草稿。等我以后有机会我会整理一下。

cat name_date.csv                                                                                                                                                                                                       
john:01.01.2001
mary:06.03.2016
sue:

cat name_city.csv                                                                                                                                                                                                       
john:london
mary:new york
bob:

import csv

with open("name_date.csv") as dt_csv:
    new_dict = {}
    dt_dictR = csv.DictReader(dt_csv, fieldnames=["name", "date"],  delimiter=':')
    for row in dt_dictR:
        if not row["date"]:
            row["date"] = '-'
        new_dict.update({row["name"]: {"date": row["date"]}})
    with open("name_city.csv") as city_csv:
        dt_dictC = csv.DictReader(city_csv, fieldnames=["name", "city"],  delimiter=':')
        print(new_dict)
        for row in dt_dictC:
            if not row["city"]: 
                row["city"] = '-' 
            if new_dict.get(row["name"]):
                new_dict[row["name"]].update({"city": row["city"]})
            else:
                new_dict.update({row["name"]: {"date": '-', "city": row["city"]}})
    with open("merged_csv", "w", newline='') as out_file:
        csv_w = csv.writer(out_file)
        csv_w.writerow(["name","town","date"])
        for item in new_dict:
            if not new_dict[item].get("city"):
                new_dict[item]["city"] = '-'
            csv_w.writerow([item, new_dict[item]["city"], new_dict[item]["date"]])

cat merged_csv                                                                                                                                                                                                          
name,town,date
john,london,01.01.2001
mary,new york,06.03.2016
sue,-,-
bob,-,-

使用defaultdict进行一些简化：

import csv 
from collections import defaultdict

with open("name_date.csv") as dt_csv:
    def cityDateDict():
        return {"city": "-", "date": "-"}
    new_dict = defaultdict(cityDateDict)
    dt_dictR = csv.DictReader(dt_csv, fieldnames=["name", "date"],  delimiter=':')
    for row in dt_dictR:
        new_dict[row["name"]]
        if row["date"].strip():
            new_dict[row["name"]]["date"] = row["date"]
    with open("name_city.csv") as city_csv:
        dt_dictC = csv.DictReader(city_csv, fieldnames=["name", "city"],  delimiter=':')
        for row in dt_dictC:
            new_dict[row["name"]] 
            if row["city"].strip():
                new_dict[row["name"]]["city"] = row["city"] 
    with open("merged_csv", "w", newline='') as out_file:
        csv_w = csv.writer(out_file)
        csv_w.writerow(["name","town","date"])
        for item in new_dict:
            csv_w.writerow([item, new_dict[item]["city"], new_dict[item]["date"]])

defaultdict允许您使用'default'值动态构建字典。在本例中，城市/日期dict的默认值为-。然后，可以使用非空值更新相应的关键字（城市/日期）以覆盖默认值。