flowtask 5.8.4__cp310-cp310-manylinux2014_x86_64.manylinux_2_17_x86_64.whl

flowtask/components/OpenWithPandas.py

@@ -0,0 +1,715 @@
+from typing import Any
+from pathlib import PurePath
+from io import BytesIO
+import aiofiles
+from xml.sax import parse
+import warnings
+import pandas
+from pandas._libs.parsers import STR_NA_VALUES
+import orjson
+import xlrd
+import numpy as np
+from ..utils import check_empty
+from ..exceptions import ComponentError, DataNotFound, EmptyFile
+from .OpenWithBase import OpenWithBase, detect_encoding, excel_based, ExcelHandler
+
+
+# Suppress specific warning
+warnings.filterwarnings("ignore", category=UserWarning)
+
+class OpenWithPandas(OpenWithBase):
+    """
+    OpenWithPandas
+
+    Overview
+
+            Open a file and return a Dataframe type
+
+    .. table:: Properties
+       :widths: auto
+
+
+    +-------------+----------+-----------+-------------------------------------------------------+
+    | Name        | Required | Summary                                                           |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    |  model      |   Yes    | A model (json) representative of the data that I am going to      |
+    |             |          | open * name of a DataModel (in-development)                       |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    |  map        |   Yes    | Map the columns against the model                                 |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    | tablename   |   Yes    | Join the data from the table in the postgres database             |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    | use_map     |   Yes    | If true, then a MAP file is used instead of a table in postgresql |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    | file_engine |   Yes    | Pandas different types of engines for different types of Excel    |
+    |             |          | * xlrd (legacy, xls type)                                         |
+    |             |          | * openpyxl (new xlsx files)                                       |
+    |             |          | * pyxlsb (to open with macros and functions)                      |
+    +-------------+----------+-----------+-------------------------------------------------------+
+    |  dtypes     |   No     | force the data type of a column ex: { order_date: datetime }      |
+    +-------------+----------+-----------+-------------------------------------------------------+
+
+
+    Return the list of arbitrary days
+    
+
+        Example:
+
+        ```yaml
+        OpenWithPandas:
+          mime: text/csv
+          process: true
+          separator: '|'
+          drop_empty: true
+          trim: true
+          pk:
+            columns:
+            - associate_oid
+            - associate_id
+            append: false
+            verify_integrity: true
+          map:
+            tablename: employees
+            schema: bacardi
+            map: employees
+            replace: false
+        ```
+
+    """
+    """
+    OpenWithPandas
+
+    Overview
+
+        This component opens various file types (CSV, Excel, HTML, JSON) into Pandas DataFrames.
+
+    .. table:: Properties
+    :widths: auto
+
+
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | Name                   | Required | Summary                                                                   |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | directory              |   No     | The directory where the files are located.                                |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | filename               |   No     | The name of the file to open.                                             |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | file                   |   No     | Pattern or file to open.                                                  |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | mime                   |   No     | The MIME type of the file. Default is "text/csv".                         |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | separator              |   No     | Separator for CSV files. Default is ",".                                  |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | force_map              |   No     | Force the use of a map file. Default is False.                            |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | parse_dates            |   No     | Columns to parse as dates. Default is an empty dictionary.                |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | filter_nan             |   No     | Filter out NaN values. Default is True.                                   |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | na_values              |   No     | List of values to recognize as NaN. Default is ["NULL", "TBD"].           |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | remove_empty_strings   |   No     | Remove empty strings. Default is True.                                    |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | no_multi               |   No     | Disable multi-file processing. Default is False.                          |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | clean_nat              |   No     | Clean NaT values. Default is False.                                       |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | flavor                 |   No     | The flavor of the database for column information. Default is "postgres". |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | pd_args                |   No     | Additional arguments for pandas. Default is an empty dictionary.          |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    |  model                 |   Yes    | A model (json) representative of the data that I am going to              |
+    |                        |          | open * name of a DataModel (in-development)                               |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    |  map                   |   Yes    | Map the columns against the model                                         |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | tablename              |   Yes    | Join the data from the table in the postgres database                     |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | use_map                |   Yes    | If true, then a MAP file is used instead of a table in postgresql         |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    | file_engine            |   Yes    | Pandas different types of engines for different types of Excel            |
+    |                        |          | * xlrd (legacy, xls type)                                                 |
+    |                        |          | * openpyxl (new xlsx files)                                               |
+    |                        |          | * pyxlsb (to open with macros and functions)                              |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+    |  dtypes                |   No     | force the data type of a column ex: { order_date: datetime }              |
+    +------------------------+----------+-----------+---------------------------------------------------------------+
+
+    Returns
+
+    This component returns a Pandas DataFrame containing the data from the opened file(s).
