Standard rules
Annotation formatting¶
Multiple annotations should be on a separate line than the annotated declaration; annotations with parameters should each be on separate lines; annotations should be followed by a space
// A single annotation (without parameters) is allowed on same line as annotated construct
@FunctionalInterface class FooBar {
@JvmField var foo: String
@Test fun bar() {}
}
// A class or function parameter may have a single annotation with parameter(s) on the same line
class Foo(@Path("fooId") val fooId: String)
class Bar(
@NotNull("fooId") val fooId: String,
@NotNull("bar") bar: String
)
// Multiple annotations (without parameters) are allowed on the same line
@Foo @Bar
class FooBar {
@Foo @Bar
var foo: String
@Foo @Bar
fun bar() {}
}
// An array of annotations (without parameters) is allowed on same line as annotated construct
@[Foo Bar] class FooBar2 {
@[Foo Bar] var foo: String
@[Foo Bar] fun bar() {}
}
// An annotation with parameter(s) is not allowed on same line as annotated construct
@Suppress("Unused") class FooBar {
@Suppress("Unused") var foo: String
@Suppress("Unused") fun bar() {}
}
// Multiple annotation on same line as annotated construct are not allowed
@Foo @Bar class FooBar {
@Foo @Bar var foo: String
@Foo @Bar fun bar() {}
}
Rule-id: annotation
(standard
rule set)
Blank line before declarations¶
Requires a blank line before any class or function declaration. No blank line is required between the class signature and the first declaration in the class. In a similar way, a blank line is required before any list of top level or class properties. No blank line is required before local properties or between consecutive properties.
Rule id: blank-line-before-declaration
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Block comment initial star alignment¶
Lines in a block comment which (exclusive the indentation) start with a *
should have this *
aligned with the *
in the opening of the block comment.
Rule id: block-comment-initial-star-alignment
(standard
rule set)
Discouraged comment location¶
Detect discouraged comment locations (no autocorrect).
Note
Kotlin allows comments to be placed almost everywhere. As this can lead to code which is hard to read, most of them will never be used in practice. Ideally each rule takes comments at all possible locations into account. Sometimes this is really hard and not worth the effort. By explicitly forbidding such comment locations, the development of those rules becomes a bit easier.
Rule id: discouraged-comment-location
(standard
rule set)
Enum entry¶
Enum entry names should be uppercase underscore-separated or upper camel-case separated.
Rule id: enum-entry-name-case
(standard
rule set)
File name¶
A file containing only one visible (e.g. non-private) class, and visible declarations related to that class only, should be named according to that element. The same applies if the file does not contain a visible class but exactly one type alias or one object declaration. Otherwise, the PascalCase notation should be used.
Rule id: filename
(standard
rule set)
Final newline¶
Ensures consistent usage of a newline at the end of each file.
This rule can be configured with .editorconfig
property insert_final_newline
.
Rule id: final-newline
(standard
rule set)
Function signature¶
Rewrites the function signature to a single line when possible (e.g. when not exceeding the max_line_length
property) or a multiline signature otherwise. In case of function with a body expression, the body expression is placed on the same line as the function signature when not exceeding the max_line_length
property.
In ktlint-official
code style, a function signature is always rewritten to a multiline signature in case the function has 2 or more parameters. This number of parameters can be set via .editorconfig
property ktlint_function_signature_rule_force_multiline_when_parameter_count_greater_or_equal_than
.
// Assume that the last allowed character is
// at the X character on the right X
fun foooooooo(
a: Any,
b: Any,
c: Any
): String {
// body
}
// Assume that the last allowed character is
// at the X character on the right X
fun bar(a: Any, b: Any, c: Any): String {
// body
}
// When wrapping of body is set to 'default'.
// Assume that the last allowed character is
// at the X character on the right X
fun f(a: Any, b: Any): String = "some-result"
.uppercase()
// When wrapping of body is set to 'multiline'
// or 'always'.
