Kotlin Files and I/O

import java.io.File

val file = File("notes.txt")
file.writeText("Buy milk\nCall Sam\n")     // creates or replaces the file
file.appendText("Book dentist\n")          // adds to the end
val text = file.readText()                 // the whole file as one String
val lines = file.readLines()               // a List<String>, one per line

• writeText replaces the whole file; appendText adds to it.
• These functions use UTF-8 by default, which handles every language's characters.
• A relative path like "notes.txt" is found from the folder where the program runs.
• They're perfect for small and medium files. For very large ones, read line by line instead.

readLines() loads every line into memory at once, which is a problem for a 2 GB log file. Read lazily instead:

file.forEachLine { line -> if ("ERROR" in line) println(line) }

val errorCount = file.useLines { lines -> lines.count { "ERROR" in it } }

useLines gives you a sequence of lines, so you can use operators like filter, map and count on it, and it closes the file when the block finishes. Only the current line is held in memory, however big the file is.

Anything that holds a file open, such as a writer or reader, must be closed, or data may never reach the disk. The use function closes it automatically, even when an exception is thrown:

File("report.csv").bufferedWriter().use { out ->
    out.write("name,score")
    out.newLine()
}

Useful File members: exists(), isDirectory, length(), delete(), mkdirs() (create a folder and its parents), name, extension and listFiles(). File("photos").walk() visits every file in a folder tree, which is handy for scripts that tidy or count files.

File operations can fail: the file is missing, locked or unreadable. Missing files throw FileNotFoundException, a kind of IOException. Check with exists() first, or wrap the operation in try/catch or runCatching from the exceptions lesson.

To read structured text such as CSV, split each line and build a data class:

val products = file.readLines().drop(1).map { line ->
    val (name, price) = line.split(",")
    Product(name, price.toDouble())
}

For real CSV with quoted commas, or for JSON, use a library such as kotlinx.serialization.

java.io.File has been around since the earliest days of Java. Newer code often uses java.nio.file.Path instead, and Kotlin's kotlin.io.path package gives it the same friendly functions you've seen on File:

import kotlin.io.path.*

val dir = Path("data")
dir.createDirectories()                  // no error if it already exists
val notes = dir / "notes.txt"            // data/notes.txt
notes.writeText("Buy milk\n")
notes.appendText("Call Sam\n")
println(notes.readLines())               // [Buy milk, Call Sam]
println(notes.name)                      // notes.txt
println(dir.listDirectoryEntries("*.txt").map { it.name })

The / operator joins path parts with the right separator for the operating system, which is safer than gluing strings together with "/" or "\\". With File, File("data", "notes.txt") does the same job.

One difference to know: reading a missing file through a Path throws NoSuchFileException, while File.readText() throws FileNotFoundException. Both are kinds of IOException, so catching IOException handles either. You can switch between the two styles at any time with file.toPath() and path.toFile().

If a program crashes halfway through writeText, the file can end up half-written, and the old contents are already gone. A common safeguard is to write to a temporary file next to the real one, then move it into place in a single step:

val tmp = dir / "notes.txt.tmp"
tmp.writeText(newContent)
tmp.moveTo(notes, overwrite = true)

Until the move happens, the original file is untouched, so readers see either the complete old version or the complete new one. Settings files and saved games are good candidates for this pattern.