+
+    """
+    def get_column_headers(self):
+        headers = []
+        for filename in self._filenames:
+            try:
+                encoding = self.check_encoding(filename)
+            except Exception:
+                encoding = "UTF-8"
+            df = pandas.read_csv(
+                filename,
+                sep=self.separator,
+                skipinitialspace=True,
+                encoding=encoding,
+                engine="python",
+                nrows=1,
+            )
+            headers.append(df.columns.values.tolist())
+        return headers
+
+    def set_datatypes(self):
+        dtypes = {}
+        for field, dtype in self.datatypes.items():
+            if dtype == "uint8":
+                dtypes[field] = np.uint8
+            elif dtype == "uint16":
+                dtypes[field] = np.uint16
+            elif dtype == "uint32":
+                dtypes[field] = np.uint32
+            elif dtype == "int8":
+                dtypes[field] = np.int8
+            elif dtype == "int16":
+                dtypes[field] = np.int16
+            elif dtype == "int32":
+                dtypes[field] = np.int32
+            elif dtype == "float":
+                dtypes[field] = float
+            elif dtype == "float32":
+                dtypes[field] = float
+            elif dtype in ("string", "varchar", "str"):
+                dtypes[field] = "str"
+            elif dtype == "object":
+                dtypes[field] = object
+            else:
+                # invalid datatype
+                raise ComponentError(
+                    f"Invalid DataType value: {field} for field {dtype}"
+                )
+        if dtypes:
+            self.args["dtype"] = dtypes
+
+    async def open_excel(
+        self, filename: str, add_columns: dict, encoding
+    ) -> pandas.DataFrame:
+        self._logger.debug(
+            f"Opening Excel file {filename} with Pandas, encoding: {encoding}"
+        )
+        if self.mime == "text/xml":
+            xmlparser = ExcelHandler()
+            parse(filename, xmlparser)
+            if hasattr(self, "skiprows"):
+                row = self.skiprows
+                columns = self.skiprows + 1
+                start = columns + 1
+            else:
+                row = 0
+                columns = 0
+                start = columns + 1
+            try:
+                if (
+                    hasattr(self, "add_columns") and hasattr(self, "rename")
+                    and self.rename is True
+                ):
+                    cols = add_columns
+                else:
+                    cols = xmlparser.tables[0][columns]
+                df = pandas.DataFrame(data=xmlparser.tables[0][start:], columns=cols)
+                return df
+            except pandas.errors.EmptyDataError as err:
+                raise EmptyFile(f"Empty File {filename}: {err}") from err
+            except pandas.errors.ParserError as err:
+                raise ComponentError(f"Parsing File {filename}: {err}") from err
+            except Exception as err:
+                raise ComponentError(
+                    f"Generic Error on file {filename}, error: {err}"
+                ) from err
+        else:
+            if (
+                self.mime == "application/vnd.openxmlformats-officedocument.spreadsheetml.sheet"
+            ):
+                # xlsx or any openxml based document
+                file_engine = self._params.get("file_engine", "openpyxl")
+            elif self.mime == "application/vnd.ms-excel.sheet.binary.macroEnabled.12":
+                file_engine = self._params.get("file_engine", "pyxlsb")
+            else:
+                try:
+                    ext = filename.suffix
+                except (AttributeError, ValueError) as e:
+                    print(f"Error detecting extension: {e}")
+                    ext = ".xls"
+                if ext == ".xls":
+                    file_engine = self._params.get("file_engine", "xlrd")
+                else:
+                    file_engine = self._params.get("file_engine", "calamine")
+            try:
+                arguments = {**self.args, **add_columns, **self.parse_dates}
+                if self._limit is not None and isinstance(self._limit, int):
+                    arguments["nrows"] = self._limit
+                if self.sheet_name is not None:
+                    arguments["sheet_name"] = self.sheet_name
+                # TODO: if sheet_name is None, then open all worksheets
+                # work with the dictionary of dataframes.
