Python时区转换问题：本地时间转UTC的日期溢出处理

问题背景 #

我有一个结构为(store_id, day, start_time_local, end_time_local)的DataFrame，数据样例如下：

1481966498820158979 0   "00:00:00"  "00:10:00"
1481966498820158979 0   "11:00:00"  "23:59:00"
1481966498820158979 1   "00:00:00"  "00:10:00"
1481966498820158979 1   "11:00:00"  "23:59:00"
1481966498820158979 2   "00:00:00"  "00:10:00"
1481966498820158979 2   "11:00:00"  "23:59:00"
1481966498820158979 3   "00:00:00"  "00:10:00"
1481966498820158979 3   "11:00:00"  "23:59:00"
1481966498820158979 4   "00:00:00"  "00:10:00"
1481966498820158979 4   "11:00:00"  "23:59:00"
1481966498820158979 5   "00:00:00"  "00:10:00"
1481966498820158979 5   "11:00:00"  "23:59:00"
1481966498820158979 6   "00:00:00"  "00:10:00"
1481966498820158979 6   "11:00:00"  "23:59:00"

该数据集的时区为本地时区，需要将其转换为UTC时区。我编写了以下函数，但无法正确处理日期溢出问题：

# Define a function to convert local time to UTC and handle day change
def local_to_utc_with_day(row, timezone_string):
    timezone_obj = pytz.timezone(timezone_string)
    local_start_time = timezone_obj.localize(row['start_time_local'])
    local_end_time = timezone_obj.localize(row['end_time_local'])
    
    utc_start_time = local_start_time.astimezone(pytz.UTC)
    utc_end_time = local_end_time.astimezone(pytz.UTC)
    
    # Check if the day has changed due to timezone shift
    if utc_start_time.date() != row['start_time_local'].date():
        row['day'] += 1
    
    row['start_time_utc_formatted'] = utc_start_time.strftime('%H:%M:%S')  # UTC time in HH:MM:SS format
    row['end_time_utc_formatted'] = utc_end_time.strftime('%H:%M:%S')
    row['start_time_local'] = local_start_time.strftime('%H:%M:%S')  # Keep HH:MM:SS format
    row['end_time_local'] = local_end_time.strftime('%H:%M:%S')  # Keep HH:MM:SS format
    return row

问题分析与修复方案 #

原函数的核心问题在于：

• localize方法误用：若row['start_time_local']是时间字符串，localize无法直接处理；若已是带时区的datetime，调用localize会报错，必须先结合day字段构建完整本地datetime对象。
• 日期溢出判断逻辑疏漏：仅判断开始时间的日期变化，未考虑结束时间跨天场景，也未处理day超过6（一周7天）的循环问题。

修复后的函数 #

import pytz
from datetime import datetime, timedelta

def local_to_utc_with_day(row, timezone_string):
    timezone_obj = pytz.timezone(timezone_string)
    
    # 构建本地日期：用基准日期+day偏移生成完整日期，基准日期可根据业务调整
    base_date = datetime(2024, 1, 1)
    local_date = base_date + timedelta(days=row['day'])
    
    # 解析本地时间字符串，移除引号后转换为时间对象
    start_time_str = row['start_time_local'].strip('"')
    end_time_str = row['end_time_local'].strip('"')
    start_time = datetime.strptime(start_time_str, '%H:%M:%S').time()
    end_time = datetime.strptime(end_time_str, '%H:%M:%S').time()
    
    # 结合日期与时间，生成带时区的本地datetime
    local_start_dt = timezone_obj.localize(datetime.combine(local_date, start_time))
    local_end_dt = timezone_obj.localize(datetime.combine(local_date, end_time))
    
    # 转换为UTC时间
    utc_start_dt = local_start_dt.astimezone(pytz.UTC)
    utc_end_dt = local_end_dt.astimezone(pytz.UTC)
    
    # 计算UTC对应的day，取模7保证在0-6范围内
    day_diff = (utc_start_dt.date() - local_date.date()).days
    row['day'] = (row['day'] + day_diff) % 7
    
    # 格式化输出时间
    row['start_time_utc_formatted'] = utc_start_dt.strftime('%H:%M:%S')
    row['end_time_utc_formatted'] = utc_end_dt.strftime('%H:%M:%S')
    row['start_time_local'] = start_time_str
    row['end_time_local'] = end_time_str
    
    return row

关键改进点 #

• 完整构建本地datetime：将day字段与时间字符串结合，生成包含日期和时区的完整datetime对象，避免localize方法误用。
• 准确计算日期偏移：通过UTC日期与本地日期的差值计算day变化，用取模7保证day始终符合一周7天的业务逻辑。
• 规范时间字符串解析：移除时间字符串的引号，确保正确转换为时间对象。

内容的提问来源于stack exchange，提问作者Sumit Vishwakarma