You are viewing a plain text version of this content. The canonical link for it is here.
Posted to user-zh@flink.apache.org by 小学生 <20...@qq.com> on 2020/06/04 06:46:23 UTC

pyflink 嵌套使用函数出错

各位大佬好,初学pyflink,有一个问题需要帮忙解决下。


代码为:
from pyflink.table import StreamTableEnvironment, DataTypes, EnvironmentSettings,TableConfig,BatchTableEnvironment
from pyflink.table.descriptors import Schema, OldCsv, FileSystem
from pyflink.table.udf import udf
from pyflink.datastream import StreamExecutionEnvironment
elements = 'aaa|bbb'
env = StreamExecutionEnvironment.get_execution_environment()
env.set_parallelism(1)
t_env = StreamTableEnvironment.create(env, environment_settings=EnvironmentSettings.new_instance().use_blink_planner().build())


@udf(input_types=[DataTypes.STRING()],
&nbsp; &nbsp; &nbsp;result_type=DataTypes.ARRAY(DataTypes.STRING()))
def split(x):
&nbsp; &nbsp; return x.strip().split("|")
# t_env.register_function("split", udf(lambda i: i.strip().split("|"), [DataTypes.STRING()], DataTypes.ARRAY(DataTypes.STRING())))
t_env.register_function("split", split)
#split拆分后为一个2元数组
@udf(input_types=[DataTypes.ARRAY(DataTypes.STRING()), DataTypes.INT()],
&nbsp; &nbsp; &nbsp;result_type=DataTypes.STRING())
def get(array, index):
&nbsp; &nbsp; return array[index]


@udf(input_types=[DataTypes.ARRAY(DataTypes.STRING())],
&nbsp; &nbsp; &nbsp;result_type=DataTypes.STRING())
def convert(array):
&nbsp; &nbsp; return get(array, 0)


t_env.register_function("convert", convert)


t_env.connect(FileSystem().path('/tmp/output')) \
&nbsp; &nbsp; .with_format(OldCsv()
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.field('b', DataTypes.STRING())) \
&nbsp; &nbsp; .with_schema(Schema()
&nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.field('b', DataTypes.STRING())) \
&nbsp; &nbsp; .create_temporary_table('mySink')


t_env.from_elements(elements)\
&nbsp; &nbsp; &nbsp;.alias('line')\
&nbsp; &nbsp; &nbsp;.select('split(line)')\
&nbsp; &nbsp; &nbsp;.alias('array')\
&nbsp; &nbsp; &nbsp;.select('convert(array) as b')\
&nbsp; &nbsp; &nbsp;.insert_into('mySink')\
&nbsp; &nbsp; &nbsp;.t_env.execute("convert_job")

报错为:
TypeError: 'UserDefinedFunctionWrapper' object is not callable

Re: pyflink 嵌套使用函数出错

Posted by 小学生 <20...@qq.com>.
您好,谢谢大佬的指导,排查了下确实是get部分的udf去掉即可,感恩

Re: pyflink 嵌套使用函数出错

Posted by Xingbo Huang <hx...@gmail.com>.
Hi, 小学生

我稍微修改了一下你的code(你的from_elements那样写按理说就没法运行)
code是能够正确运行的,你可以参考一下,你去掉的是不是有问题,或者你把你修改后的代码贴上来,再一起看看
from pyflink.table import StreamTableEnvironment, DataTypes,
EnvironmentSettings, TableConfig, \
    BatchTableEnvironment
from pyflink.table.descriptors import Schema, OldCsv, FileSystem
from pyflink.table.udf import udf
from pyflink.datastream import StreamExecutionEnvironment


def test_test():
    elements = 'aaa|bbb'
    env = StreamExecutionEnvironment.get_execution_environment()
    env.set_parallelism(1)
    t_env = StreamTableEnvironment.create(env,

environment_settings=EnvironmentSettings.new_instance().use_blink_planner().build())

    @udf(input_types=[DataTypes.STRING()],
         result_type=DataTypes.ARRAY(DataTypes.STRING()))
    def split(x):
        print(x.strip().split("|"))
        return x.strip().split("|")

    # t_env.register_function("split", udf(lambda i: i.strip().split("|"),
[DataTypes.STRING()], DataTypes.ARRAY(DataTypes.STRING())))
    t_env.register_function("split", split)

