updated the doc string

kerchunk/grib2.py

@@ -3,10 +3,12 @@
 import io
 import logging
 from collections import defaultdict
-from typing import Iterable, List, Dict, Set, TYPE_CHECKING
-
+from typing import Iterable, List, Dict, Set, TYPE_CHECKING, Optional
 import ujson
 
+if TYPE_CHECKING:
+    import pandas as pd
+
 try:
     import cfgrib
 except ModuleNotFoundError as err:  # pragma: no cover
@@ -584,23 +586,20 @@ def correct_hrrr_subhf_step(group: Dict) -> Dict:
     return group
 
 
-if TYPE_CHECKING:
-    import pandas as pd
-
-
 def parse_grib_idx(
     fs: fsspec.AbstractFileSystem,
     *,
     basename: str,
     suffix: str = "idx",
+    tstamp: Optional["pd.Timestamp"] = None,
     validate: bool = False,
 ) -> "pd.DataFrame":
     """
     Standalone method used to extract metadata from a grib2 idx file(text) from NODD.
 
-    The function takes idx file, extracts the metadata especially the attrs (variables)
-    from each idx entry and converts it into pandas DataFrame. The dataframe is later
-    to build the one-to-one mapping to the grib file.
+    The function takes idx file, extracts the metadata known as attrs (variables with
+    level and forecast time) from each idx entry and converts it into pandas
+    DataFrame. The dataframe is later to build the one-to-one mapping to the grib file metadata.
 
     Parameters
     ----------
@@ -610,6 +609,8 @@ def parse_grib_idx(
         The base name is the full path to the grib file.
     suffix : str
         The suffix is the ending for the idx file.
+    tstamp : Optional[pd.Timestamp]
+        The timestamp to use for when the data was indexed
     validate : bool
         The validation if the metadata table has duplicate attrs.
 
@@ -631,7 +632,7 @@ def parse_grib_idx(
         result["attrs"] = (
             result["attrs"] + ":" + result["level"] + ":" + result["forecast"]
         )
-        result = result.drop(columns=["level", "forecast"])
+        result.drop(columns=["level", "forecast"], inplace=True)
     except Exception as e:
         raise ValueError(f"Could not parse {fname}") from e
 
@@ -641,7 +642,7 @@ def parse_grib_idx(
         ),
         idx_uri=fname,
         grib_uri=basename,
-        indexed_at=pd.Timestamp.now(),
+        indexed_at=tstamp if tstamp else pd.Timestamp.now(),
     )
 
     if validate and not result["attrs"].is_unique: