Skip to content

cs-au-dk/tajs_vr_experiments

Repository files navigation

Building TAJS-VR-Experiments

Make sure you clone not only the TAJS-VR-Experiments repository but also the submodules, for example by running

git submodule update --init --recursive

The simplest way to build TAJS-VR is to run Ant:

ant

This will build two jar files: dist/tajsvr.jar (contains only TAJS-VR itself) and dist/tajsvr-all.jar (includes the relevant extra libraries).

Running TAJS-VR-Experiments

The analysis can be run as follows:

java -jar dist/tajsvr-all.jar [EXPERIMENT] [ARGUMENTS]

where [EXPERIMENT] can be one of:

  • micro-benchmarks
    • Description: Runs micro-benchmarks with TAJS and TAJS-VR.
    • [ARGUMENTS]: Takes no arguments.
    • Output: Table similar to table 1 in the paper.
  • library-benchmarks
    • Description: Runs library-benchmarks with TAJS-VR.
    • [ARGUMENTS]: Takes a single argument, which specifies a comma-separated of suites. Valid test-suites are jquery, jsai, prototype, scriptaculous, underscore, lodash3 and lodash4. To run all test-suites, the argument ALL can be used instead.
    • Output: Table similar to table 2, table 3 and the appendix tables for the specified suites.
  • single-test
    • Description: Runs TAJS-VR on a specified test.
    • [ARGUMENTS]: arguments are directly passed to TAJS, so see the TAJS readme for details.
    • Output: Same output as running TAJS without value refinement

Reproducing results from paper

Note that the paper has run experiments with a JVM with 10GB RAM, so we use the Java option -Xmx10G to enable this. Reproduce results from Table 1 in the paper (expected time is few seconds):

java -Xmx10G -jar dist/tajsvr-all.jar micro-benchmarks

Reproduce the rest of the tables (expected time is around 8 hours):

java -Xmx10G -jar dist/tajsvr-all.jar library-benchmarks ALL

Since recalculating all results takes a long time, smaller versions of the tables can be produced, by only running some of the library test-suites.

java -Xmx10G -jar dist/tajsvr-all.jar library-benchmarks scriptaculous,underscore

Expected running time for each test-suite:

  • jquery: 1 hour and 14 minutes
  • jsai: 29 minutes
  • prototype: 15 minutes
  • scriptaculous: 5 minutes
  • underscore: 13 minutes
  • lodash3: 44 minutes
  • lodash4: 4 hours and 25 minutes

About

No description, website, or topics provided.

Resources

Stars

Watchers

Forks

Releases

No releases published

Packages

No packages published