使用 UDF 解析 PySpark Dataframe 中的嵌套 XML 字段

Question

我有一个场景，数据框列中有 XML 数据。

性别	更新时间	访客
F	1574264158	<?xml版本=“1.0”编码=“utf-8”

我想使用 UDF 解析 - 访客列 - 将嵌套的 XML 字段解析到 Dataframe 中的列中

XML 格式

<?xml version="1.0" encoding="utf-8"?> <visitors> <visitor id="9615" age="68" sex="F" /> <visitor id="1882" age="34" sex="M" /> <visitor id="5987" age="23" sex="M" /> </visitors>

Answer 1

您可以xpath在不使用 UDF 的情况下使用查询：

df = spark.createDataFrame([['<?xml version="1.0" encoding="utf-8"?> <visitors> <visitor id="9615" age="68" sex="F" /> <visitor id="1882" age="34" sex="M" /> <visitor id="5987" age="23" sex="M" /> </visitors>']], ['visitors'])

df.show(truncate=False)
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
|visitors                                                                                                                                                                          |
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+
|<?xml version="1.0" encoding="utf-8"?> <visitors> <visitor id="9615" age="68" sex="F" /> <visitor id="1882" age="34" sex="M" /> <visitor id="5987" age="23" sex="M" /> </visitors>|
+----------------------------------------------------------------------------------------------------------------------------------------------------------------------------------+


df2 = df.selectExpr(
    "xpath(visitors, './visitors/visitor/@id') id",
    "xpath(visitors, './visitors/visitor/@age') age",
    "xpath(visitors, './visitors/visitor/@sex') sex"
).selectExpr(
    "explode(arrays_zip(id, age, sex)) visitors"
).select('visitors.*')

df2.show(truncate=False)
+----+---+---+
|id  |age|sex|
+----+---+---+
|9615|68 |F  |
|1882|34 |M  |
|5987|23 |M  |
+----+---+---+

如果您坚持使用 UDF：

import xml.etree.ElementTree as ET
import pyspark.sql.functions as F

@F.udf('array<struct<id:string, age:string, sex:string>>')
def parse_xml(s):
    root = ET.fromstring(s)
    return list(map(lambda x: x.attrib, root.findall('visitor')))
    
df2 = df.select(
    F.explode(parse_xml('visitors')).alias('visitors')
).select('visitors.*')

df2.show()
+----+---+---+
|  id|age|sex|
+----+---+---+
|9615| 68|  F|
|1882| 34|  M|
|5987| 23|  M|
+----+---+---+