// Assume that the last allowed character is
// at the X character on the right X
fun f(a: Any, b: Any): String =
"some-result"
.uppercase()
// Assume that the last allowed character is
// at the X character on the right X
fun foooooooo(a: Any, b: Any, c: Any): String {
// body
}
// Assume that the last allowed character is
// at the X character on the right X
fun bar(
a: Any,
b: Any,
c: Any
): String {
// body
}
// When wrapping of body is set to 'default'.
// Assume that the last allowed character is
// at the X character on the right X
fun f(a: Any, b: Any): String =
"some-result"
.uppercase()
// When wrapping of body is set to 'multiline'
// or 'always'.
// Assume that the last allowed character is
// at the X character on the right X
fun f(a: Any, b: Any): String = "some-result"
.uppercase()
Rule id: function-signature
(standard
rule set)
If else bracing¶
If at least one branch of an if-else statement or an if-else-if statement is wrapped between curly braces then all branches should be wrapped between braces.
Rule id: if-else-bracing
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Import ordering¶
Ensures that imports are ordered consistently (see Import Layouts for configuration).
Rule id: import-ordering
(standard
rule set)
Indentation¶
Indentation formatting - respects .editorconfig
indent_size
with no continuation indent (see EditorConfig section for more).
Note
This rule handles indentation for many different language constructs which can not be summarized with a few examples. See the unit tests for more details.
Rule id: indent
(standard
rule set)
Naming¶
Class naming¶
Enforce naming of class and objects.
Note
Functions in files which import a class from package org.junit.jupiter.api
are considered to be test functions and are allowed to have a name specified between backticks and do not need to adhere to the normal naming convention. Although, the Kotlin coding conventions does not allow this explicitly for class identifiers, ktlint
does allow it.
This rule can also be suppressed with the IntelliJ IDEA inspection suppression ClassName
.
Rule id: class-naming
(standard
rule set)
Function naming¶
Enforce naming of function.
Note
Functions in files which import a class from package org.junit
, org.testng
or kotlin.test
are considered to be test functions. Functions in such classes are allowed to have underscores in the name. Or function names can be specified between backticks and do not need to adhere to the normal naming convention.
This rule can also be suppressed with the IntelliJ IDEA inspection suppression FunctionName
.
Rule id: function-naming
(standard
rule set)
Package name¶
Validates that the package name matches the regular expression [a-z][a-zA-Z\d]*(\.[a-z][a-zA-Z\d]*)*
.
Rule id: package-name
(standard
rule set)
Property naming¶
Enforce naming of property.
Note
This rule can not reliably detect all situations in which incorrect property naming is used. So it only detects in which it is certain that naming is incorrect.
val foo1 = Foo() // In case developer want to communicate that Foo is mutable
val FOO1 = Foo() // In case developer want to communicate that Foo is deeply immutable
const val FOO_BAR = "FOO-BAR" // By definition deeply immutable
var foo2: Foo = Foo() // By definition not immutable
class Bar {
val foo1 = Foo() // In case developer want to communicate that Foo is mutable
val FOO1 = Foo() // In case developer want to communicate that Foo is deeply immutable
const val FOO_BAR = "FOO-BAR" // By definition deeply immutable
var foo2: Foo = Foo() // By definition not immutable
// Backing property
private val _elementList = mutableListOf<Element>()
val elementList: List<Element>
get() = _elementList
}
const val fooBar = "FOO-BAR" // By definition deeply immutable
var FOO2: Foo = Foo() // By definition not immutable
class Bar {
val FOO_BAR = "FOO-BAR" // Class properties always start with lowercase, const is not allowed
// Incomplete backing property as public property 'elementList1' is missing
private val _elementList1 = mutableListOf<Element>()
// Invalid backing property as '_elementList2' is not a private property
val _elementList2 = mutableListOf<Element>()
val elementList2: List<Element>
get() = _elementList2
}
This rule can also be suppressed with the IntelliJ IDEA inspection suppression PropertyName
.
Rule id: property-naming
(standard
rule set)
No blank lines in list¶
Disallow blank lines to be used in lists before the first element, between elements, and after the last element.
