Utiliser Aspose.Cells pour Python via .NET comme moteur Excel pour pandas

Pourquoi utiliser Aspose.Cells pour Python via .NET ?

Aspose.Cells offre :

  • Support avancé d’Excel (formules, graphiques, mise en forme, cellules fusionnées, etc.)
  • Support pour plusieurs formats : .xls, .xlsx, .xlsb, .ods, .csv, .html
  • Meilleure précision pour les feuilles de calcul complexes par rapport à openpyxl ou xlrd

Prerequisites

  • Install a C compiler
  • Here, we use the Windows platform as an example for explanation.If you have installed Visual Studio 2022 on your Windows system, you can open the x64 Native Tools Command Prompt for VS 2022 and run the cl command to check the version of the C++ compiler. Make sure the compiler version is 19.3x or higher before proceeding with the following build steps.
  • Make sure that you have cloned the repository
git clone https://github.com/pandas-dev/pandas.git
  • cd vers le répertoire source pandas que vous venez de créer avec la commande clone

Step 1: Create an isolated environment

# Set up virtual environment
python -m venv .venv
.\.venv\Scripts\activate      # on Windows
# source .venv/bin/activate  # on Linux/macOS

# Install aspose-cells-python
pip install aspose-cells-python
# Install the build dependencies
python -m pip install -r requirements-dev.txt

Step 2: Create Aspose engine adapter

Create a new file:
pandas/io/excel/_asposecells.py
Add the following content:

# pandas/io/excel/_asposecells.py
import pandas as pd
from aspose.cells import Workbook

class AsposeCellsExcelReader:
    def __init__(self, filepath_or_buffer, sheet_name=0, header=0, **kwargs):
        self.filepath = filepath_or_buffer
        self.sheet_name = sheet_name
        self.header = header

    def parse(self, sheet_name, header=0, **kwargs):
        wb = Workbook(self.filepath)
        worksheet = wb.worksheets[sheet_name] if isinstance(sheet_name, int) else wb.worksheets.get(sheet_name)
        # Get the Cells collection from the worksheet
        cells = worksheet.cells

        # Calculate number of columns: max_col - min_col (both are 0-based)
        col_count = cells.max_data_column - cells.min_data_column

        # Initialize a list to hold all the row data
        output_data = []

        # Get the index of the first row that contains data
        first_data_row_Index = cells.min_data_row

        # Iterate through all the rows
        for row in cells.rows:
            if row is None:
                continue  # Skip if the row is not initialized

            row_data = []
            for cell in row:
                row_data.append(cell.value)
            output_data.append(row_data)

        # Prepare the column names
        columns = []
        if header is not None:
            row = cells.rows[first_data_row_Index]
            for cell in row:
                columns.append(cell.value)
            # Remove the header row from the data
            output_data = output_data[1:]
        else:
            # If no header, generate default column names like "Unnamed: 0", "Unnamed: 1", ...
            columns = [f"Unnamed: {i}" for i in range(col_count + 1)]

        # Convert the data into a pandas DataFrame
        return pd.DataFrame(output_data, columns=columns)

    def close(self):
        pass  # Required by pandas API

Step 3: Register the Aspose Engine in Pandas

In pandas/io/excel/_base.py, find the class ExcelFile class, and add Add the following import line after the existing from pandas.io.excel._xlrd import XlrdReader:

from pandas.io.excel._asposecells import AsposeCellsExcelReader

et ajoutez alors le code suivant dans _engines: Mapping[str, Any]

_engines: Mapping[str, Any] = {
    ...
    "asposecells": AsposeCellsExcelReader,
}

Step 4: Build and install pandas

# build and install pandas
python -m pip install -ve . --no-build-isolation

✅ Si vous rencontrez une erreur de compilation lors du processus de construction, comme :Cython.Compiler.Errors.InternalError: Internal compiler error: ‘free_threading_config.pxi’ not found, vous pouvez essayer d’exécuter la commande suivante, puis recompiler.

# remove all untracked files, directories, and ignored files from the working directory.
git clean -xfd

Step 5: Use the Engine

You can use the following Excel file for testing.

import pandas as pd
# asposecells
df = pd.read_excel("test.xlsx", engine="asposecells", sheet_name=0, header=0)
# print and check DataFrame
print(df)

Après l’avoir exécutée, vous devriez obtenir un résultat comme celui-ci.

todo:image_alt_text

Notes

  • For production use, a valid Aspose.Cells for Python via .NET license is required.
  • This approach is ideal for testing or local enhancement of read_excel behavior.
  • You can also use Aspose.Cells for Python via .NET directly outside of pandas if desired.

Voir aussi

Convertir une classe définie par l’utilisateur en Excel