关于CSV格式

逗号分隔值（Comma-Separated Values，CSV，有时也称为字符分隔值，因为分隔字符也可以不是逗号），其文件以纯文本形式存储表格数据（数字和文本）。“CSV”并不是一种单一的、定义明确的格式（尽管RFC 4180有一个被通常使用的定义）。

python中csv模块中定义的函数：

csv.reader(csvfile, dialect=‘excel’, **fmtparams)

返回一个可以遍历csv文件的reader对象。dialect参数可以用于定义一组特定的csv方言参数，是Dialect类的子类或者list_dialects()函数返回的字符串。从csv文件读取的每一行都作为字符串列表返回。除非指定了QUOTE_NONNUMERIC格式选项(在这种情况下，未加引号的字段将转换为浮点数)，否则不会执行自动的数据类型转换。

待处理CSV文件

此文件是外部接口提供的文件，由于时间是比较久远的软件，或者，其他原因，内容有些散乱，如下图所示：

示例数据如下：

"","","","","","","","","","","","","","","","","油品销售明细表","","","","","","","","","","","",""
"","","加油站名称：","","","","广州*********加油站        ","","","","","","","","","","","","","","","","","","","","","",""
"","从：","","2021-10-01 00:00:00","","","","","","","到：","2022-10-01 23:59:59","","","","","","","","","","","","","","","","",""
"","流水号","","","","","","交易时间","","","","","油枪号码","","油品名称","","","油品单价","","体积","","交易金额","","起泵码","","止泵码","","","备注"
"","","","","479171","","","","2021-10-23 16:21:00","","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","58.83","","479.49","","380693.19","","","380752.02",""
"","","","","","259635","","","","2021-10-23 16:32:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","60.90","","493.00","","380752.02","","","380812.92",""
"","","","","","259636","","","","2021-10-23 16:34:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","56.19","","446.95","","380812.92","","","380869.11",""
"","","","","","479251","","","","2021-10-23 18:30:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","70.76","","573.94","","380869.11","","","380939.87",""
"","","","","","86765","","","","2021-10-23 18:35:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","44.03","","361.49","","380939.87","","","380983.90",""
"","","","","479289","","","","","2021-10-23 20:11:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","6.09","","50.00","","380983.90","","","380989.99",""
"","","","","","86775","","","","2021-10-23 20:30:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","49.13","","393.53","","380989.99","","","381039.12",""
"","","","","479309","","","","","2021-10-23 21:23:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","24.36","","200.00","","381039.12","","","381063.48",""
"","","","","","479413","","","","2021-10-24 03:29:00","","","","1","","95号 车用汽油(ⅥA)","","","8.21","","33.47","","271.29","","381063.48","","","381096.95",""
"打印时间：2022-10-28","","","","","","","","","","","","","","","","","","","","","","","","","","填表人：","",""
"","流水号","","","交易时间","","油枪号码","","油品名称","","油品单价","","体积","","交易金额","","起泵码","","止泵码","","","备注"
"","","","86814","","2021-10-24 09:47:00","","1","","95号 车用汽油(ⅥA)","","8.21","","52.29","","429.30","","381157.85","","","381210.14",""
"","","259822","","","2021-10-24 09:59:00","","1","","95号 车用汽油(ⅥA)","","8.21","","46.09","","374.90","","381210.14","","","381256.23",""

python使用csv模块解析数据

方法一，是按单元格逐行个性化解析，例如参考上次XLS格式数据处理《Python按单元格读取复杂电子表格（Excel）数据实践》，这个方法，挺麻烦的，发现第二个方法后，过段放弃此方法。

方法二，提取有效数据解析，由于CSV格式数据不跨行，可以逐行剔除空项，而直接取有效数据，代码非常简单，如下所示：

import csv
import pandas as pd# 以读方式打开文件
dat_row = []
with open("油品销售明细202110-202210.CSV", mode="r") as f:    # 基于打开的文件，创建csv.reader实例reader = csv.reader(f)# 逐行获取数据，并输出for row in reader:dat_col = [v for v in row if len(v)>0]n = n + 1if len(dat_col)==9:dat_row.append(dat_col)cols_list = ['流水号', '交易时间', '油枪号码', '油品名称', '油品单价', '体积', '交易金额', '起泵码', '止泵码']
df = pd.DataFrame(dat_row,columns=cols_list)
df.to_csv('detail.csv',encoding='utf_8_sig',index=False)

注：其中，“dat_col = [v for v in row if len(v)>0]”代码是按行，过滤没有数据的单元格。

小结

对于没有合并单元格（此处为跨行）的数据文件解析，使用适当的方法还是很简单的，非常喜欢简单的方法！

参考：

快乐江小鱼. Python基础 - csv文件格式. CSDN博客. 2022.08
肖永威. 《Python按单元格读取复杂电子表格（Excel）数据实践》. CSDN博客. 2022.11