+                df = pandas.read_excel(
+                    filename,
+                    na_values=self.na_values,
+                    na_filter=self.filter_nan,
+                    engine=file_engine,
+                    keep_default_na=False,
+                    **arguments,
+                )
+                return df
+            except (IndexError, xlrd.biffh.XLRDError) as err:
+                raise ComponentError(
+                    f"Excel Index error on File {filename}: {err}"
+                ) from err
+            except pandas.errors.EmptyDataError as err:
+                raise EmptyFile(f"Empty File {filename}: {err}") from err
+            except pandas.errors.ParserError as err:
+                raise ComponentError(f"Error Parsing File {filename}: {err}") from err
+            except Exception as err:
+                raise ComponentError(
+                    f"Generic Error on file {filename}, error: {err}"
+                ) from err
+
+    async def open_html(
+        self, filename: str, add_columns: dict, encoding: str
+    ) -> pandas.DataFrame:
+        self._logger.debug(
+            f"Opening an HTML file {filename} with Pandas, encoding={encoding}"
+        )
+        if "dtype" in self.args:
+            del self.args["dtype"]
+        if "skiprows" in self.args:
+            del self.args["skiprows"]
+        try:
+            dfs = pandas.read_html(
+                filename,
+                keep_default_na=False,
+                flavor="html5lib",
+                na_values=self.na_values,
+                encoding=encoding,
+                **self.parse_dates,
+                **self.args,
+            )
+            if dfs:
+                df = dfs[0]
+            else:
+                df = None
+            if "names" in add_columns:
+                df.columns = add_columns["names"]
+            return df
+        except pandas.errors.EmptyDataError as err:
+            raise EmptyFile(message=f"Empty File {filename}: {err}") from err
+        except pandas.errors.ParserError as err:
+            raise ComponentError(message=f"Parsing File {filename}: {err}") from err
+        except Exception as err:
+            raise ComponentError(
+                message=f"Generic Error on file {filename}: {err}"
+            ) from err
+
+    async def open_parquet(
+        self, filename: str, add_columns: dict, encoding
+    ) -> pandas.DataFrame:
+        pass
+
+    async def open_sql(
+        self, filename: str, add_columns: dict, encoding
+    ) -> pandas.DataFrame:
+        pass
+
+    async def open_json(
+        self, filename: str, add_columns: dict, encoding: str
+    ) -> pandas.DataFrame:
+        self._logger.debug(
+            f"Opening a JSON file {filename} with Pandas, encoding={encoding}"
+        )
+        # TODO: add columns functionality.
+        try:
+            df = pandas.read_json(
+                filename, orient="records", encoding=encoding, **self.args
+            )
+            return df
+        except pandas.errors.EmptyDataError as err:
+            raise EmptyFile(message=f"Empty File {filename}: {err}") from err
+        except pandas.errors.ParserError as err:
+            raise ComponentError(
+                message=f"Error Parsing File {filename}: {err}"
+            ) from err
+        except Exception as err:
+            raise ComponentError(
+                message=f"Generic Error on file {filename}: {err}"
+            ) from err
+
+    async def open_csv(
+        self, filename: str, add_columns: dict, encoding
+    ) -> pandas.DataFrame:
+        self._logger.debug(
+            f"Opening CSV file {filename} with Pandas, encoding={encoding}"
+        )
+        try:
+            add_columns["low_memory"] = False
+            add_columns["float_precision"] = "high"
+        except KeyError:
+            pass
+        try:
+            # can we use pyarrow.
