output. BOOLEAN - if the query is matched (TRUE/FALSE). ONE - one matching element. ALL - all matching elements. NOTE: The default setting for benchmarking is BOOLEAN. The query results are output to "result.txt" in the current working directory, if the --output option is not set to "indiv". --output=indiv Output query results to individual files. NOTE: Once output is set to "indiv", the query results are output to multiple files in the working directory, one per XML document, in order of document execution. --nfa_opt=LEVEL The NFA execution level. LEVEL can be one of the following: 2 -- for NFA execution and result collection 1 -- for NFA execution and simplified result collection if queries don't contain predicates 0 -- for NFA execution only NOTE: The NFA execution level partially depends on the result level. nfa_opt can be set to 1, only if the result level is NONE, BOOLEAN, or ONE. nfa_opt can be set to 0, only if the result level is NONE. The default setting for benchmarking is: result = BOOLEAN, nfa_opt = 1. --msg_preparse=FLAG Preparse messages. FLAG can be TRUE or FALSE. --warmup=VALUE Warm-up the filter. VALUE can be one of the following two cases: 1) an XML document file used for warming up or, 2) a positive integer, indicating the number of documents to be used for warming up if the input is an XML directory. DIAGNOSTICS BUGS