<p>This document serves as the <strong>complete</strong> definition of Google's coding standards for
source code in the Java™ Programming Language. A Java source file is described as being <em>in
Google Style</em> if and only if it adheres to the rules herein.</p><p>Like other programming style guides, the issues covered span not only aesthetic issues of
formatting, but other types of conventions or coding standards as well. However, this document
focuses primarily on the <strong>hard-and-fast rules</strong> that we follow universally, and
avoids giving <em>advice</em> that isn't clearly enforceable (whether by human or tool).
<p>In this document, unless otherwise clarified:</p><ol><li>The term <em>class</em> is used inclusively to mean an "ordinary" class, enum class,
interface or annotation type (<codeclass="prettyprint lang-java">@interface</code>).</li><li>The term <em>comment</em> always refers to <em>implementation</em> comments. We do not
use the phrase "documentation comments", instead using the common term "Javadoc."</li></ol><p>Other "terminology notes" will appear occasionally throughout the document.</p><aname="s1.2-guide-notes"/>
<p>Aside from the line terminator sequence, the <strong>ASCII horizontal space
character</strong> (<strong>0x20</strong>) is the only whitespace character that appears
anywhere in a source file. This implies that:</p><ol><li>All other whitespace characters in string and character literals are escaped.</li><li>Tab characters are <strong>not</strong> used for indentation.</li></ol><aname="s2.3.2-special-escape-sequences"/>
<h4>2.3.2 Special escape sequences <ahref="#s2.3.2-special-escape-sequences"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<p>For any character that has a special escape sequence
(<codeclass="prettyprint lang-java">\b</code>,
<codeclass="prettyprint lang-java">\t</code>,
<codeclass="prettyprint lang-java">\n</code>,
<codeclass="prettyprint lang-java">\f</code>,
<codeclass="prettyprint lang-java">\r</code>,
<codeclass="prettyprint lang-java">\"</code>,
<codeclass="prettyprint lang-java">\'</code> and
<codeclass="prettyprint lang-java">\\</code>), that sequence
is used rather than the corresponding octal
(e.g.<codeclass="prettyprint lang-java badcode">\012</code>) or Unicode
<p>For the remaining non-ASCII characters, either the actual Unicode character
(e.g.<codeclass="prettyprint lang-java">∞</code>) or the equivalent Unicode escape
(e.g.<codeclass="prettyprint lang-java">\u221e</code>) is used, depending only on which
makes the code <strong>easier to read and understand</strong>.</p><pclass="tip"><b>Tip:</b> in the Unicode escape case, and occasionally even when actual Unicode characters
are used, an explanatory comment can be very helpful.</p><p>Examples:</p><table><tr><th>Example</th><th>Discussion</th></tr><tr><td><codeclass="prettyprint lang-java">String unitAbbrev = "μs";</code></td><td>Best: perfectly clear even without a comment.</td></tr><tr><td><codeclass="prettyprint lang-java">String unitAbbrev = "\u03bcs"; // "μs"</code></td><td>Allowed, but there's no reason to do this.</td></tr><tr><td><codeclass="prettyprint lang-java">String unitAbbrev = "\u03bcs";
// Greek letter mu, "s"</code></td><td>Allowed, but awkward and prone to mistakes.</td></tr><tr><td><codeclass="prettyprint lang-java badcode">String unitAbbrev = "\u03bcs";</code></td><td>Poor: the reader has no idea what this is.</td></tr><tr><td><codeclass="prettyprint lang-java">return '\ufeff' + content;
// byte order mark</code></td><td>Good: use escapes for non-printable characters, and comment if necessary.</td></tr></table><pclass="tip"><b>Tip:</b> Never make your code less readable simply out of fear that some programs might
not handle non-ASCII characters properly. If that should happen, those programs