+            engine = self.args["engine"]
+            del self.args["engine"]
+        except KeyError:
+            engine = "c"
+        if self._limit is not None and isinstance(self._limit, int):
+            add_columns["nrows"] = self._limit
+        # try to fix the encoding problem on files:
+        _, new_encoding = detect_encoding(filename, encoding)
+        if new_encoding != encoding:
+            self._logger.warning(
+                f"Encoding on file: {new_encoding} and \
+                declared by Task ({encoding}) are different"
+            )
+            # encoding = new_encoding
+        # open file:
+        if hasattr(self, "bigfile"):
+            try:
+                tp = pandas.read_csv(
+                    filename,
+                    sep=self.separator,
+                    decimal=",",
+                    engine=engine,
+                    keep_default_na=False,
+                    na_values=self.na_values,
+                    na_filter=self.filter_nan,
+                    encoding=encoding,
+                    skipinitialspace=True,
+                    iterator=True,
+                    chunksize=int(self.chunksize),
+                    **add_columns,
+                    **self.parse_dates,
+                    **self.args,
+                )
+                return pandas.concat(tp, ignore_index=True)
+            except pandas.errors.EmptyDataError as err:
+                raise ComponentError(
+                    f"Empty Data File on: {filename}, error: {err}"
+                ) from err
+            except Exception as err:
+                raise ComponentError(
+                    f"Generic Error on file: {filename}, error: {err}"
+                ) from err
+        else:
+            try:
+                return pandas.read_csv(
+                    filename,
+                    sep=self.separator,
+                    quotechar='"',
+                    decimal=",",
+                    engine=engine,
+                    keep_default_na=False,
+                    na_values=self.na_values,
+                    na_filter=self.filter_nan,
+                    encoding=encoding,
+                    skipinitialspace=True,
+                    **add_columns,
+                    **self.parse_dates,
+                    **self.args,
+                )
+            except UnicodeDecodeError as exc:
+                self._logger.warning(
+                    f"Invalid Encoding {encoding}: {exc}"
+                )
+                # fallback to a default unicode:
+                _, encoding = detect_encoding(filename, encoding)
+                self._logger.debug(f"Detected Encoding > {encoding!s}")
+                last_encoding = None
+                fname = filename
+                if hasattr(self, 'clean_null_bytes'):
+                    async with aiofiles.open(filename, 'rb') as file:
+                        # Removing all null bytes
+                        content = await file.read()
+                        content = content.replace(b'\x00', b'')
+                    fname = BytesIO(content)
+                for enc in ('utf-8', 'latin1', 'ascii'):
+                    last_encoding = enc
+                    try:
+                        return pandas.read_csv(
+                            fname,
+                            sep=self.separator,
+                            quotechar='"',
+                            decimal=",",
+                            engine=engine,
+                            keep_default_na=False,
+                            na_values=self.na_values,
+                            na_filter=self.filter_nan,
+                            encoding=enc,
+                            skipinitialspace=True,
+                            on_bad_lines='warn',
+                            **add_columns,
+                            **self.parse_dates,
+                            **self.args,
+                        )
+                    except Exception as e:
+                        print(e)
+                        continue
+                else:
+                    # No encoding match
+                    raise ComponentError(
+                        f"Cannot Open the file with encoding {last_encoding}"
+                    )
+            except ValueError as exc:
+                # Open Pandas with default settings for detect discrepancies
+                df = pandas.read_csv(
+                    filename,
+                    sep=self.separator,
+                    quotechar='"',
+                    decimal=",",
+                    engine=engine,
+                    encoding=encoding,
+                    dtype=str,
+                    header=None,
+                )
+                # columns in Pandas:
+                num_cols = int(df.shape[1])
+                expected = len(add_columns.get('names', []))
+                if expected > 0 and num_cols - expected > 0:
+                    # some extra columns were found:
+                    raise ComponentError(
+                        (
+                            f"There are more columns in FILE than expected. "
+                            f"There are {num_cols} in File received vs "
+                            f"{expected} columns in Mapping definition."
