In this Repo, XXXX-2-BM corresponds to MutBM and XXXX-2 corresponds to AnonM

View benchmarks used

• We present results for 3 benchmarks: XData-BM, LargeJoin, TPCH.
• /benchamrks/<benchmark-name> directory contains following files.
  • Schema file (.sql). (eg. DDL.sql)
  • Queries file (.txt). Format to write queries is "queryID|queryType|query"
  • Mutants file (.txt). Format to write queries is "queryID|mutantType|mutant". Note mutants corresponding to each queryID are given.

View precomputed results for benchmarks

• Precomputed <benchmark>-mutant-result.xslx files are present in /results folder

Set Up - XData

Note: Currently supported for Linux and Windows

• Clone this repo using git clone command

  • git clone https://gitlab.com/xdata/xdatapublic.git

• Install Docker to setup the container.

  • Follow the steps here: https://docs.docker.com/engine/install/

• To build container for the first time

  • Run chmod +x xdata.sh then
  • Run ./xdata.sh build (run with sudo if user does not have permission)
  • This will take approximately 5 mins

• To run the container

  • Run ./xdata.sh run (run with sudo if user does not have permission)

Run XData

• Run source vrtuenv/bin/activate to start virtual env

• Run python3 execute_tests.py <benchmark> | tee <benchmark>-datagen.log

  • <benchmark> options:
    • XData-BM: Runpython3 execute_tests.py XData-BM | tee XData-BM-datagen.log
      • avg time to run 30 mins
    • LargeJoin: Run python3 execute_tests.py LargeJoin | tee LargeJoin-datagen.log
      • avg time to run 5 mins
    • TPCH: Run python3 execute_tests.py TPCH | tee TPCH-datagen.log
      • avg time to run 10 mins
    • Other: Run python3 execute_tests.py Other | tee Other-datagen.log
  • First three benchmarks to be used to replicate results
  • To run XData on any other set of query mutant pairs
    • Modify `benchmarks/Other/DDL.sql' with postgreSQL compatible schema
    • Modify `benchmarks/Other/queries.txt'
    • Modify `benchmarks/Other/mutants.txt'
    • Then rebuild the setup and run execute_tests.py

• To run xdata three input files are required. Place all the three files in benchmarks folder

  • Schema file (.sql). (eg. DDL.sql)
  • Queries file (.txt). Format to write queries is "queryID|queryType|query"
  • Mutants file (.txt). Format to write queries is "queryID|mutantType|mutant". Note mutants corresponding to each queryID are given.
  • This version does not require sample data as input

• Run python3 execute_query.py to run a single query. The schema and query paths are to be mentioned at query_config.cfg

Reading Output after running XData

• /home/xdata/output/<benchmark>-mutant-results.out file contains output state(Killed/NotKilled/Unsupported) for each mutant-query pair.
  • Note: Precomputed .xslx files are present in /results folder (outside of the container)
• Directories /home/xdata/tmp/temp_smt/<query_id> contains the datasets (.sql files) which are generated for each original query