are <strong>broken</strong> and they must be <strong>fixed</strong>.</p><aname="filestructure"/><aname="s3-source-file-structure"/>
<div><p>A source file consists of, <strong>in order</strong>:</p><ol><li>License or copyright information, if present</li><li>Package statement</li><li>Import statements</li><li>Exactly one top-level class</li></ol></div><p><strong>Exactly one blank line</strong> separates each section that is present.</p><aname="s3.1-copyright-statement"/>
<h3>3.1 License or copyright information, if present <ahref="#s3.1-copyright-statement"><imgheight="21"width="21"src="javaguidelink.png"/></a></h3>
<p>If license or copyright information belongs in a file, it belongs here.</p><aname="s3.2-package-statement"/>
<code>sun</code></li></ul></li><li><code>java</code> imports</li><li><code>javax</code> imports</li></ol><p>Within a group there are no blank lines, and the imported names appear in ASCII sort
order. (<strong>Note:</strong> this is not the same as the import <em>statements</em> being in
ASCII sort order; the presence of semicolons warps the result.)</p><aname="s3.4-class-declaration"/>
<h3>3.4 Class declaration <ahref="#s3.4-class-declaration"><imgheight="21"width="21"src="javaguidelink.png"/></a></h3>
<h4>3.4.1 Exactly one top-level class declaration <ahref="#s3.4.1-one-top-level-class"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
for <em>nonempty</em> blocks and block-like constructs:</p><ul><li>No line break before the opening brace.</li><li>Line break after the opening brace.</li><li>Line break before the closing brace.</li><li>Line break after the closing brace <em>if</em> that brace terminates a statement or the body
of a method, constructor or <em>named</em> class. For example, there is <em>no</em> line break
after the brace if it is followed by <codeclass="prettyprint lang-java">else</code> or a
</p><p><strong>Exceptions:</strong></p><ol><li>Lines where obeying the column limit is not possible (for example, a long URL in Javadoc,
or a long JSNI method reference).</li><li><codeclass="prettyprint lang-java">package</code> and
<codeclass="prettyprint lang-java">import</code> statements (see Sections
3.2 <ahref="#s3.2-package-statement">Package statement</a> and
3.3 <ahref="#s3.3-import-statements">Import statements</a>).</li><li>Command lines in a comment that may be cut-and-pasted into a shell.</li></ol><aname="s4.5-line-wrapping"/>
<pclass="terminology"><b>Terminology Note:</b> When code that might otherwise legally occupy a single line is divided into
multiple lines, typically to avoid overflowing the column limit, this activity is called
<em>line-wrapping</em>.</p><p>There is no comprehensive, deterministic formula showing <em>exactly</em> how to line-wrap in
every situation. Very often there are several valid ways to line-wrap the same piece of code.</p><pclass="tip"><b>Tip:</b> extracting a method or local variable may solve the problem without the need to
<h4>4.5.1 Where to break <ahref="#s4.5.1-line-wrapping-where-to-break"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<p>The prime directive of line-wrapping is: prefer to break at a
<strong>higher syntactic level</strong>. Also:</p><ol><li>When a line is broken at a <em>non-assignment</em> operator the break comes <em>before</em>
the symbol. (Note that this is not the same practice used in Google style for other languages,
such as C++ and JavaScript.)
<ul><li>This also applies to the following "operator-like" symbols: the dot separator
(<codeclass="prettyprint lang-java">.</code>), the ampersand in type bounds
(<codeclass="prettyprint lang-java"><T extends Foo & Bar></code>), and the pipe in
catch blocks
(<codeclass="prettyprint lang-java">catch (FooException | BarException e)</code>).</li></ul></li><li>When a line is broken at an <em>assignment</em> operator the break typically comes
<em>after</em> the symbol, but either way is acceptable.