Super type
Type argument list
Type constraint list
class BiAdapter<C : RecyclerView.ViewHolder, V1 : C, V2 : C, out A1, out A2>(
val adapter1: A1,
val adapter2: A2
) : RecyclerView.Adapter<C>()
where A1 : RecyclerView.Adapter<V1>, A1 : ComposableAdapter.ViewTypeProvider,
A2 : RecyclerView.Adapter<V2>, A2 : ComposableAdapter.ViewTypeProvider {
// body
}
class BiAdapter<C : RecyclerView.ViewHolder, V1 : C, V2 : C, out A1, out A2>(
val adapter1: A1,
val adapter2: A2
) : RecyclerView.Adapter<C>()
where
A1 : RecyclerView.Adapter<V1>, A1 : ComposableAdapter.ViewTypeProvider,
A2 : RecyclerView.Adapter<V2>, A2 : ComposableAdapter.ViewTypeProvider
{
// body
}
Type parameter list
Value argument list
Value parameter list
Rule id: no-blank-line-in-list
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
No consecutive comments¶
Consecutive comments are disallowed in following cases: - Any mix of a consecutive kdoc, a block comment or an EOL comment unless separated by a blank line in between - Consecutive KDocs (even when separated by a blank line) - Consecutive block comments (even when separated by a blank line)
Consecutive EOL comments are always allowed as they are often used instead of a block comment.
// An EOL comment
// may be followed by another EOL comment
val foo = "foo"
// Different comment types (including KDoc) may be consecutive ..
/*
* ... but do need to be separated by a blank line ...
*/
/**
* ... but a KDoc can not be followed by an EOL or a block comment or another KDoc
*/
fun bar() = "bar"
Rule id: no-consecutive-comments
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
No empty file¶
A kotlin (script) file should not be empty. It needs to contain at least one declaration. Files only contain a package and/or import statements are as of that disallowed.
Rule id: no-empty-file
No empty first line at start in class body¶
Detect blank lines at start of a class body.
Rule id: no-empty-first-line-in-class-body
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
No single line block comment¶
A single line block comment should be replaced with an EOL comment when possible.
Rule id: no-single-line-block-comment
(standard
rule set)
Ktlint-suppression rule¶
The ktlint-disable
and ktlint-enable
directives are no longer supported as of ktlint version 0.50.0
. This rule migrates the directives to Suppress or SuppressWarnings annotations.
Identifiers in the @Suppress and @SuppressWarnings annotations to suppress ktlint rules are checked for validity and autocorrected when possible.
/* ktlint-disable standard:no-wildcard-imports */
class FooBar {
val foo = "some longggggggggggggggggggg text" // ktlint-disable standard:max-line-length
fun bar() =
listOf(
/* ktlint-disable standard:no-multi-spaces */
"1 One",
"10 Ten",
"100 Hundred",
/* ktlint-enable standard:no-multi-spaces */
)
}
Rule id: ktlint-suppression
(standard
rule set)
Note
This rule can not be disabled in the .editorconfig
.
Max line length¶
Ensures that lines do not exceed the given length of .editorconfig
property max_line_length
(see EditorConfig section for more). This rule does not apply in a number of situations. For example, in the case a line exceeds the maximum line length due to a comment that disables ktlint rules then that comment is being ignored when validating the length of the line. The .editorconfig
property ktlint_ignore_back_ticked_identifier
can be set to ignore identifiers which are enclosed in backticks, which for example is very useful when you want to allow longer names for unit tests.
// Assume that the last allowed character is
// at the X character on the right X
// Lines below are accepted although the max
// line length is exceeded.
package com.toooooooooooooooooooooooooooo.long
import com.tooooooooooooooooooooooooooooo.long
val foo =
"""
fooooooooooooooooooooooooooooooooooooooooo
"""
@Test
fun `Test description which is toooooooooooo long`() {
}
Rule id: max-line-length
(standard
rule set)
Modifier order¶
Consistent order of modifiers
Rule id: modifier-order
(standard
rule set)
Multiline if-else¶
Braces required for multiline if/else statements.
