compare.py: add support for filtering fields (#250)

Small PR to add the option -f to be able to filter the fields one wants to check with giving a substr as argument. Only fields including the substrings are then checked and not all of them. Allows to also compare one field or some types of fields quickly which speeds up the workflow quite a lot (for me at least).

Author: elsagermann

AUTHORS

@@ -14,3 +14,4 @@ Carlos Osuna (cosunae), MeteoSwiss
 Willem Deconinck (wdeconinck), ECMWF
 Yannick Boetzel (yboetz), MeteoSwiss
 Rhea George (rheacangeo), Vulcan Inc.
+Elsa Germann (elsagermann), ETH Zurich (C2SM)

src/serialbox-python/compare/compare.py

@@ -307,7 +307,7 @@ def compare_fields(serializers, field, savepoint, dim_bounds):
     return False
 
 
-def compare(serializers, field_to_check, dim_bounds):
+def compare(serializers, field_to_check, dim_bounds, substr):
     """ Compare the data and info at every savepoint of field_1 to field_2 and returns
         True on success
     """
@@ -350,6 +350,10 @@ def get_time(start):
         else:
             fields_to_check = [field_to_check] if field_to_check in fields_at_savepoint else []
 
+        # Only check fields with name including substr if argument substr given
+        if substr != '':
+            fields_to_check = [field for field in fields_to_check if not field.find(substr)]
+
         savepoint_start_time = time()
         if fields_to_check:
             report(10 * "-", "Savepoint {}".format(savepoint))
@@ -449,6 +453,9 @@ def main(arguments=None):
     parser.add_argument("-q", "--info-only", dest="field_info_only", action="store_true",
                         help="only compare field meta-info (no data comparison) "
                              "(default: {})".format(get_config().FIELD_INFO_ONLY))
+    parser.add_argument("-f", "--filter-string", dest="substr", type=str,
+                        default='',
+                        help="only compare field names with SUBSTR in it")
     for dim in ["i", "j", "k", "l"]:
         parser.add_argument("-{}".format(dim), dest="{}".format(dim), metavar="START[:END]",
                             help="only compare the {} dimension 'START' or if 'END' is supplied "
@@ -492,7 +499,7 @@ def main(arguments=None):
     # Perform comparison
     ret = 1
     try:
-        ret = compare([serializer_1, serializer_2], field_1, dim_bounds)
+        ret = compare([serializer_1, serializer_2], field_1, dim_bounds, args.substr)
     except ser.SerialboxError as e:
         fatal_error(e)
     return ret