<ul><li>This also applies to the "assignment-operator-like" colon in an enhanced
<codeclass="prettyprint lang-java">for</code> ("foreach") statement.</li></ul></li><li>A method or constructor name stays attached to the open parenthesis
(<codeclass="prettyprint lang-java">(</code>) that follows it.</li><li>A comma (<codeclass="prettyprint lang-java">,</code>) stays attached to the token that
<h4>4.5.2 Indent continuation lines at least +4 spaces <ahref="#s4.5.2-line-wrapping-indent"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<p>When line-wrapping, each line after the first (each <em>continuation line</em>) is indented
at least +4 from the original line.</p><p>When there are multiple continuation lines, indentation may be varied beyond +4 as
desired. In general, two continuation lines use the same indentation level if and only if they
begin with syntactically parallel elements.</p><p>Section 4.6.3 on <ahref="#s4.6.3-horizontal-alignment">Horizontal alignment</a> addresses
the discouraged practice of using a variable number of spaces to align certain tokens with
<ul><li><spanclass="exception"><b>Exception:</b> a blank line between two consecutive fields (having no other code
between them) is optional. Such blank lines are used as needed to create <em>logical
groupings</em> of fields.</span></li></ul></li><li>Within method bodies, as needed to create <em>logical groupings</em> of statements.</li><li><em>Optionally</em> before the first member or after the last member of the class (neither
encouraged nor discouraged).</li><li>As required by other sections of this document (such as Section 3.3,
<ahref="#s3.3-import-statements">Import statements</a>).</li></ol><p><em>Multiple</em> consecutive blank lines are permitted, but never required (or encouraged).</p><aname="s4.6.2-horizontal-whitespace"/>
<p>Beyond where required by the language or other style rules, and apart from literals, comments and
Javadoc, a single ASCII space also appears in the following places <strong>only</strong>.</p><ol><li>Separating any reserved word, such as
<codeclass="prettyprint lang-java">if</code>,
<codeclass="prettyprint lang-java">for</code> or
<codeclass="prettyprint lang-java">catch</code>, from an open parenthesis
(<codeclass="prettyprint lang-java">(</code>)
that follows it on that line</li><li>Separating any reserved word, such as
<codeclass="prettyprint lang-java">else</code> or
<codeclass="prettyprint lang-java">catch</code>, from a closing curly brace
(<codeclass="prettyprint lang-java">}</code>) that precedes it on that line</li><li>Before any open curly brace
(<codeclass="prettyprint lang-java">{</code>), with two exceptions:
<ul><li><codeclass="prettyprint lang-java">@SomeAnnotation({a, b})</code> (no space is used)</li><li><codeclass="prettyprint lang-java">String[][] x = {{"foo"}};</code> (no space is required
between <codeclass="prettyprint lang-java">{{</code>, by item 8 below)</li></ul></li><li>On both sides of any binary or ternary operator. This also applies to the following
"operator-like" symbols:
<ul><li>the ampersand in a conjunctive type bound:
<codeclass="prettyprint lang-java"><T extends Foo & Bar></code></li><li>the pipe for a catch block that handles multiple exceptions:
<codeclass="prettyprint lang-java">catch (FooException | BarException e)</code></li><li>the colon (<codeclass="prettyprint lang-java">:</code>) in an enhanced
<codeclass="prettyprint lang-java">for</code> ("foreach") statement</li></ul></li><li>After <codeclass="prettyprint lang-java">,:;</code> or the closing parenthesis
(<codeclass="prettyprint lang-java">)</code>) of a cast</li><li>On both sides of the double slash (<codeclass="prettyprint lang-java">//</code>) that
begins an end-of-line comment. Here, multiple spaces are allowed, but not required.</li><li>Between the type and variable of a declaration:
<codeclass="prettyprint lang-java">List<String> list</code></li><li><em>Optional</em> just inside both braces of an array initializer
<ul><li><codeclass="prettyprint lang-java">new int[] {5, 6}</code> and
<codeclass="prettyprint lang-java">new int[] { 5, 6 }</code> are both valid</li></ul></li></ol><pclass="note"><b>Note:</b> this rule never requires or forbids additional space at the start or end of a
line, only <em>interior</em> space.</p><aname="s4.6.3-horizontal-alignment"/>
<h4>4.6.3 Horizontal alignment: never required <ahref="#s4.6.3-horizontal-alignment"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<pclass="terminology"><b>Terminology Note:</b> Horizontal alignment is the practice of adding a variable number of
additional spaces in your code with the goal of making certain tokens appear directly below certain
other tokens on previous lines.</p><p>This practice is permitted, but is <strong>never required</strong> by Google Style. It is not
even required to <em>maintain</em> horizontal alignment in places where it was already used.</p><p>Here is an example without alignment, then using alignment:</p><preclass="prettyprint lang-java">
private int x; // this is fine
private Color color; // this too
private int x; // permitted, but future edits
private Color color; // may leave it unaligned
</pre><pclass="tip"><b>Tip:</b> Alignment can aid readability, but it creates problems for future maintenance.
