Hide keyboard shortcuts

Hot-keys on this page

r m x p   toggle line displays

j k   next/prev highlighted chunk

0   (zero) top of page

1   (one) first highlighted chunk

1

2

3

4

5

6

7

8

9

10

11

12

13

14

15

16

17

18

19

20

21

22

23

24

25

26

27

28

29

30

31

32

33

34

35

36

37

38

39

40

41

42

43

44

45

46

47

48

49

50

51

52

53

54

55

56

57

58

59

60

61

62

63

64

65

66

67

68

69

70

71

72

73

74

75

76

77

78

79

80

81

82

83

84

85

86

87

88

89

90

91

92

93

94

95

96

97

98

99

100

101

102

103

104

105

106

107

108

109

110

111

112

113

114

115

116

117

118

119

120

121

122

123

124

125

126

127

128

129

130

131

132

133

134

135

136

137

138

139

140

141

142

143

144

145

146

147

148

149

150

151

152

153

154

155

156

157

158

159

160

161

162

163

164

165

166

167

168

169

170

171

172

173

174

175

176

177

178

179

180

181

182

183

184

185

186

187

188

189

190

191

192

193

194

195

196

197

198

199

200

201

202

203

204

205

206

207

208

209

210

211

# 

# Licensed to the Apache Software Foundation (ASF) under one or more 

# contributor license agreements. See the NOTICE file distributed with 

# this work for additional information regarding copyright ownership. 

# The ASF licenses this file to You under the Apache License, Version 2.0 

# (the "License"); you may not use this file except in compliance with 

# the License. You may obtain a copy of the License at 

# 

# http://www.apache.org/licenses/LICENSE-2.0 

# 

# Unless required by applicable law or agreed to in writing, software 

# distributed under the License is distributed on an "AS IS" BASIS, 

# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied. 

# See the License for the specific language governing permissions and 

# limitations under the License. 

# 

 

import os 

import shutil 

import tempfile 

 

from pyspark.sql.functions import col 

from pyspark.sql.readwriter import DataFrameWriterV2 

from pyspark.sql.types import StructType, StructField, StringType 

from pyspark.testing.sqlutils import ReusedSQLTestCase 

 

 

class ReadwriterTests(ReusedSQLTestCase): 

 

def test_save_and_load(self): 

df = self.df 

tmpPath = tempfile.mkdtemp() 

shutil.rmtree(tmpPath) 

df.write.json(tmpPath) 

actual = self.spark.read.json(tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

schema = StructType([StructField("value", StringType(), True)]) 

actual = self.spark.read.json(tmpPath, schema) 

self.assertEqual(sorted(df.select("value").collect()), sorted(actual.collect())) 

 

df.write.json(tmpPath, "overwrite") 

actual = self.spark.read.json(tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

df.write.save(format="json", mode="overwrite", path=tmpPath, 

noUse="this options will not be used in save.") 

actual = self.spark.read.load(format="json", path=tmpPath, 

noUse="this options will not be used in load.") 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

defaultDataSourceName = self.spark.conf.get("spark.sql.sources.default", 

"org.apache.spark.sql.parquet") 

self.spark.sql("SET spark.sql.sources.default=org.apache.spark.sql.json") 

actual = self.spark.read.load(path=tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

self.spark.sql("SET spark.sql.sources.default=" + defaultDataSourceName) 

 

csvpath = os.path.join(tempfile.mkdtemp(), 'data') 

df.write.option('quote', None).format('csv').save(csvpath) 

 

shutil.rmtree(tmpPath) 

 

def test_save_and_load_builder(self): 

df = self.df 

tmpPath = tempfile.mkdtemp() 

shutil.rmtree(tmpPath) 

df.write.json(tmpPath) 

actual = self.spark.read.json(tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

schema = StructType([StructField("value", StringType(), True)]) 

actual = self.spark.read.json(tmpPath, schema) 

self.assertEqual(sorted(df.select("value").collect()), sorted(actual.collect())) 

 

df.write.mode("overwrite").json(tmpPath) 

actual = self.spark.read.json(tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

df.write.mode("overwrite").options(noUse="this options will not be used in save.")\ 

.option("noUse", "this option will not be used in save.")\ 

.format("json").save(path=tmpPath) 

actual =\ 

self.spark.read.format("json")\ 

.load(path=tmpPath, noUse="this options will not be used in load.") 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

 

defaultDataSourceName = self.spark.conf.get("spark.sql.sources.default", 

"org.apache.spark.sql.parquet") 

self.spark.sql("SET spark.sql.sources.default=org.apache.spark.sql.json") 

actual = self.spark.read.load(path=tmpPath) 

self.assertEqual(sorted(df.collect()), sorted(actual.collect())) 

self.spark.sql("SET spark.sql.sources.default=" + defaultDataSourceName) 

