#337: don't create folders in case of dry-run, change input arg and u…

…pdate README

README.md

@@ -17,9 +17,14 @@ module can be extended with other necessary data management tools. Moving
 can be done with the `move_data` script, e.g.:
 
 ```
-move_data /path/to/files-from-git/mmd-xml-<env> /path/to/new/storage "/path/to/data/files/*.nc" --dmci-update
+move_data /path/to/files-from-git/mmd-xml-<env> /path/to/old/storage /path/to/new/storage "%Y/%m/%d/*.nc" --dmci-update
 ```
 
+The two last arguments provide a search pattern in case the netCDF files are
+stored in subfolders, and to directly updated the metadata catalog,
+respectively. If --dmci-update is not provided, local MMD files will not be
+pushed to the catalog.
+
 # Installation
 
 To avoid problems with conflicting versions, we recommend using the [Conda](

py_mmd_tools/mmd_operations.py

@@ -157,7 +157,7 @@ def move_data(mmd_repository_path, old_file_location_base, new_file_location_bas
     for nc_file in existing:
         # Error message
         emsg = ""
-        nfl = new_file_location(nc_file, new_file_location_base, old_file_location_base)
+        nfl = new_file_location(nc_file, new_file_location_base, old_file_location_base, dry_run)
         mmd_orig = get_local_mmd_git_path(nc_file, mmd_repository_path)
 
         # Check permissions before doing anything
@@ -216,13 +216,14 @@ def move_data(mmd_repository_path, old_file_location_base, new_file_location_bas
     return not_updated, updated
 
 
-def new_file_location(nc_file, new_base_loc, existing_base_loc):
+def new_file_location(nc_file, new_base_loc, existing_base_loc, dry_run):
     """Return the name of the new folder where the netcdf file will be
     stored. Subfolders of new_base_loc will be created.
     """
     if not os.path.isdir(new_base_loc):
         raise ValueError(f"Folder does not exist: {new_base_loc}")
     file_path = nc_file.replace(existing_base_loc, new_base_loc)
     new_folder = os.path.dirname(os.path.abspath(file_path))
-    os.makedirs(new_folder)
+    if not dry_run:
+        os.makedirs(new_folder)
     return new_folder

py_mmd_tools/script/move_data.py

@@ -35,11 +35,11 @@ def create_parser():
         "new_file_location_base", type=str,
         help="Base or exact path to the folder to which the data file(s) will be moved.")
     parser.add_argument(
-        "--ext_pattern", type=str, default=None,
+        "--ext-pattern", type=str, default=None,
         help="Pathname pattern extending old_file_location_base, i.e., extending the "
-             "existing file *base* location(s). Allows parsing date/times from a path "
-             "given a glob pattern intertwined with date/time format akin to "
-             "strptime/strftime format.")
+             "existing file *base* location(s) with, e.g, the year and month as a "
+             "glob pattern intertwined with date/time format akin to "
+             "strptime/strftime format (e.g., '%Y/%m').")
     parser.add_argument(
         '--dmci-update', action='store_true',
         help='Directly update the online catalog with the changed MMD files.'

tests/test_mmd_operations.py

@@ -251,12 +251,12 @@ def test_new_file_location(monkeypatch):
     existing_base_loc = "/some/old/loc"
 
     with pytest.raises(ValueError):
-        new_file_location(file, new_base, existing_base_loc)
+        new_file_location(file, new_base, existing_base_loc, True)
 
     with monkeypatch.context() as mp:
         mp.setattr("py_mmd_tools.mmd_operations.os.path.isdir", lambda *a, **k: True)
         mp.setattr("py_mmd_tools.mmd_operations.os.makedirs", lambda *a, **k: None)
-        assert new_file_location(file, new_base, existing_base_loc) == \
+        assert new_file_location(file, new_base, existing_base_loc, False) == \
             "/some/where/else/2024/06/19"