Rule id: multiline-if-else
(standard
rule set)
No blank lines before }
¶
No blank lines before }
.
Rule id: no-blank-line-before-rbrace
(standard
rule set)
No blank lines in chained method calls¶
Rule id: no-blank-lines-in-chained-method-calls
(standard
rule set)
No consecutive blank lines¶
Rule id: no-consecutive-blank-lines
(standard
rule set)
No empty ({}
) class bodies¶
Rule id: no-empty-class-body
(standard
rule set)
No leading empty lines in method blocks¶
Rule id: no-empty-first-line-in-method-block
(standard
rule set)
No line break after else¶
Disallows line breaks after the else keyword if that could lead to confusion, for example:
Rule id: no-line-break-after-else
(standard
rule set)
No line break before assignment¶
When a line is broken at an assignment (=
) operator the break comes after the symbol.
Rule id: no-line-break-before-assignment
(standard
rule set)
No multi spaces¶
Except in indentation and in KDoc's it is not allowed to have multiple consecutive spaces.
Rule id: no-multi-spaces
(standard
rule set)
No semicolons¶
No semicolons (unless used to separate multiple statements on the same line).
Rule id: no-semi
(standard
rule set)
No trailing whitespaces¶
Rule id: no-trailing-spaces
(standard
rule set)
No Unit
as return type¶
The Unit
type is not allowed as return type of a function.
returns (fun fn {}
instead of fun fn: Unit {}
)
Rule id: no-unit-return
(standard
rule set)
No unused imports¶
Warning
This rule is not able to detect all unused imports as mentioned in this issue comment.
Rule id: no-unused-imports
(standard
rule set)
No wildcard imports¶
No wildcard imports except imports listed in .editorconfig
property ij_kotlin_packages_to_use_import_on_demand
.
Warning
In case property ij_kotlin_packages_to_use_import_on_demand
is not explicitly set, it allows wildcards imports like java.util.*
by default to keep in sync with IntelliJ IDEA behavior. To disallow all wildcard imports, add property below to your .editorconfig
:
Rule id: no-wildcard-imports
(standard
rule set)
Spacing¶
Angle bracket spacing¶
No spaces around angle brackets when used for typing.
Rule id: spacing-around-angle-brackets
(standard
rule set)
Annotation spacing¶
Annotations should be separated by a single line break.
Rule id: annotation-spacing
(standard
rule set)
Blank line between declarations with annotations¶
Declarations with annotations should be separated by a blank line.
Rule id: spacing-between-declarations-with-annotations
(standard
rule set)
Blank line between declaration with comments¶
Declarations with comments should be separated by a blank line.
Rule id: spacing-between-declarations-with-comments
(standard
rule set)
Colon spacing¶
Consistent spacing around colon.
Rule id: colon-spacing
(standard
rule set)
Comma spacing¶
Consistent spacing around comma.
Rule id: comma-spacing
(standard
rule set)
Comment spacing¶
The end of line comment sign //
should be preceded and followed by exactly a space.
Rule id: comment-spacing
(standard
rule set)
Curly spacing¶
Consistent spacing around curly braces.
Rule id: curly-spacing
(standard
rule set)
Dot spacing¶
Consistent spacing around dots.
Rule id: dot-spacing
(standard
rule set)
Double colon spacing¶
No spaces around ::
.
Rule id: double-colon-spacing
(standard
rule set)
Function return type spacing¶
Consistent spacing around the function return type.
Rule id: function-return-type-spacing
(standard
rule set)
Function start of body spacing¶
Consistent spacing before start of function body.
Rule id: function-start-of-body-spacing
(standard
rule set)
Function type reference spacing¶
Consistent spacing in the type reference before a function.
Rule id: function-type-reference-spacing
(standard
rule set)
Fun keyword spacing¶
Consistent spacing after the fun keyword.
Rule id: fun-keyword-spacing
(standard
rule set)
Kdoc wrapping¶
A KDoc comment should start and end on a line that does not contain any other element.
Rule id: kdoc-wrapping
(standard
rule set)
Keyword spacing¶
Consistent spacing around keywords.