+                        )
+                    )
+                try:
+                    del self.args['dtype']
+                except KeyError:
+                    pass
+                self._logger.error(
+                    (
+                        f"Some columns have wrong type in Model: {exc}, "
+                        "Opening file with default settings (str)"
+                    )
+                )
+                try:
+                    return pandas.read_csv(
+                        filename,
+                        sep=self.separator,
+                        quotechar='"',
+                        decimal=",",
+                        engine=engine,
+                        keep_default_na=False,
+                        na_values=self.na_values,
+                        na_filter=self.filter_nan,
+                        encoding=encoding,
+                        **add_columns,
+                        **self.parse_dates,
+                        **self.args,
+                    )
+                except Exception as ex:
+                    raise ComponentError(
+                        f"Invalid types of columns found on file {filename}, {ex}"
+                    )
+            except pandas.errors.EmptyDataError as err:
+                raise ComponentError(
+                    f"Empty Data in file: {filename}, error: {err}"
+                ) from err
+            except pandas.errors.ParserError as err:
+                raise ComponentError(
+                    f"Error parsing File: {filename}, error: {err}"
+                ) from err
+            except Exception as err:
+                raise ComponentError(
+                    f"Generic Error on file: {filename}, error: {err}"
+                ) from err
+
+    async def run(self) -> Any:
+        await super(OpenWithPandas, self).run()
+        add_columns = await self.colinfo()
+        result = []
+        df = None
+        ## Define NA Values:
+        default_missing = STR_NA_VALUES.copy()
+        if self.remove_empty_strings is True:
+            try:
+                default_missing.remove("")
+            except KeyError:
+                pass
+        for val in self.na_values:  # pylint: disable=E0203
+            default_missing.add(val)
+            default_missing.add(val)
+        self.na_values = default_missing
+        if self._filenames is None and not check_empty(self._data):
+            if isinstance(self._data, list):
+                for file in self._data:
+                    try:
+                        df = pandas.DataFrame(
+                            data=file, **add_columns, **self.parse_dates, **self.args
+                        )
+                        result.append(df)
+                    except pandas.errors.EmptyDataError as err:
+                        raise ComponentError(
+                            f"Error on Empty Data: error: {err}"
+                        ) from err
+                    except ValueError as err:
+                        raise ComponentError(
+                            f"Error parsing Data: error: {err}"
+                        ) from err
+                    except Exception as err:
+                        raise ComponentError(
+                            f"Generic Error on Data: error: {err}"
+                        ) from err
+            if df is None or df.empty:
+                raise DataNotFound("Dataframe is Empty: Data not found")
+        else:
+            # itereate over all files or data
+            self._variables["FILENAMES"] = self._filenames
+            for filename in self._filenames:
+                try:
+                    encoding = self.check_encoding(filename)
+                except Exception:
+                    encoding = "UTF-8"
+                if self.mime == "text/csv" or self.mime == "text/plain":
+                    try:
+                        df = await self.open_csv(filename, add_columns, encoding)
+                        if isinstance(filename, PurePath):
+                            self.add_metric(f"{filename.name}", len(df.index))
+                        else:
+                            self.add_metric(f"{filename}", len(df.index))
+                    except Exception as err:
+                        raise ComponentError(f"Encoding Error: {err}") from err
+                    if hasattr(self, "add_columns") and hasattr(self, "rename"):
+                        if self.rename is True:
+                            df = df.drop(df.index[0])
+                elif self.mime in excel_based:
+                    try:
+                        df = await self.open_excel(filename, add_columns, encoding)
+                    except Exception as err:
+                        raise ComponentError(
+                            f"Error parsing Excel: {err}"
+                        ) from err
+                elif self.mime == "text/html" or self.mime == "application/html":
+                    try:
+                        df = await self.open_html(filename, add_columns, encoding)
+                    except Exception as err:
+                        raise ComponentError(f"Error parsing XML: {err}") from err
+                elif self.mime == "application/json":
+                    try:
+                        df = await self.open_json(filename, add_columns, encoding)
+                    except Exception as err:
+                        raise ComponentError(f"Error parsing JSON: {err}") from err
+                else:
+                    raise ComponentError(f"Try to Open invalid MIME Type: {self.mime}")
+                if df is None or df.empty:
+                    raise EmptyFile(f"Empty File {filename}")
+                result.append(df)
+        # at the end, concat the sources:
+        if len(result) == 1:
+            df = result[0]
