Strings & Basistypen

Two everyday things read very differently in Kotlin: strings and basic types. String templates end the "..." + variable + "..." concatenation ritual, and triple-quoted strings handle multi-line text cleanly. And Kotlin's type system has no primitives - Int, Long, Double, Boolean are all real classes, unifying the type hierarchy while the compiler still uses efficient JVM primitives under the hood. Both are small daily wins for a Java developer.

String templates

In Java you build strings with + concatenation or String.format. Kotlin has string templates - embed a variable or expression directly with $:

val name = "Ada"
val age = 36
println("$name is $age years old")               // simple variable: $name
println("Next year she'll be ${age + 1}")        // an expression: ${...}
println("Name length: ${name.length}")           // any expression in ${}

$variable inserts a simple variable; ${expression} inserts any expression. Compare the Java: System.out.println(name + " is " + age + " years old") - the template version reads like the sentence it produces, with no + noise. This is the same idea as Java's newer text blocks and formatted, but inline and used constantly.

Multi-line and raw strings

Triple quotes create a raw string that spans lines and doesn't process escape sequences - ideal for JSON, SQL, or regex where Java's \\ escaping is painful:

val json = """
    {
        "name": "$name",
        "age": $age
    }
""".trimIndent()                                  // trimIndent() removes the common leading indentation

val regex = """\d{3}-\d{4}""".toRegex()           // no double-backslash escaping

trimIndent() is the idiomatic companion - it strips the leading whitespace you added for code readability, so the string's actual content starts at the left margin. (Templates still work inside triple-quoted strings, as $name shows.)

No primitives: a unified type system

Java splits types into primitives (int, double, boolean - lowercase, not objects) and their boxed wrappers (Integer, Double, Boolean - objects). This split causes autoboxing surprises and means primitives can't be used as generic type arguments (List<int> is illegal; you need List<Integer>).

Kotlin has only the class types - Int, Long, Short, Byte, Double, Float, Boolean, Char:

val count: Int = 42          // a real type, capital I - no separate 'int'
val big: Long = 10_000_000_000
val ratio: Double = 3.14
val flag: Boolean = true
val list: List<Int> = listOf(1, 2, 3)   // Int works directly as a generic argument

There's no int-vs-Integer distinction to trip over. But here's the clever part: the compiler still uses JVM primitives under the hood wherever possible. An Int local variable compiles to a JVM int (fast, no boxing); it only boxes to an Integer when it must (e.g. as a nullable Int? or a generic type argument), exactly when Java would. So you get a clean, unified type model in the source and efficient primitive performance in the bytecode - the best of both.

Numeric literals and conversions

A couple of details Java developers should note:

• Underscores in literals for readability: 1_000_000, 0xFF_EC_DE.
• No implicit widening. Kotlin won't silently convert an Int to a Long; you convert explicitly with .toLong(), .toDouble(), etc. This is stricter than Java (which auto-widens) and prevents a class of subtle bugs, at the cost of an occasional explicit .toX().