Rule id: keyword-spacing
(standard
rule set)
Modifier list spacing¶
Consistent spacing between modifiers in and after the last modifier in a modifier list.
Rule id: modifier-list-spacing
(standard
rule set)
Nullable type spacing¶
No spaces in a nullable type.
Rule id: nullable-type-spacing
(standard
rule set)
Operator spacing¶
Consistent spacing around operators.
Rule id: op-spacing
(standard
rule set)
Parameter list spacing¶
Consistent spacing inside the parameter list.
Rule id: parameter-list-spacing
(standard
rule set)
Parenthesis spacing¶
Consistent spacing around parenthesis.
Rule id: paren-spacing
(standard
rule set)
Range spacing¶
Consistent spacing around range operators.
Rule id: range-spacing
(standard
rule set)
Spacing between function name and opening parenthesis¶
Consistent spacing between function name and opening parenthesis.
Rule id: spacing-between-function-name-and-opening-parenthesis
(standard
rule set)
Try catch finally spacing¶
Enforce consistent spacing in try { .. } catch { .. } finally { .. }
.
Rule id: try-catch-finally-spacing
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Type argument list spacing¶
Spacing before and after the angle brackets of a type argument list.
Rule id: type-argument-list-spacing
(standard
rule set)
Type parameter list spacing¶
Spacing after a type parameter list in function and class declarations.
Rule id: type-parameter-list-spacing
(standard
rule set)
Unary operator spacing¶
No spaces around unary operators.
Rule id: unary-op-spacing
(standard
rule set)
String template¶
Consistent string templates ($v
instead of ${v}
, ${p.v}
instead of ${p.v.toString()}
)
Rule id: string-template
(standard
rule set)
String template indent¶
Enforce consistent string template indentation for multiline string templates which are post-fixed with .trimIndent()
. The opening and closing """
are placed on separate lines and the indentation of the content of the template is aligned with the """
.
Rule id: string-template-indent
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Trailing comma on call site¶
Consistent removal (default) or adding of trailing commas on call site.
Important
KtLint uses the IntelliJ IDEA .editorconfig
property ij_kotlin_allow_trailing_comma_on_call_site
to configure the rule. When this property is enabled, KtLint enforces the usage of the trailing comma at call site while IntelliJ IDEA default formatter only allows to use the trailing comma but leaves it to the developer's discretion to actually use it (or not). KtLint values consistent formatting more than a per-situation decision.
Note
In KtLint 0.48.x the default value for using the trailing comma on call site has been changed to true
except when codestyle android
is used.
Although the Kotlin coding conventions leaves it to the developer's discretion to use trailing commas on the call site, it also states that usage of trailing commas has several benefits:
- It makes version-control diffs cleaner – as all the focus is on the changed value.
- It makes it easy to add and reorder elements – there is no need to add or delete the comma if you manipulate elements.
- It simplifies code generation, for example, for object initializers. The last element can also have a comma.
Note
Trailing comma on call site is automatically disabled if the Wrapping rule (or, before version 0.45.0
, the Indentation rule) is disabled or not loaded. Because it cannot provide proper formatting with unwrapped calls. (see dependencies).
Rule id: trailing-comma-on-call-site
(standard
rule set)
Trailing comma on declaration site¶
Consistent removal (default) or adding of trailing commas on declaration site.
Important
KtLint uses the IntelliJ IDEA .editorconfig
property ij_kotlin_allow_trailing_comma
to configure the rule. When this property is enabled, KtLint enforces the usage of the trailing comma at declaration site while IntelliJ IDEA default formatter only allows to use the trailing comma but leaves it to the developer's discretion to actually use it (or not). KtLint values consistent formatting more than a per-situation decision.
Note
In KtLint 0.48.x the default value for using the trailing comma on declaration site has been changed to true
except when codestyle android
is used.
The Kotlin coding conventions encourages the usage of trailing commas on the declaration site, but leaves it to the developer's discretion to use trailing commas on the call site. But next to this, it also states that usage of trailing commas has several benefits:
- It makes version-control diffs cleaner – as all the focus is on the changed value.