+        else:
+            ## fix Pandas Concat
+            if self.no_multi is True:  # get only one element
+                df = result.pop()
+            else:
+                try:
+                    df = pandas.concat(
+                        result  # , ignore_index=True  # , sort=False, axis=0,
+                    )  # .reindex(result[0].index)
+                except Exception as err:
+                    raise ComponentError(
+                        f"Error Combining Resultset Dataframes: {err}"
+                    ) from err
+        # post-processing:
+        if hasattr(self, "remove_scientific_notation"):
+            pandas.set_option("display.float_format", lambda x: "%.3f" % x)
+        if hasattr(self, "drop_empty"):
+            df.dropna(axis=1, how="all", inplace=True)
+            df.dropna(axis=0, how="all", inplace=True)
+            df = df.loc[:, ~df.columns.str.contains("^Unnamed")]
+        if hasattr(self, "dropna"):
+            df.dropna(subset=self.dropna, how="all", inplace=True)
+        if hasattr(self, "trim"):
+            # cols = list(df.columns)
+            cols = df.select_dtypes(include=["object", "string"])
+            # def utrim(x): return x.strip() if isinstance(x, str) else x
+            # u.applymap(utrim)
+            for col in cols:
+                df[col] = df[col].astype(str).str.strip()
+        # define the primary keys for DataFrame
+        if hasattr(self, "pk"):
+            try:
+                columns = self.pk["columns"]
+                del self.pk["columns"]
+                df.reset_index().set_index(columns, inplace=True, drop=False, **self.pk)
+            except Exception as err:
+                self._logger.error(f"OpenWith: Error setting index: {err}")
+        if self.clean_nat is True:
+            df.replace({pandas.NaT: None}, inplace=True)
+        if self._colinfo:
+            # fix the datatype for every column in dataframe (if needed)
+            for column, dtype in self._colinfo.items():
+                # print(column, '->', dtype, '->', df[column].iloc[0])
+                try:
+                    if (
+                        dtype == "timestamp without time zone"
+                        or dtype == "timestamp with time zone"
+                        or dtype == "date"
+                    ):
+                        if df[column].dtype != "datetime64[ns]":
+                            df[column] = pandas.to_datetime(df[column], errors="coerce")
+                            df[column] = df[column].astype("datetime64[ns]")
+                    elif (
+                        dtype == "character varying"
+                        or dtype == "character"
+                        or dtype == "text"
+                        or dtype == "varchar"
+                    ):
+                        # print(column, '->', dtype, '->', df[column].iloc[0])
+                        df[column] = df[column].replace([np.nan], "", regex=True)
+                        # df[column] = df[column].astype(str)
+                        # df[column].fillna("", inplace=True)
+                        df[column] = df[column].fillna("")
+                        # df[column].astype(str, inplace=True, errors='coerce')
+                        df[column] = df[column].astype("string", errors="raise")
+                        # df[column].fillna(None, inplace=True)
+                    elif dtype == "smallint":
+                        df[column] = pandas.to_numeric(df[column], errors="coerce")
+                        df[column] = df[column].fillna("").astype("Int8")
+                    elif dtype == "integer" or dtype == "bigint":
+                        try:
+                            ctype = df[column].dtypes[0].name
+                        except (TypeError, KeyError):
+                            ctype = df[column].dtype
+                        if ctype not in ("Int8", "Int32", "Int64"):
+                            df[column] = pandas.to_numeric(df[column], errors="raise")
+                            df[column] = df[column].astype("Int64", errors="raise")
+                        else:
+                            df[column] = df[column].astype("Int64", errors="raise")
+                    elif dtype == "numeric" or dtype == "float":
+                        df[column] = pandas.to_numeric(df[column], errors="coerce")
+                        df[column] = df[column].astype("float64")
+                    elif dtype == "double precision" or dtype == "real":
+                        df[column] = pandas.to_numeric(df[column], errors="coerce")
+                        df[column] = df[column].astype("float64")
+                    elif dtype == "jsonb":
+                        df[column] = df[column].apply(orjson.loads)
+                    elif dtype == "object":
+                        df[column] = df[column].replace([np.nan], "", regex=True)
+                except Exception as err:
+                    print("ERR ::", column, dtype, err, type(err))
+                    self._logger.warning(
+                        f"Cannot set data type for column {column}: {err}"
+                    )
+                    continue
+        self._result = df
+        numrows = len(df.index)
+        self._variables["_numRows_"] = numrows
+        self._variables[f"{self.StepName}_NUMROWS"] = numrows
+        self.add_metric("NUMROWS", numrows)
+        self.add_metric("OPENED_FILES", self._filenames)
+        if self._debug is True:
+            print(df)
+            print("::: Printing Column Information === ")
+            columns = list(df.columns)
+            for column, t in df.dtypes.items():
+                print(column, "->", t, "->", df[column].iloc[0])
+            self._logger.debug(f"Opened File(s) with Pandas {self._filenames}")
+        return self._result