 

shutil.rmtree(tmpPath) 

 

def test_bucketed_write(self): 

data = [ 

(1, "foo", 3.0), (2, "foo", 5.0), 

(3, "bar", -1.0), (4, "bar", 6.0), 

] 

df = self.spark.createDataFrame(data, ["x", "y", "z"]) 

 

def count_bucketed_cols(names, table="pyspark_bucket"): 

"""Given a sequence of column names and a table name 

query the catalog and return number o columns which are 

used for bucketing 

""" 

cols = self.spark.catalog.listColumns(table) 

num = len([c for c in cols if c.name in names and c.isBucket]) 

return num 

 

with self.table("pyspark_bucket"): 

# Test write with one bucketing column 

df.write.bucketBy(3, "x").mode("overwrite").saveAsTable("pyspark_bucket") 

self.assertEqual(count_bucketed_cols(["x"]), 1) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

# Test write two bucketing columns 

df.write.bucketBy(3, "x", "y").mode("overwrite").saveAsTable("pyspark_bucket") 

self.assertEqual(count_bucketed_cols(["x", "y"]), 2) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

# Test write with bucket and sort 

df.write.bucketBy(2, "x").sortBy("z").mode("overwrite").saveAsTable("pyspark_bucket") 

self.assertEqual(count_bucketed_cols(["x"]), 1) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

# Test write with a list of columns 

df.write.bucketBy(3, ["x", "y"]).mode("overwrite").saveAsTable("pyspark_bucket") 

self.assertEqual(count_bucketed_cols(["x", "y"]), 2) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

# Test write with bucket and sort with a list of columns 

(df.write.bucketBy(2, "x") 

.sortBy(["y", "z"]) 

.mode("overwrite").saveAsTable("pyspark_bucket")) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

# Test write with bucket and sort with multiple columns 

(df.write.bucketBy(2, "x") 

.sortBy("y", "z") 

.mode("overwrite").saveAsTable("pyspark_bucket")) 

self.assertSetEqual(set(data), set(self.spark.table("pyspark_bucket").collect())) 

 

def test_insert_into(self): 

df = self.spark.createDataFrame([("a", 1), ("b", 2)], ["C1", "C2"]) 

with self.table("test_table"): 

df.write.saveAsTable("test_table") 

self.assertEqual(2, self.spark.sql("select * from test_table").count()) 

 

df.write.insertInto("test_table") 

self.assertEqual(4, self.spark.sql("select * from test_table").count()) 

 

df.write.mode("overwrite").insertInto("test_table") 

self.assertEqual(2, self.spark.sql("select * from test_table").count()) 

 

df.write.insertInto("test_table", True) 

self.assertEqual(2, self.spark.sql("select * from test_table").count()) 

 

df.write.insertInto("test_table", False) 

self.assertEqual(4, self.spark.sql("select * from test_table").count()) 

 

df.write.mode("overwrite").insertInto("test_table", False) 

self.assertEqual(6, self.spark.sql("select * from test_table").count()) 

 

 

class ReadwriterV2Tests(ReusedSQLTestCase): 

def test_api(self): 

df = self.df 

writer = df.writeTo("testcat.t") 

self.assertIsInstance(writer, DataFrameWriterV2) 

self.assertIsInstance(writer.option("property", "value"), DataFrameWriterV2) 

self.assertIsInstance(writer.options(property="value"), DataFrameWriterV2) 

self.assertIsInstance(writer.using("source"), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy("id"), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(col("id")), DataFrameWriterV2) 

self.assertIsInstance(writer.tableProperty("foo", "bar"), DataFrameWriterV2) 

 

def test_partitioning_functions(self): 

import datetime 

from pyspark.sql.functions import years, months, days, hours, bucket 

 

df = self.spark.createDataFrame( 

[(1, datetime.datetime(2000, 1, 1), "foo")], 

("id", "ts", "value") 

) 

 

writer = df.writeTo("testcat.t") 

 

self.assertIsInstance(writer.partitionedBy(years("ts")), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(months("ts")), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(days("ts")), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(hours("ts")), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(bucket(11, "id")), DataFrameWriterV2) 

self.assertIsInstance(writer.partitionedBy(bucket(11, col("id"))), DataFrameWriterV2) 

self.assertIsInstance( 

writer.partitionedBy(bucket(3, "id"), hours(col("ts"))), DataFrameWriterV2 

) 

 

 

if __name__ == "__main__": 

import unittest 

from pyspark.sql.tests.test_readwriter import * # noqa: F401 

 

try: 

import xmlrunner # type: ignore[import] 

testRunner = xmlrunner.XMLTestRunner(output='target/test-reports', verbosity=2) 

except ImportError: 

testRunner = None 

unittest.main(testRunner=testRunner, verbosity=2)