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
|
/*
* 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.
*/
package org.apache.spark.sql.execution.datasources.csv
import java.nio.charset.{Charset, StandardCharsets}
import org.apache.hadoop.conf.Configuration
import org.apache.hadoop.fs.FileStatus
import org.apache.hadoop.io.{LongWritable, Text}
import org.apache.hadoop.mapred.TextInputFormat
import org.apache.hadoop.mapreduce._
import org.apache.spark.rdd.RDD
import org.apache.spark.sql.SparkSession
import org.apache.spark.sql.catalyst.InternalRow
import org.apache.spark.sql.catalyst.expressions.JoinedRow
import org.apache.spark.sql.catalyst.expressions.codegen.GenerateUnsafeProjection
import org.apache.spark.sql.execution.datasources._
import org.apache.spark.sql.sources._
import org.apache.spark.sql.types.{StringType, StructField, StructType}
import org.apache.spark.util.SerializableConfiguration
/**
* Provides access to CSV data from pure SQL statements.
*/
class DefaultSource extends FileFormat with DataSourceRegister {
override def shortName(): String = "csv"
override def toString: String = "CSV"
override def hashCode(): Int = getClass.hashCode()
override def equals(other: Any): Boolean = other.isInstanceOf[DefaultSource]
override def inferSchema(
sparkSession: SparkSession,
options: Map[String, String],
files: Seq[FileStatus]): Option[StructType] = {
val csvOptions = new CSVOptions(options)
// TODO: Move filtering.
val paths = files.filterNot(_.getPath.getName startsWith "_").map(_.getPath.toString)
val rdd = baseRdd(sparkSession, csvOptions, paths)
val firstLine = findFirstLine(csvOptions, rdd)
val firstRow = new LineCsvReader(csvOptions).parseLine(firstLine)
val header = if (csvOptions.headerFlag) {
firstRow
} else {
firstRow.zipWithIndex.map { case (value, index) => s"C$index" }
}
val parsedRdd = tokenRdd(sparkSession, csvOptions, header, paths)
val schema = if (csvOptions.inferSchemaFlag) {
CSVInferSchema.infer(parsedRdd, header, csvOptions.nullValue)
} else {
// By default fields are assumed to be StringType
val schemaFields = header.map { fieldName =>
StructField(fieldName.toString, StringType, nullable = true)
}
StructType(schemaFields)
}
Some(schema)
}
override def prepareWrite(
sparkSession: SparkSession,
job: Job,
options: Map[String, String],
dataSchema: StructType): OutputWriterFactory = {
val conf = job.getConfiguration
val csvOptions = new CSVOptions(options)
csvOptions.compressionCodec.foreach { codec =>
CompressionCodecs.setCodecConfiguration(conf, codec)
}
new CSVOutputWriterFactory(csvOptions)
}
override def buildReader(
sparkSession: SparkSession,
dataSchema: StructType,
partitionSchema: StructType,
requiredSchema: StructType,
filters: Seq[Filter],
options: Map[String, String]): (PartitionedFile) => Iterator[InternalRow] = {
val csvOptions = new CSVOptions(options)
val headers = requiredSchema.fields.map(_.name)
val conf = new Configuration(sparkSession.sessionState.hadoopConf)
val broadcastedConf = sparkSession.sparkContext.broadcast(new SerializableConfiguration(conf))
(file: PartitionedFile) => {
val lineIterator = {
val conf = broadcastedConf.value.value
new HadoopFileLinesReader(file, conf).map { line =>
new String(line.getBytes, 0, line.getLength, csvOptions.charset)
}
}
CSVRelation.dropHeaderLine(file, lineIterator, csvOptions)
val unsafeRowIterator = {
val tokenizedIterator = new BulkCsvReader(lineIterator, csvOptions, headers)
val parser = CSVRelation.csvParser(dataSchema, requiredSchema.fieldNames, csvOptions)
tokenizedIterator.flatMap(parser(_).toSeq)
}
// Appends partition values
val fullOutput = requiredSchema.toAttributes ++ partitionSchema.toAttributes
val joinedRow = new JoinedRow()
val appendPartitionColumns = GenerateUnsafeProjection.generate(fullOutput, fullOutput)
unsafeRowIterator.map { dataRow =>
appendPartitionColumns(joinedRow(dataRow, file.partitionValues))
}
}
}
private def baseRdd(
sparkSession: SparkSession,
options: CSVOptions,
inputPaths: Seq[String]): RDD[String] = {
readText(sparkSession, options, inputPaths.mkString(","))
}
private def tokenRdd(
sparkSession: SparkSession,
options: CSVOptions,
header: Array[String],
inputPaths: Seq[String]): RDD[Array[String]] = {
val rdd = baseRdd(sparkSession, options, inputPaths)
// Make sure firstLine is materialized before sending to executors
val firstLine = if (options.headerFlag) findFirstLine(options, rdd) else null
CSVRelation.univocityTokenizer(rdd, header, firstLine, options)
}
/**
* Returns the first line of the first non-empty file in path
*/
private def findFirstLine(options: CSVOptions, rdd: RDD[String]): String = {
if (options.isCommentSet) {
val comment = options.comment.toString
rdd.filter { line =>
line.trim.nonEmpty && !line.startsWith(comment)
}.first()
} else {
rdd.filter { line =>
line.trim.nonEmpty
}.first()
}
}
private def readText(
sparkSession: SparkSession,
options: CSVOptions,
location: String): RDD[String] = {
if (Charset.forName(options.charset) == StandardCharsets.UTF_8) {
sparkSession.sparkContext.textFile(location)
} else {
val charset = options.charset
sparkSession.sparkContext
.hadoopFile[LongWritable, Text, TextInputFormat](location)
.mapPartitions(_.map(pair => new String(pair._2.getBytes, 0, pair._2.getLength, charset)))
}
}
}
|
