fix(file-based): Switch Excel parser from calamine to openpyxl engine

Switch the Excel parser engine from calamine to openpyxl to prevent
crashes when parsing Excel files with invalid date values.

The calamine engine (Rust-based) panics when encountering date values
that result in years outside Python's datetime range (1-9999), causing
the entire sync to fail. The openpyxl engine (pure Python) handles
these edge cases more gracefully, allowing syncs to complete even with
data quality issues.

This fixes crashes like:
  pyo3_runtime.PanicException: failed to construct date: PyErr {
    type: <class 'ValueError'>,
    value: ValueError('year 20225 is out of range')
  }

Trade-off: openpyxl is slower than calamine, but reliability is more
important than speed for production syncs.

Fixes: airbytehq/oncall#10097
Co-Authored-By: unknown <>

Author: devin-ai-integration[bot]

airbyte_cdk/sources/file_based/file_types/excel_parser.py

@@ -191,4 +191,4 @@ def open_and_parse_file(fp: Union[IOBase, str, Path]) -> pd.DataFrame:
         Returns:
             pd.DataFrame: Parsed data from the Excel file.
         """
-        return pd.ExcelFile(fp, engine="calamine").parse()  # type: ignore [arg-type, call-overload, no-any-return]
+        return pd.ExcelFile(fp, engine="openpyxl").parse()  # type: ignore [arg-type, call-overload, no-any-return]