您可以获取df中的max date，并与原始Dataframe进行交叉连接，然后使用一些when表达式来设置description列：

from pyspark.sql import functions as F
# convert date into date type
df = df.withColumn("date", F.to_date(F.col("date").cast("string"), "yyyyMMdd"))
max_date_df = df.agg(F.max("date").alias("max_date"))
result = df.crossJoin(max_date_df) \
    .withColumn(
    "text",
    F.when(
        F.year("date") < F.year("max_date"),
        F.lit("prev year")
    ).when(
        (F.year("date") == F.year("max_date")) & (F.quarter("date") == F.quarter("max_date") - 1),
        F.lit("prev quarter")
    ).when(
        (F.year("date") == F.year("max_date")) & (F.quarter("date") == F.quarter("max_date")),
        F.lit("current quarter")
    ).otherwise(F.lit("recent year"))
).select("date", "text")
result.show()
# +----------+---------------+
# |      date|           text|
# +----------+---------------+
# |2017-09-30|      prev year|
# |2017-10-31|      prev year|
# |2017-11-30|      prev year|
# |2018-01-31|    recent year|
# |2018-02-28|    recent year|
# |2018-03-31|    recent year|
# |2018-04-30|    recent year|
# |2018-05-31|    recent year|
# |2018-06-30|    recent year|
# |2018-07-31|   prev quarter|
# |2018-07-31|   prev quarter|
# |2018-08-31|   prev quarter|
# |2018-08-31|   prev quarter|
# |2018-09-30|   prev quarter|
# |2018-09-30|   prev quarter|
# |2018-09-30|   prev quarter|
# |2018-10-31|current quarter|
# |2018-10-31|current quarter|
# |2018-11-30|current quarter|
# |2018-11-30|current quarter|
# +----------+---------------+