    # split拆分后为一个2元数组
    def get(arr, index):
        return arr[index]

    @udf(input_types=[DataTypes.ARRAY(DataTypes.STRING())],
result_type=DataTypes.STRING())
    def convert(arr):
        return get(arr, 0)


    t_env.register_function("convert", convert)

    t_env.connect(FileSystem().path('/tmp/output')) \
        .with_format(OldCsv()
                     .field('b', DataTypes.STRING())) \
        .with_schema(Schema()
                     .field('b', DataTypes.STRING())) \
        .create_temporary_table('mySink')

    t_env.from_elements([(elements,), ], ["a"]) \
        .alias('line') \
        .select('split(line) as arr') \
        .select('convert(arr) as b') \
        .insert_into('mySink')
    t_env.execute("convert_job")


if __name__ == '__main__':
    test_test()

Best,
Xingbo

小学生 <20...@qq.com> 于2020年6月4日周四 下午3:38写道:

> 大佬好,去掉了运行还是出错,一样的错误

Re: pyflink 嵌套使用函数出错

Posted by 小学生 <20...@qq.com>.
大佬好,去掉了运行还是出错,一样的错误

Re: pyflink 嵌套使用函数出错

Posted by Xingbo Huang <hx...@gmail.com>.
Hi, 小学生。
把函数get的标签udf给去掉,它只是普通的Python函数,不要加上@udf,加上之后就不是python的函数了。只有Python的UDF你才要加上@udf

Best,
Xingbo

小学生 <20...@qq.com> 于2020年6月4日周四 下午2:46写道:

> 各位大佬好,初学pyflink,有一个问题需要帮忙解决下。
>
>
> 代码为:
> from pyflink.table import StreamTableEnvironment, DataTypes,
> EnvironmentSettings,TableConfig,BatchTableEnvironment
> from pyflink.table.descriptors import Schema, OldCsv, FileSystem
> from pyflink.table.udf import udf
> from pyflink.datastream import StreamExecutionEnvironment
> elements = 'aaa|bbb'
> env = StreamExecutionEnvironment.get_execution_environment()
> env.set_parallelism(1)
> t_env = StreamTableEnvironment.create(env,
> environment_settings=EnvironmentSettings.new_instance().use_blink_planner().build())
>
>
> @udf(input_types=[DataTypes.STRING()],
> &nbsp; &nbsp; &nbsp;result_type=DataTypes.ARRAY(DataTypes.STRING()))
> def split(x):
> &nbsp; &nbsp; return x.strip().split("|")
> # t_env.register_function("split", udf(lambda i: i.strip().split("|"),
> [DataTypes.STRING()], DataTypes.ARRAY(DataTypes.STRING())))
> t_env.register_function("split", split)
> #split拆分后为一个2元数组
> @udf(input_types=[DataTypes.ARRAY(DataTypes.STRING()), DataTypes.INT()],
> &nbsp; &nbsp; &nbsp;result_type=DataTypes.STRING())
> def get(array, index):
> &nbsp; &nbsp; return array[index]
>
>
> @udf(input_types=[DataTypes.ARRAY(DataTypes.STRING())],
> &nbsp; &nbsp; &nbsp;result_type=DataTypes.STRING())
> def convert(array):
> &nbsp; &nbsp; return get(array, 0)
>
>
> t_env.register_function("convert", convert)
>
>
> t_env.connect(FileSystem().path('/tmp/output')) \
> &nbsp; &nbsp; .with_format(OldCsv()
> &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.field('b',
> DataTypes.STRING())) \
> &nbsp; &nbsp; .with_schema(Schema()
> &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp; &nbsp;.field('b',
> DataTypes.STRING())) \
> &nbsp; &nbsp; .create_temporary_table('mySink')
>
>
> t_env.from_elements(elements)\
> &nbsp; &nbsp; &nbsp;.alias('line')\
> &nbsp; &nbsp; &nbsp;.select('split(line)')\
> &nbsp; &nbsp; &nbsp;.alias('array')\
> &nbsp; &nbsp; &nbsp;.select('convert(array) as b')\
> &nbsp; &nbsp; &nbsp;.insert_into('mySink')\
> &nbsp; &nbsp; &nbsp;.t_env.execute("convert_job")
>
> 报错为:
> TypeError: 'UserDefinedFunctionWrapper' object is not callable