Consider a future change that needs to touch just one line. This change may leave the
formerly-pleasing formatting mangled, and that is <strong>allowed</strong>. More often it prompts
the coder (perhaps you) to adjust whitespace on nearby lines as well, possibly triggering a
cascading series of reformattings. That one-line change now has a "blast radius." This can at worst
result in pointless busywork, but at best it still corrupts version history information, slows down
reviewers and exacerbates merge conflicts.</p><aname="parentheses"/><aname="s4.7-grouping-parentheses"/>
<p>After each comma that follows an enum constant, a line-break is optional.</p><p>An enum class with no methods and no documentation on its constants may optionally be formatted
as if it were an array initializer:</p><preclass="prettyprint lang-java">
</pre><p>Since enum classes <em>are classes</em>, all other rules for formatting classes apply.</p><aname="localvariables"/><aname="s4.8.2-variable-declarations"/>
<h5>4.8.3.1 Array initializers: can be "block-like" <ahref="#s4.8.3.1-array-initializers"><imgheight="21"width="21"src="javaguidelink.png"/></a></h5>
<p>Any array initializer may <em>optionally</em> be formatted as if it were a "block-like
construct." For example, the following are all valid (<strong>not</strong> an exhaustive
list):</p><preclass="prettyprint lang-java">
new int[] { new int[] {
0, 1, 2, 3 0,
} 1,
2,
new int[] { 3,
0, 1, }
2, 3
} new int[]
{0, 1, 2, 3}
</pre><aname="s4.8.3.2-array-declarations"/>
<h5>4.8.3.2 No C-style array declarations <ahref="#s4.8.3.2-array-declarations"><imgheight="21"width="21"src="javaguidelink.png"/></a></h5>
<p>The square brackets form a part of the <em>type</em>, not the variable:
<codeclass="prettyprint lang-java">String[] args</code>, not
<p>As with any other block, the contents of a switch block are indented +2.</p><p>After a switch label, a newline appears, and the indentation level is increased +2, exactly as
if a block were being opened. The following switch label returns to the previous indentation
level, as if a block had been closed.</p><aname="fallthrough"/><aname="s4.8.4.2-switch-fall-through"/>
<p>Block comments are indented at the same level as the surrounding code. They may be in
<codeclass="prettyprint lang-java">/* ... */</code> style or
<codeclass="prettyprint lang-java">// ...</code> style. For multi-line
<codeclass="prettyprint lang-java">/* ... */</code> comments, subsequent lines must start with
<code>*</code> aligned with the <code>*</code> on the previous line.</p><preclass="prettyprint lang-java">
/*
* This is // And so /* Or you can
* okay. // is this. * even do this. */
*/
</pre><p>Comments are not enclosed in boxes drawn with asterisks or other characters.</p><pclass="tip"><b>Tip:</b> When writing multi-line comments, use the
<codeclass="prettyprint lang-java">/* ... */</code> style if you want automatic code formatters to
re-wrap the lines when necessary (paragraph-style). Most formatters don't re-wrap lines in
<p>Class names are written in <ahref="#s5.3-camel-case">UpperCamelCase</a>.</p><p>Class names are typically nouns or noun phrases. For example,
<codeclass="prettyprint lang-java">Character</code> or
<codeclass="prettyprint lang-java">ImmutableList</code>. Interface names may also be nouns or
noun phrases (for example, <codeclass="prettyprint lang-java">List</code>), but may sometimes be
adjectives or adjective phrases instead (for example,
<codeclass="prettyprint lang-java">Readable</code>).</p><p>There are no specific rules or even well-established conventions for naming annotation types.</p><p><em>Test</em> classes are named starting with the name of the class they are testing, and ending
