Convert parquet timestamps directly to usable iso time format (#40)

* fix: fixed duplicated removal of first index / header

* convert parquet timestamps directly to usable iso timeformat instead of nanosecond timestamp object

* fix ruff linter

---------

Co-authored-by: Marvin Klerx <marvin.klerx@imbus.de>

Author: MarvKler

src/Tables/utils/file_reader.py

@@ -230,7 +230,20 @@ def read_excel(self, path: Path, **kwargs) -> DataFrame:
 
     def read_parquet(self, path: Path) -> DataFrame:
         """ """
-        return pd.read_parquet(path)
+        df: DataFrame = pd.read_parquet(path)
+
+        # try to transform to ISO timeformat -> if transformation fails, just return original parquet dataframe
+        try:
+            return self._parquet_transform_to_iso_timeformat(df)
+        except Exception:
+            return pd.read_parquet(path)
+
+    def _parquet_transform_to_iso_timeformat(self, df: DataFrame) -> DataFrame:
+        ts_cols = df.select_dtypes(include=["datetime64[ns]", "datetime64[ns, UTC]"]).columns
+        for col in ts_cols:
+            df[col] = df[col].dt.strftime("%Y-%m-%dT%H:%M:%S.%fZ")
+            df[col] = df[col].str[:-3] + "Z"
+        return df
 
     def read_table_file(self, path: Path) -> DataFrame:
         """

tests/atest/test_reader.robot

@@ -206,6 +206,13 @@ Read Parquet File - Without Header
     ${result} =    BuiltIn.Evaluate    "_time" not in "${content}"
     BuiltIn.Should Be True    ${result}
 
+Open Table - Parquet
+    Tables.Configure Ignore Header    False
+    ${alias} =    Tables.Open Table    ${CURDIR}${/}testdata${/}example_03.parquet
+    @{data} =    Tables.Get Table
+    Should Contain    ${data}[1][0]    2025
+    Tables.Count Table    ${alias}    Rows    ==    ${1001}
+
 Get Table Cell - Parquet
     Tables.Configure Ignore Header    False
     Tables.Open Table    ${CURDIR}${/}testdata${/}example_03.parquet