With the load_workbook readonly option, what was earlier seen as a execution seen visibly waiting for many seconds happened with milliseconds. The solution could however be still improved. import pandas as pd from openpyxl import load_workbook class ExcelFile: def __init__(self, **kwargs): ........ ..... self._SheetNames = list(load_workbook(self._name,read_only=True,keep_links=False).sheetnames) The Excelfile.parse takes the same time as reading the complete xls in order of 10s of sec. This result was obtained with windows 10 operating system with below package versions C:\>python -V Python 3.9.1 C:\>pip list Package Version --------------- ------- et-xmlfile 1.0.1 numpy 1.20.2 openpyxl 3.0.7 pandas 1.2.3 pip 21.0.1 python-dateutil 2.8.1 pytz 2021.1 pyxlsb 1.0.8 setuptools 49.2.1 six 1.15.0 xlrd 2.0.1

如果你:

关注业绩 在执行时不需要文件中的数据。 想要使用传统的库还是使用自己的解决方案

下面是一个~10Mb的xlsx, xlsb文件的基准测试。

xls, xlsx

from openpyxl import load_workbook

def get_sheetnames_xlsx(filepath):
    wb = load_workbook(filepath, read_only=True, keep_links=False)
    return wb.sheetnames

基准测试:~ 14倍的速度提升

# get_sheetnames_xlsx vs pd.read_excel
225 ms ± 6.21 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)
3.25 s ± 140 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

xlsb

from pyxlsb import open_workbook

def get_sheetnames_xlsb(filepath):
  with open_workbook(filepath) as wb:
     return wb.sheets

基准测试:~ 56倍的速度提升

# get_sheetnames_xlsb vs pd.read_excel
96.4 ms ± 1.61 ms per loop (mean ± std. dev. of 7 runs, 10 loops each)
5.36 s ± 162 ms per loop (mean ± std. dev. of 7 runs, 1 loop each)

注:

这是一个很好的资源 http://www.python-excel.org/ XLRD从2020年起不再维持

你仍然可以使用ExcelFile类(和sheet_names属性):

xl = pd.ExcelFile('foo.xls')

xl.sheet_names  # see all sheet names

xl.parse(sheet_name)  # read a specific sheet to DataFrame

更多选项参见文档解析…

您应该显式地将第二个参数(sheetname)指定为None。是这样的:

 df = pandas.read_excel("/yourPath/FileName.xlsx", None);

"df"都是一个数据帧字典，你可以通过运行这个来验证:

df.keys()

结果是这样的:

[u'201610', u'201601', u'201701', u'201702', u'201703', u'201704', u'201705', u'201706', u'201612', u'fund', u'201603', u'201602', u'201605', u'201607', u'201606', u'201608', u'201512', u'201611', u'201604']

详情请参考熊猫文档:https://pandas.pydata.org/pandas-docs/stable/generated/pandas.read_excel.html

With the load_workbook readonly option, what was earlier seen as a execution seen visibly waiting for many seconds happened with milliseconds. The solution could however be still improved. import pandas as pd from openpyxl import load_workbook class ExcelFile: def __init__(self, **kwargs): ........ ..... self._SheetNames = list(load_workbook(self._name,read_only=True,keep_links=False).sheetnames) The Excelfile.parse takes the same time as reading the complete xls in order of 10s of sec. This result was obtained with windows 10 operating system with below package versions C:\>python -V Python 3.9.1 C:\>pip list Package Version --------------- ------- et-xmlfile 1.0.1 numpy 1.20.2 openpyxl 3.0.7 pandas 1.2.3 pip 21.0.1 python-dateutil 2.8.1 pytz 2021.1 pyxlsb 1.0.8 setuptools 49.2.1 six 1.15.0 xlrd 2.0.1

from openpyxl import load_workbook

sheets = load_workbook(excel_file, read_only=True).sheetnames

对于我正在使用的5MB Excel文件，没有read_only标志的load_workbook花了8.24秒。对于read_only标志，只需要39.6 ms。如果您仍然希望使用Excel库而不使用xml解决方案，那么这比解析整个文件的方法要快得多。