descriptor 'decode' requires a 'bytes' object but received a 'NoneType'

記錄在使用python過程中踩的坑------
使用xlwt庫對excel文件進行保存時報錯 descriptor 'decode' requires a 'bytes' object but received a 'NoneType'

log:

Traceback (most recent call last):
File "F:/xxxxxx/util/ExcelUtil.py", line 110, in
excel = write_excel("Outbound_Template.xls", 4, 9, "AT2019110912")
File "F:/xxxxxx/util/ExcelUtil.py", line 48, in write_excel
newWb.save(excel_path)
File "F:\Python\lib\site-packages\xlwt\Workbook.py", line 710, in save
doc.save(filename_or_stream, self.get_biff_data())
File "F:\Python\lib\site-packages\xlwt\Workbook.py", line 667, in get_biff_data
before += self.__all_fonts_num_formats_xf_styles_rec()
File "F:\Python\lib\site-packages\xlwt\Workbook.py", line 570, in __all_fonts_num_formats_xf_styles_rec
return self.__styles.get_biff_data()
File "F:\Python\lib\site-packages\xlwt\Style.py", line 185, in get_biff_data
result += self._all_num_formats()
File "F:\Python\lib\site-packages\xlwt\Style.py", line 209, in _all_num_formats
result += NumberFormatRecord(fmtidx, fmtstr).get()
File "F:\Python\lib\site-packages\xlwt\BIFFRecords.py", line 785, in init
ufmtstr = upack2(fmtstr)
File "F:\Python\lib\site-packages\xlwt\UnicodeUtils.py", line 50, in upack2
us = unicode(s, encoding)
TypeError: descriptor 'decode' requires a 'bytes' object but received a 'NoneType'

1.原文件用MS Excel編輯，后執行程序，可以運行
2.原文件用MS Excel編輯，后執行程序，更新后的文件，再用MS Excel編輯，保存，再執行程序，出現上面的錯誤

我的代碼：

# coding:utf-8
def write_excel(file_name, row, col, value, sheet_name="Template"):
    excel_path = Configure.read_config("project_path", "path") + "\\data\\" + file_name
    # formatting_info = True，得以保存之前數據的格式
    old_wb = open_workbook(excel_path, formatting_info=True)
    # 將操作文件對象拷貝，變成可寫的workbook對象
    new_wb = copy(old_wb)
    # 獲得sheet的對象
    new_ws = new_wb.get_sheet(sheet_name)
    # 寫入數據
    new_ws.write(row, col, value)
    # 另存為excel文件，並將文件命名
    new_wb.save(excel_path)

原因：
用xlutils.copy 得到的excel文件，編碼格式變了。然后用Office365或WPS去編輯再保存，得到的文件，decode會是None，xlwt save時會出錯。（UnicodeUtils.py中的upack2沒有考慮到會有None的情況，網上有的解決方式是自己在源碼里加多判斷None的情況）

源碼：

def upack2(s, encoding='ascii'):
    # If not unicode, make it so.
    if isinstance(s, unicode_type):
        us = s
    else:
        us = unicode(s, encoding)
    # Limit is based on number of content characters
    # (not on number of bytes in packed result)
    len_us = len(us)
    if len_us > 32767:
        raise Exception('String longer than 32767 characters')
    try:
        encs = us.encode('latin1')
        # Success here means all chars are in U+0000 to U+00FF
        # inclusive, meaning that we can use "compressed format".
        flag = 0
        n_items = len_us
    except UnicodeEncodeError:
        encs = us.encode('utf_16_le')
        flag = 1
        n_items = len(encs) // 2
        # n_items is the number of "double byte characters" i.e. MS C wchars
        # Can't use len(us).
        # len(u"\U0001D400") -> 1 on a wide-unicode build 
        # and 2 on a narrow-unicode build.
        # We need n_items == 2 in this case.
    return pack('<HB', n_items, flag) + encs

解決方法：用OpenOffice去編輯文件保存就不會有問題