转载请注明：陈熹 chenx6542@foxmail.com （简书号：半为花间酒）
若公众号内转载请联系公众号：早起Python

本例可以学到的知识点：pandas和openpyxl协同操作excel文件

数据表格：https://pan.baidu.com/s/1mw4IagrRxoCkAvD0Wh5LRA
提取码：l9e8

读者的基础需求如下：

简单来说就是每天都会对Excel中多个sheet进行更新，需要操作完后可以用程序完成第一张sheet 汇总表 的更新

这里假设我手上现在是只有一个sheet的空表，第一步需要完成的工作是：

一、创建多个sheet和随机数据（非核心步骤）

为了尽可能模拟读者的生产环境多了这一步。数据已经充分则可以跳到第二步

我已经把这个excel文件命名好放在桌面的data文件夹中

from openpyxl import load_workbook
import os
import glob
import random

def GetDesktopPath():
    return os.path.join(os.path.expanduser("~"), 'Desktop')

# 调用glob可以利用通配符获取指定命名格式的文件
path = glob.glob(f'{GetDesktopPath()}/data/*.xls*')[0]
workbook = load_workbook(filename=path)
sheet_init = workbook.active

接着是创建一些供随机的内容，可以随意写

name_lst = ['皮卡丘', '小火龙', '杰尼龟', '妙蛙种子', '风速狗', '小拳石', '飞天螳螂']
place_lst = [chr(i).upper() for i in range(97, 123)] 
# 我忘记大写字母的码了哈哈哈 这样变着法换大写字母
activity_lst = ['椭圆机', '篮球', '足球', '羽毛球', '跳绳']
source_lst = ['朋友介绍', '微信聊天', '网页弹窗', '其他']

for i in range(30):
    # 不断去拷贝第一页并重命名
    sheet = workbook.copy_worksheet(sheet_init)
    sheet.title = f'{i+1}日'
    for j in range(random.randint(10, 30)):
        # 从第三行开始行遍历
        for row in sheet.iter_rows(min_row=3+j, max_row=3+j):
            info = [f'{j+1}', f'{i+1}日', f'{random.choice(name_lst)}', f'{random.choice(place_lst)}馆',
                    f'{random.choice(activity_lst)}', f'{random.choice(source_lst)}', f'{random.randint(1, 10)}',
                    '无', f'{random.choice(["Y", "N"])}', f'{random.choice(["Y", "N"])}', f'{random.choice(["Y", "N"])}']
            # 嵌套循环，对当前行的格子进行遍历把内容写入
            for index, k in enumerate(info):
                row[index].value = k
    print(f'第{i+1}日已完成')

workbook.save(filename=f'{GetDesktopPath()}/data/results.xlsx')

数据就创建好了，然后正式进入读者的问题：

二、合并多个sheet并写入汇总sheet

由于后面多个表的更新后需要按日期顺序在汇总表里呈现，因此有一个策略是利用openpyxl按顺序遍历各表然后写回汇总表

但注意，表格中存在边框、居中等样式修改

这种情况下，openpyxl会识别样式，认为这些行是已经有数据的，故纯粹的sheet.append()方法是无法将数据写入这些所谓的空行，而会从没有样式的行开始写入

所以需要在各表写入的时候不断计算所在行，并利用sheet.iter_rows()定位

是不是有点麻烦？

因此换了简单的思路：利用pandas，其方便的地方在于无视表格样式

path_new = glob.glob(f'{GetDesktopPath()}/data/results.xls*')[0]
workbook_new = load_workbook(filename=path_new)
# 方便获取总表数便于遍历
sheetnames = workbook.sheetnames
df_lst = []

for i in range(1, len(sheetnames)):
    df = pd.read_excel(path_new , encoding='utf-8', sheet_name=i, skiprows=1)
    df_lst.append(df)

# 把获取的各表纵向合并，注意纵向合并常常需要重置索引
df_total = pd.concat(df_lst,axis=0,ignore_index=True)
# 索引是从0开始，利用索引+1重置各记录的编号
df_total['编号'] = df_total.index + 1

将生成的表写回汇总表即可，涉及的内容稍微比较复杂。由于直接使用dataframe.to_excel会覆盖原excel导致只有一张sheet，其他全部丢失，需要利用pd.ExcelWriter，具体见代码

删除原来的 汇总表 并写入新的 汇总表。因为新写入的sheet会置于末尾，可以用list.insert(0, list.pop())将最后一个元素置于开头

writer = pd.ExcelWriter(path_new, engine='openpyxl')
writer.book = workbook
workbook.remove(workbook['汇总表'])
df_total.to_excel(excel_writer=writer, sheet_name=u'汇总表', index=None)
writer.close()
workbook._sheets.insert(0, workbook._sheets.pop())
workbook.save(filename=f'{GetDesktopPath()}/data/results.xlsx')

