mirror of
https://github.com/datahub-project/datahub.git
synced 2025-11-03 20:27:50 +00:00
33 lines
921 B
Python
33 lines
921 B
Python
from pyspark.sql import SparkSession
|
|
|
|
DATA_DIR = "../resources/data"
|
|
TEST_CASE_NAME = "PythonHdfsIn2HdfsOut2"
|
|
|
|
spark = SparkSession \
|
|
.builder \
|
|
.appName(TEST_CASE_NAME) \
|
|
.enableHiveSupport() \
|
|
.getOrCreate()
|
|
|
|
df1 = spark.read.option("header", "true").csv(DATA_DIR + "/in1.csv")
|
|
df2 = spark.read.option("header", "true").csv(DATA_DIR + "/in2.csv")
|
|
df1.createOrReplaceTempView("v1")
|
|
df2.createOrReplaceTempView("v2")
|
|
|
|
df = spark.sql("select v1.c1 as a, v1.c2 as b, v2.c1 as c, v2.c2 as d from v1 join v2 on v1.id = v2.id")
|
|
|
|
#InsertIntoHadoopFsRelationCommand
|
|
df.write.mode('overwrite').csv(DATA_DIR + "/" + TEST_CASE_NAME+"/out1.csv")
|
|
|
|
df_o = spark.sql("select v1.c1 as a1, v1.c2 as b1, v2.c1 as c1, v2.c2 as d1 from v1 join v2 on v1.id = v2.id")
|
|
#InsertIntoHadoopFsRelationCommand
|
|
df_o.write.mode('overwrite').csv(DATA_DIR + "/" + TEST_CASE_NAME+"/out2.csv")
|
|
|
|
spark.sparkContext.stop()
|
|
|
|
spark.stop()
|
|
|
|
|
|
|
|
|