<p>Constant names use <codeclass="prettyprint lang-java">CONSTANT_CASE</code>: all uppercase
letters, with words separated by underscores. But what <em>is</em> a constant, exactly?</p><p>Every constant is a static final field, but not all static final fields are constants. Before
choosing constant case, consider whether the field really <em>feels like</em> a constant. For
example, if any of that instance's observable state can change, it is almost certainly not a
constant. Merely <em>intending</em> to never mutate the object is generally not
<p>Parameter names are written in <ahref="#s5.3-camel-case">lowerCamelCase</a>.</p><p>One-character parameter names should be avoided.</p><aname="s5.2.7-local-variable-names"/>
<h4>5.2.7 Local variable names <ahref="#s5.2.7-local-variable-names"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<p>Local variable names are written in <ahref="#s5.3-camel-case">lowerCamelCase</a>, and can be
abbreviated more liberally than other types of names.</p><p>However, one-character names should be avoided, except for temporary and looping variables.</p><p>Even when final and immutable, local variables are not considered to be constants, and should not
be styled as constants.</p><aname="s5.2.8-type-variable-names"/>
<h4>5.2.8 Type variable names <ahref="#s5.2.8-type-variable-names"><imgheight="21"width="21"src="javaguidelink.png"/></a></h4>
<p>Each type variable is named in one of two styles:</p><ul><li>A single capital letter, optionally followed by a single numeral (such as
<p>Sometimes there is more than one reasonable way to convert an English phrase into camel case,
such as when acronyms or unusual constructs like "IPv6" or "iOS" are present. To improve
predictability, Google Style specifies the following (nearly) deterministic scheme.</p><p>Beginning with the prose form of the name:</p><ol><li>Convert the phrase to plain ASCII and remove any apostrophes. For example, "Müller's
algorithm" might become "Muellers algorithm".</li><li>Divide this result into words, splitting on spaces and any remaining punctuation (typically
hyphens).
<ul><li><em>Recommended:</em> if any word already has a conventional camel-case appearance in common
usage, split this into its constituent parts (e.g., "AdWords" becomes "adwords"). Note
that a word such as "iOS" is not really in camel case <em>per se</em>; it defies <em>any</em>
convention, so this recommendation does not apply.</li></ul></li><li>Now lowercase <em>everything</em> (including acronyms), then uppercase only the first
character of:
<ul><li>... each word, to yield <em>upper camel case</em>, or</li><li>... each word except the first, to yield <em>lower camel case</em></li></ul></li><li>Finally, join all the words into a single identifier.</li></ol><p>Note that the casing of the original words is almost entirely disregarded. Examples:</p><table><tr><th>Prose form</th><th>Correct</th><th>Incorrect</th></tr><tr><td>"XML HTTP request"</td><td><codeclass="prettyprint lang-java">XmlHttpRequest</code></td><td><codeclass="prettyprint lang-java badcode">XMLHTTPRequest</code></td></tr><tr><td>"new customer ID"</td><td><codeclass="prettyprint lang-java">newCustomerId</code></td><td><codeclass="prettyprint lang-java badcode">newCustomerID</code></td></tr><tr><td>"inner stopwatch"</td><td><codeclass="prettyprint lang-java">innerStopwatch</code></td><td><codeclass="prettyprint lang-java badcode">innerStopWatch</code></td></tr><tr><td>"supports IPv6 on iOS?"</td><td><codeclass="prettyprint lang-java">supportsIpv6OnIos</code></td><td><codeclass="prettyprint lang-java badcode">supportsIPv6OnIOS</code></td></tr><tr><td>"YouTube importer"</td><td><codeclass="prettyprint lang-java">YouTubeImporter</code><br/><codeclass="prettyprint lang-java">YoutubeImporter</code>*</td><td/></tr></table><p>*Acceptable, but not recommended.</p><pclass="note"><b>Note:</b> Some words are ambiguously hyphenated in the English language: for example
