README
Pure Kotlin CSV Reader/Writer.
- easy to setup
- use DSL so easy to read
- in Java, we always need to close file. but it's boilerplate code and not friendly for non-JVM user.
- provide interfaces which automatically close file without being aware.
- Kotlin Multiplatform projects support.
// Gradle Kotlin DSL
implementation("com.github.doyaaaaaken:kotlin-csv-jvm:1.9.2") // for JVM platform
implementation("com.github.doyaaaaaken:kotlin-csv-js:1.9.2") // for Kotlin JS platform
// Gradle Groovy DSL
implementation 'com.github.doyaaaaaken:kotlin-csv-jvm:1.9.2' // for JVM platform
implementation 'com.github.doyaaaaaken:kotlin-csv-js:1.9.2' // for Kotlin JS platform
<dependency>
<groupId>com.github.doyaaaaaken</groupId>
<artifactId>kotlin-csv-jvm</artifactId>
<version>1.9.2</version>
</dependency>
<dependency>
<groupId>com.github.doyaaaaaken</groupId>
<artifactId>kotlin-csv-js</artifactId>
<version>1.9.2</version>
</dependency>
@file:DependsOn("com.github.doyaaaaaken:kotlin-csv-jvm:1.9.2") // for JVM platform
@file:DependsOn("com.github.doyaaaaaken:kotlin-csv-js:1.9.2") // for Kotlin JS platform
Simple case
You can read csv file from
String, java.io.File or java.io.InputStream object.
No need to do any I/O handling. (No need to call use, close and flush method.)// read from `String`
val csvData: String = "a,b,c\nd,e,f"
val rows: List<List<String>> = csvReader().readAll(csvData)
// read from `java.io.File`
val file: File = File("test.csv")
val rows: List<List<String>> = csvReader().readAll(file)
Read with header
val csvData: String = "a,b,c\nd,e,f"
val rows: List<Map<String, String>> = csvReader().readAllWithHeader(csvData)
println(rows) //[{a=d, b=e, c=f}]
Read as
SequenceSequence type allows to execute lazily.
It starts to process each rows before reading all row data.csvReader().open("test1.csv") {
readAllAsSequence().forEach { row: List<String> ->
//Do something
println(row) //[a, b, c]
}
}
csvReader().open("test2.csv") {
readAllWithHeaderAsSequence().forEach { row: Map<String, String> ->
//Do something
println(row) //{id=1, name=doyaaaaaken}
}
}
NOTE:
readAllAsSequence and readAllWithHeaderAsSequence methods can only be called within the open lambda block. The input stream is closed after the open lambda block.Read line by line
If you want to handle line-by-line, you can do it by using
open method. Use open method and then use readNext method inside nested block to read row.csvReader().open("test.csv") {
readNext()
}
Read in a
Suspending FunctioncsvReader().openAsync("test.csv") {
val container = mutalbeListOf<List<String>>()
delay(100) //other suspending task
readAllAsSequence().asFlow().collect { row ->
delay(100) // other suspending task
container.add(row)
}
}
Note:
openAsync can be and only be accessed through a coroutine or another suspending functionCustomize
When you create CsvReader, you can choose read options:
// this is tsv reader's option
val tsvReader = csvReader {
charset = "ISO_8859_1"
quoteChar = '"'
delimiter = '\t'
escapeChar = '\\'
}
Option | default value | description |
|---|---|---|
logger | no-op | Logger instance for logging debug information at runtime. |
charset | UTF-8 | |
quoteChar | " | Character used to quote fields. |
delimiter | , | Character used as delimiter between each field.
Use "\t" if reading TSV file. |
escapeChar | " | Character to escape quote inside field string.
Normally, you don't have to change this option.
See detail comment on ICsvReaderContext. |
skipEmptyLine | false | Whether to skip or error out on empty lines. |
autoRenameDuplicateHeaders | false | Whether to auto rename duplicate headers or throw an exception. |
skipMissMatchedRow | false | Deprecated. Replace with appropriate values in excessFieldsRowBehaviour and insufficientFieldsRowBehaviour, e.g. both set to IGNORE. Whether to skip an invalid row. If ignoreExcessCols is true, only rows with less than the expected number of columns will be skipped. |
excessFieldsRowBehaviour | ERROR | Behaviour to use when a row has more fields (columns) than expected. ERROR (default), IGNORE (skip the row) or TRIM (remove the excess fields at the end of the row to match the expected number of fields). |
insufficientFieldsRowBehaviour | ERROR | Behaviour to use when a row has fewer fields (columns) than expected. ERROR (default), IGNORE (skip the row) or EMPTY_STRING (replace missing fields with an empty string). |
Simple case
You can start writing csv in one line, no need to do any I/O handling (No need to call
use, close and flush method.):val rows = listOf(listOf("a", "b", "c"), listOf("d", "e", "f"))
csvWriter().writeAll(rows, "test.csv")
// if you'd append data on the tail of the file, assign `append = true`.
csvWriter().writeAll(rows, "test.csv", append = true)
// You can also write into OutpusStream.
csvWriter().writeAll(rows, File("test.csv").outputStream())
You can also write a csv file line by line by
open method:val row1 = listOf("a", "b", "c")
val row2 = listOf("d", "e", "f")
csvWriter().open("test.csv") {
writeRow(row1)
writeRow(row2)
writeRow("g", "h", "i")
writeRows(listOf(row1, row2))
}
Write in a
Suspending Functionval rows = listOf(listOf("a", "b", "c"), listOf("d", "e", "f")).asSequence()
csvWriter().openAsync(testFileName) {
delay(100) //other suspending task
rows.asFlow().collect {
delay(100) // other suspending task
writeRow(it)
}
}
Write as String
val rows = listOf(listOf("a", "b", "c"), listOf("d", "e", "f"))
val csvString: String = csvWriter().writeAllAsString(rows) //a,b,c\r\nd,e,f\r\n
long-running write (manual control for file close)
If you want to close a file writer manually for performance reasons (e.g. streaming scenario), you can use
openAndGetRawWriter and get a raw CsvFileWriter.
DO NOT forget to close the writer!val row1 = listOf("a", "b", "c")
@OptIn(KotlinCsvExperimental::class)
val writer = csvWriter().openAndGetRawWriter("test.csv")
writer.writeRow(row1)
writer.close()
Customize
When you create a CsvWriter, you can choose write options.
val writer = csvWriter {
charset = "ISO_8859_1"
delimiter = '\t'
nullCode = "NULL"
lineTerminator = "\n"
outputLastLineTerminator = true
quote {
mode = WriteQuoteMode.ALL
char = '\''
}
}
Option | default value | description |
|---|---|---|
charset | UTF-8 | |
delimiter | , | Character used as delimiter between each fields.
Use "\t" if reading TSV file. |
nullCode | (empty string) | Character used when a written field is null value. |
lineTerminator | \r | Character used as line terminator. |
outputLastLineTerminator | true | Output line break at the end of file or not. |
prependBOM | false | Output BOM (Byte Order Mark) at the beginning of file or not. |
quote.char | " | Character to quote each fields. |
quote.mode | CANONICAL | Quote mode.
- CANONICAL: Not quote normally, but quote special characters (quoteChar, delimiter, line feed). This is the specification of CSV.
- ALL: Quote all fields.
- NON_NUMERIC: Quote non-numeric fields. (ex. 1,"a",2.3) |
Documents
Libraries which use kotlin-csv
Contributions, issues and feature requests are welcome! If you have questions, ask away in Kotlin Slack's
kotlin-csv room.Give a ⭐️ if this project helped you!
Last modified 1mo ago