EXCEL file vectorization data type exception issue

[sagitar-dj]

When using EXCEL files to parse by table, the parsing method will analyze the data columns based on the previous rows of data, and errors will occur for special data, such as character type 0101 will be recognized as a number, and some will be recognized as a date. Then for large models, data types are not so important, and they should all be parsed by character type.

使用EXCEL文件按table方式解析时，解析方法会根据前几行数据分析数据列，对于特殊数据会发生错误如：字符类型 0101 会被识别为数字，有些会被识别为日期。然后对于大模型来说数据类型不是那么重要，应该全部按字符类型进行解析即可。

[Llala0928]

@sagitar-dj Thank you for bringing up this detail. You can first convert the file to CSV. We will also consider improvements, such as adding an "Enforce Text Mode" switch, which will default all columns to be treated as strings, so you can directly upload Excel files without errors.