"nonempty" and "non-empty" are both correct, so the method names
<codeclass="prettyprint lang-java">checkNonempty</code> and
<codeclass="prettyprint lang-java">checkNonEmpty</code> are likewise both correct.</p><aname="s6-programming-practices"/>
<h3>6.2 Caught exceptions: not ignored <ahref="#s6.2-caught-exceptions"><imgheight="21"width="21"src="javaguidelink.png"/></a></h3>
<p>Except as noted below, it is very rarely correct to do nothing in response to a caught
exception. (Typical responses are to log it, or if it is considered "impossible", rethrow it as an
<codeclass="prettyprint lang-java">AssertionError</code>.)</p><p>When it truly is appropriate to take no action whatsoever in a catch block, the reason this is
justified is explained in a comment.</p><preclass="prettyprint lang-java">
try {
int i = Integer.parseInt(response);
return handleNumericResponse(i);
} catch (NumberFormatException ok) {
// it's not numeric; that's fine, just continue
}
return handleTextResponse(response);
</pre><pclass="exception"><b>Exception:</b> in tests, a caught exception may be ignored without comment <em>if</em> it is
named <codeclass="prettyprint lang-java">expected</code>. The following is a very common idiom
for ensuring that the method under test <em>does</em> throw an exception of the expected type, so
a comment is unnecessary here.</p><preclass="prettyprint lang-java">
try {
emptyStack.pop();
fail();
} catch (NoSuchElementException expected) {
}
</pre><aname="s6.3-static-members"/>
<h3>6.3 Static members: qualified using class <ahref="#s6.3-static-members"><imgheight="21"width="21"src="javaguidelink.png"/></a></h3>
<p>When a reference to a static class member must be qualified, it is qualified with that class's
name, not with a reference or expression of that class's type.</p><preclass="prettyprint lang-java">
Foo aFoo = ...;
Foo.aStaticMethod(); // good
<spanclass="badcode">aFoo.aStaticMethod();</span> // bad
<spanclass="badcode">somethingThatYieldsAFoo().aStaticMethod();</span> // very bad
<h3>6.4 Finalizers: not used <ahref="#s6.4-finalizers"><imgheight="21"width="21"src="javaguidelink.png"/></a></h3>
<p>It is <strong>extremely rare</strong> to override <codeclass="prettyprint lang-java">Object.finalize</code>.</p><pclass="tip"><b>Tip:</b> Don't do it. If you absolutely must, first read and understand
<p>Javadoc is optional for "simple, obvious" methods like
<codeclass="prettyprint lang-java">getFoo</code>, in cases where there <em>really and truly</em> is
nothing else worthwhile to say but "Returns the foo".</p><p>The test methods of a unit test class are perhaps the most common example of this exemption.
These methods can <em>usually</em> be named descriptively enough that no additional documentation is
needed.</p><pclass="tip"><b>Tip:</b> <strong>Important:</strong> it is not appropriate to cite this exception to justify
omitting relevant information that a typical reader might need to know. For example, for a method
named <codeclass="prettyprint lang-java">getCanonicalName</code>, don't omit its documentation
(with the rationale that it would say only <codeclass="prettyprint lang-java badcode">/** Returns
the canonical name. */</code>) if a typical reader may have no idea what the term "canonical name"