- It makes it easy to add and reorder elements – there is no need to add or delete the comma if you manipulate elements.
- It simplifies code generation, for example, for object initializers. The last element can also have a comma.
Note
Trailing comma on declaration site is automatically disabled if the Wrapping rule (or, before version 0.45.0
, the Indentation rule) is disabled or not loaded. Because it cannot provide proper formatting with unwrapped declarations. (see dependencies).
Rule id: trailing-comma-on-declaration-site
(standard
rule set)
Unnecessary parenthesis before trailing lambda¶
An empty parentheses block before a lambda is redundant.
Rule id: unnecessary-parentheses-before-trailing-lambda
(standard
rule set)
Wrapping¶
Argument list wrapping¶
All arguments should be on the same line, or every argument should be on a separate line.
Rule-id: argument-list-wrapping
(standard
rule set)
Chain wrapping¶
When wrapping chained calls .
, ?.
and ?:
should be placed on the next line
Rule id: chain-wrapping
(standard
rule set)
Comment wrapping¶
A block comment should start and end on a line that does not contain any other element.
Rule id: comment-wrapping
(standard
rule set)
Content receiver wrapping¶
Wraps the content receiver list to a separate line regardless of maximum line length. If the maximum line length is configured and is exceeded, wrap the context receivers and if needed its projection types to separate lines.
// ALways wrap regardless of whether max line length is set
context(Foo)
fun fooBar()
// Wrap each context receiver to a separate line when the
// entire context receiver list does not fit on a single line
context(
Fooooooooooooooooooo1,
Foooooooooooooooooooooooooooooo2
)
fun fooBar()
// Wrap each context receiver to a separate line when the
// entire context receiver list does not fit on a single line.
// Also, wrap each of it projection types in case a context
// receiver does not fit on a single line after it has been
// wrapped.
context(
Foooooooooooooooo<
Foo,
Bar
>
)
fun fooBar()
// Should be wrapped regardless of whether max line length is set
context(Foo) fun fooBar()
// Should be wrapped when the entire context receiver list does not
// fit on a single line
context(Fooooooooooooooooooo1, Foooooooooooooooooooooooooooooo2)
fun fooBar()
// Should be wrapped when the entire context receiver list does not
// fit on a single line. Also, it should wrap each of it projection
// type in case a context receiver does not fit on a single line
// after it has been wrapped.
context(Foooooooooooooooo<Foo, Bar>)
fun fooBar()
Rule id: context-receiver-wrapping
(standard
rule set)
Enum wrapping¶
An enum should be a single line, or each enum entry has to be placed on a separate line. In case the enumeration contains enum entries and declarations those are to be separated by a blank line.
Rule id: enum-wrapping
(standard
rule set)
If else wrapping¶
A single line if-statement should be kept simple. It may contain no more than one else-branch. The branches may not be wrapped in a block.
Rule id: if-else-wrapping
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Multiline expression wrapping¶
Multiline expression on the right hand side of an expression are forced to start on a separate line. Expressions in return statement are excluded as that would result in a compilation error.
Rule id: multiline-expression-wrapping
(standard
rule set)
Note
This rule is only run when ktlint_code_style
is set to ktlint_official
or when the rule is enabled explicitly.
Parameter list wrapping¶
When class/function signature doesn't fit on a single line, each parameter must be on a separate line
Rule id: parameter-list-wrapping
(standard
rule set)
Parameter wrapping¶
When a function or class parameter doesn't fit on a single line, wrap the type or value to a separate line
Rule id: parameter-wrapping
(standard
rule set)
Property wrapping¶
When a property doesn't fit on a single line, wrap the type or value to a separate line
Rule id: property-wrapping
(standard
rule set)
Statement wrapping¶
A function, class/object body or other block body statement has to be placed on different line than the braces of the body block.
Rule id: statement-wrapping
Wrapping¶
Inserts missing newlines (for example between parentheses of a multi-line function call).
Rule id: wrapping
(standard
rule set)