sandbox-test-run
sandbox-test-run.sh: perform a test run to exercise sandbox functions
Usage:
sandbox-test-run.sh [ -L ] [ -R ] SCRIPT
where SCRIPT is a script name (e.g., Scripts/HabEx-foobar.json).
By default, the SCRIPT is copied to a temporary file (in the above example, Scripts/HabEx-foobar-testrun.json) before running tests, so that the test runs land in sims/HabEx-foobar-testrun/... rather than a possibly-useful existing directory.
This means that SCRIPT can be an existing script name, and a test run will not pollute the corresponding sims/... files.
Options are:
-L: use the SCRIPT literally. This does *not* use the copied script-file
scheme above, so the new sims and plots (if they succeed)
will land in sims/SCRIPT/... among whatever might be already there.
-R: remove the simulation directory corresponding to the given SCRIPT, before
the test run, so the run is clean (except for EXOSIMS file caches).
This code prompts for confirmation of -R.
Note that using -R and -L together will delete the possibly-existing sims/SCRIPT directory, and all its contents.
Description:
The test sequence exercises several sandbox functions within the processing pipeline. This is approximately: - single test run (add-sims.sh ... =777) - ensemble of runs (add-sims.sh ... 100) and then various invocations of "make": - reduce data (make reduce) - basic plots (make html) - movies and final frames - webpage update (make html-only) - other ancillary graphics [] - final webpage update (make html-only) The intent is to shake out any errors in the infrastructure by exercising most of the mainline functions, all at once. Note that the ancillary graphics [] above can be fragile for some scripts, and they need a 3-year mission.
See also: python-version-check.py
turmon dec 2021