使用Glue python的S3文件的当前时间戳与上次修改时间戳之间的差异

import sys
from datetime import datetime
from awsglue.transforms import *
from awsglue.utils import getResolvedOptions
from pyspark.context import SparkContext
from awsglue.context import GlueContext
from pyspark.sql.functions import unix_timestamp

# 初始化GlueContext和SparkContext
sc = SparkContext()
glueContext = GlueContext(sc)
spark = glueContext.spark_session

# 获取S3的bucket名称和文件路径参数
args = getResolvedOptions(sys.argv, ['S3Bucket', 'FilePath'])
s3_bucket = args['S3Bucket']
file_path = args['FilePath']

# 读取S3文件
df = spark.read.format("csv").option("header", "true").load("s3://{}/{}".format(s3_bucket, file_path))

file_stats = df.select("timestamp_column").describe()
current_timestamp = datetime.now()
last_modified_timestamp = file_stats.collect()[0]["timestamp_column"]

time_diff = current_timestamp - last_modified_timestamp

# 判断时间差异是否超过1小时
if time_diff.total_seconds() > 3600:
    print("文件已超过1小时未更新")

# 将时间差异转换成分钟格式
time_diff_minutes = divmod(time_diff.total_seconds(), 60)
print("时间差异（分钟）：", time_diff_minutes)