这就完成了吗？没有

pandas的优势“无视样式”也成为了它的缺陷：写入文件时没有样式信息
因此最后再用openpyxl对第一页的样式调整

三、设置样式

# 设置对齐、线性、边框、字体
from openpyxl.styles import Alignment
from openpyxl.styles import Side, Border
from openpyxl.styles import Font

sheet = workbook[sheetnames[0]]
sheet.insert_rows(idx=0) # 插入第一行
font = Font(name='宋体', size=18, bold=True)
sheet['A1'] = '皮卡丘体育2020年06月新学员信息登记表'
sheet['A1'].font = font # 设置字体大小和加粗

req = ':(\w)'
weight = re.findall(req, sheet.dimensions)[0]
sheet.merge_cells(f'A1:{weight}1')

# 样式先准备好
alignment = Alignment(horizontal='center', vertical='center')
side = Side(style='thin', color='000000')
border = Border(left=side, right=side, top=side, bottom=side)

# 遍历cell设置样式
rows = sheet[f'{sheet.dimensions}']
for row in rows:
    for cell in row:
        cell.alignment = alignment
        cell.border = border

# 设置前两行的行高
sheet.row_dimensions[1].height = 38
sheet.row_dimensions[2].height = 38

# 设置列宽
letter_lst = [chr(i+64).upper() for i in range(2, ord(weight)-ord('A')+1+1)]
sheet.column_dimensions['A'].width = 8
for i in letter_lst:
     sheet.column_dimensions[f'{i}'].width = 14

workbook.save(filename=f'{GetDesktopPath()}/data/results.xlsx')

样式就修改完成了！

完整代码

from openpyxl import load_workbook
import os
import glob
import random
import pandas as pd
import re
from openpyxl.styles import Alignment
from openpyxl.styles import Side, Border
from openpyxl.styles import Font

def GetDesktopPath():
    return os.path.join(os.path.expanduser("~"), 'Desktop')

path = glob.glob(f'{GetDesktopPath()}/data/*.xls*')[0]
workbook = load_workbook(filename=path)
sheet_init = workbook.active
name_lst = ['皮卡丘', '小火龙', '杰尼龟', '妙蛙种子', '风速狗', '小拳石', '飞天螳螂']
place_lst = [chr(i).upper() for i in range(97, 123)]
activity_lst = ['椭圆机', '篮球', '足球', '羽毛球', '跳绳']
source_lst = ['朋友介绍', '微信聊天', '网页弹窗', '其他']
for i in range(30):
    sheet = workbook.copy_worksheet(sheet_init)
    sheet.title = f'{i+1}日'
    for j in range(random.randint(10, 30)):
        for row in sheet.iter_rows(min_row=3+j, max_row=3+j):
            info = [f'{j+1}', f'{i+1}日', f'{random.choice(name_lst)}', f'{random.choice(place_lst)}馆',
                    f'{random.choice(activity_lst)}', f'{random.choice(source_lst)}', f'{random.randint(1, 10)}',
                    '无', f'{random.choice(["Y", "N"])}', f'{random.choice(["Y", "N"])}', f'{random.choice(["Y", "N"])}']
            for index, k in enumerate(info):
                row[index].value = k

workbook.save(filename=f'{GetDesktopPath()}/data/results.xlsx')

path_new = glob.glob(f'{GetDesktopPath()}/data/results.xls*')[0]
# 方便获取总表数便于遍历
workbook = load_workbook(path_new)
sheetnames = workbook.sheetnames
df_lst = []
for i in range(1, len(sheetnames)):
    df = pd.read_excel(path_new, encoding='utf-8', sheet_name=i, skiprows=1)
    df_lst.append(df)

df_total = pd.concat(df_lst, axis=0, ignore_index=True)
df_total['编号'] = df_total.index + 1

writer = pd.ExcelWriter(path_new, engine='openpyxl')
writer.book = workbook
workbook.remove(workbook['汇总表'])
df_total.to_excel(excel_writer=writer, sheet_name=u'汇总表', index=None)
writer.close()
workbook._sheets.insert(0, workbook._sheets.pop())

sheet = workbook[sheetnames[0]]
sheet.insert_rows(idx=0)
font = Font(name='宋体', size=18, bold=True)
sheet['A1'] = '皮卡丘体育2020年06月新学员信息登记表'
sheet['A1'].font = font

req = ':(\w)'
weight = re.findall(req, sheet.dimensions)[0]
sheet.merge_cells(f'A1:{weight}1')

alignment = Alignment(horizontal='center', vertical='center')
side = Side(style='thin', color='000000')
border = Border(left=side, right=side, top=side, bottom=side)

rows = sheet[f'{sheet.dimensions}']
for row in rows:
    for cell in row:
        cell.alignment = alignment
        cell.border = border

sheet.row_dimensions[1].height = 38
sheet.row_dimensions[2].height = 38

letter_lst = [chr(i+64).upper() for i in range(2, ord(weight)-ord('A')+1+1)]
sheet.column_dimensions['A'].width = 8
for i in letter_lst:
     sheet.column_dimensions[f'{i}'].width = 14

workbook.save(filename=f'{GetDesktopPath()}/data/results.xlsx')