file_min_digits: 5->6

[ax3l]

100k+ step runs are quite common in WarpX. To simplify post-processing scripts, increase the default to pad to 6 digits.

This might break some hand-written scripts that use ????? wild-cards over * wildcards in regex and thus need to be updated. But it at the same time simplifies regexes for analysis of data series and listing of such file series.

Notes:

• contains work-arounds for restart tests: do not hard-code 5 digits AMReX-Codes/regression_testing#119 (ok for now)
• skipped: for plotfiles, let's also take the chance to replace <diag_name><step> with <diag_name>_<step>. Reason: no regex can differentiate diag10004 from step 4 or 10'004 ? -> another time, too much to change yet again

[ax3l]

Found a couple of scripts that need updates:

Regression/Checksum/checksumAPI.py
83:    plotfile_list = glob.glob(path_to_all_plotfiles + '*_plt?????',

Docs/source/developers/checksum.rst
26:This can also be included in an existing analysis script. Note that the plotfile must be ``<test name>_plt?????``, as is generated by the CI framework.

Docs/source/dataanalysis/plot_parallel.rst
27:will loop through plotfiles named ``plt?????`` (e.g., ``plt00000``, ``plt00100`` etc.)

Examples/Tests/collision/analysis_collision_3d.py
55:fn_list = glob(last_fn[:-5] + "?????")

Examples/Tests/collision/analysis_collision_2d.py
55:fn_list = glob(last_fn[:-5] + "?????")

Tools/DevUtils/compare_wx_w_3d.ipynb
57:    "file_list_warpx = glob.glob(path_warpx + 'plt?????')\n",
71:    "file_list_warp = glob.glob(path_warp + 'diags/hdf5/data????????.h5')\n",

Tools/PostProcessing/video_yt.py
36:file_list = glob.glob('./diags/plotfiles/plt?????')

Tools/PostProcessing/plot_parallel.py
47:                    help='path to plotfiles, defaults to diags/plotfiles. Plotfiles names must be plt?????')
236:file_list = glob.glob(os.path.join(path, 'plt?????'))

Tools/PostProcessing/yt3d_mpi.py
39:file_list = glob.glob('plotfiles/plt?????')

Tools/PostProcessing/plot_nci_growth_rate.ipynb
58:    "file_list_warpx = glob.glob(path_wx + 'diag1?????')\n",

Tools/PostProcessing/plot_particle_path.py
131:    fn_list = glob.glob("plt?????")

Tools/LibEnsemble/read_sim_output.py
87:    file_list = glob.glob('diags/plotfiles/plt?????')

[ax3l]

The replacement for glob can potentially be as easy as this:

glob('./diags/plotfiles/plt[0-9]+')

Simple regex that is matching trailing digits if we introduce _ in plotfiles: .*_(\d+)$

[EZoni]

To fix the tests uniform_plasma_restart, restart, restart_psatd and restart_psatd_time_avg, we will need to change 5 to 6 in the following line of the Python file regtest.py within the regression_testing repository (https://github.com/ECP-WarpX/regression_testing/blob/9aab727cd00dc4527ff2cc6fdd5772c9ebb9ffab/regtest.py#L736-L737):

--- a/regtest.py
+++ b/regtest.py
@@ -734,7 +734,7 @@ def test_suite(argv):
                 shutil.move(test.diffDir, orig_diff_dir)
 
             # get the file number to restart from
-            restart_file = "%s_chk%5.5d" % (test.name, test.restartFileNum)
+            restart_file = "%s_chk%6.6d" % (test.name, test.restartFileNum)

[ax3l]

regression_testing

I for now just overwrote the 4 checkpoint-restart tests we have to use 5 digits.

[dpgrote]

This looks good to me.

[EZoni]

Awesome, looks great to me!