Gitiles
Code Review Sign In
android-review.linaro.org / platform / external / google-styleguide / refs/heads/android14-d1-s4-release / . / javaguide.html
blob: bbdbc1c92dbf59343c7973cb09ddd09de4478e9b [file] [log] [blame]
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]1<!DOCTYPE html>
2<html>
3<head>
4<meta http-equiv="content-type" content="text/html; charset=UTF-8">
5<title>Google Java Style Guide</title>
6<link rel="stylesheet" type="text/css" href="javaguide.css">
7<script language="javascript" src="include/styleguide.js"></script>
8<link rel="shortcut icon" type="image/x-icon" href="https://www.google.com/favicon.ico" />
9<script src="https://cdn.rawgit.com/google/code-prettify/master/loader/run_prettify.js"></script>
10</head>
11<body onload="initStyleGuide();">
12<div id="content">
13<h1>Google Java Style Guide</h1>
14<div class="vertical_toc" id="tocDiv"></div>
15
16<div class="main_body">
17
18<h2 id="s1-introduction">1 Introduction</h2>
19
20<p>This document serves as the <strong>complete</strong> definition of Google's coding standards for
21source code in the Java&#8482; Programming Language. A Java source file is described as being <em>in
22Google Style</em> if and only if it adheres to the rules herein.</p>
23
24<p>Like other programming style guides, the issues covered span not only aesthetic issues of
25formatting, but other types of conventions or coding standards as well. However, this document
26focuses primarily on the <strong>hard-and-fast rules</strong> that we follow universally, and
27avoids giving <em>advice</em> that isn't clearly enforceable (whether by human or tool).
28</p>
29
30
31
32<h3 id="s1.1-terminology">1.1 Terminology notes</h3>
33
34<p>In this document, unless otherwise clarified:</p>
35
36<ol>
37 <li>The term <em>class</em> is used inclusively to mean an "ordinary" class, enum class,
38 interface or annotation type (<code class="prettyprint lang-java">@interface</code>).</li>
39
40 <li>The term <em>member</em> (of a class) is used inclusively to mean a nested class, field,
41 method, <em>or constructor</em>; that is, all top-level contents of a class except initializers
42 and comments.
43
44 </li><li>The term <em>comment</em> always refers to <em>implementation</em> comments. We do not
45 use the phrase "documentation comments", instead using the common term "Javadoc."</li>
46</ol>
47
48<p>Other "terminology notes" will appear occasionally throughout the document.</p>
49
50<h3 id="s1.2-guide-notes">1.2 Guide notes</h3>
51
52<p>Example code in this document is <strong>non-normative</strong>. That is, while the examples
53are in Google Style, they may not illustrate the <em>only</em> stylish way to represent the
54code. Optional formatting choices made in examples should not be enforced as rules.</p>
55
56
57<h2 id="s2-source-file-basics">2 Source file basics</h2>
58
59<h3 id="s2.1-file-name">2.1 File name</h3>
60
61<p>The source file name consists of the case-sensitive name of the top-level class it contains
62(of which there is <a href="#s3.4.1-one-top-level-class">exactly one</a>), plus the
63<code>.java</code> extension.</p>
64
65<h3 id="s2.2-file-encoding">2.2 File encoding: UTF-8</h3>
66
67<p>Source files are encoded in <strong>UTF-8</strong>.</p>
68
69<h3 id="s2.3-special-characters">2.3 Special characters</h3>
70
71<h4 id="s2.3.1-whitespace-characters">2.3.1 Whitespace characters</h4>
72
73<p>Aside from the line terminator sequence, the <strong>ASCII horizontal space
74character</strong> (<strong>0x20</strong>) is the only whitespace character that appears
75anywhere in a source file. This implies that:</p>
76
77<ol>
78 <li>All other whitespace characters in string and character literals are escaped.</li>
79
80 <li>Tab characters are <strong>not</strong> used for indentation.</li>
81</ol>
82
83<h4 id="s2.3.2-special-escape-sequences">2.3.2 Special escape sequences</h4>
84
85<p>For any character that has a
86<a href="http://docs.oracle.com/javase/tutorial/java/data/characters.html">
87 special escape sequence</a>
88(<code class="prettyprint lang-java">\b</code>,
89<code class="prettyprint lang-java">\t</code>,
90<code class="prettyprint lang-java">\n</code>,
91<code class="prettyprint lang-java">\f</code>,
92<code class="prettyprint lang-java">\r</code>,
93<code class="prettyprint lang-java">\"</code>,
94<code class="prettyprint lang-java">\'</code> and
95<code class="prettyprint lang-java">\\</code>), that sequence
96is used rather than the corresponding octal
97(e.g.&#160;<code class="badcode">\012</code>) or Unicode
98(e.g.&#160;<code class="badcode">\u000a</code>) escape.</p>
99
100<h4 id="s2.3.3-non-ascii-characters">2.3.3 Non-ASCII characters</h4>
101
102<p>For the remaining non-ASCII characters, either the actual Unicode character
103(e.g.&#160;<code class="prettyprint lang-java">&#8734;</code>) or the equivalent Unicode escape
104(e.g.&#160;<code class="prettyprint lang-java">\u221e</code>) is used. The choice depends only on
105which makes the code <strong>easier to read and understand</strong>, although Unicode escapes
106outside string literals and comments are strongly discouraged.</p>
107
108<p class="tip"><strong>Tip:</strong> In the Unicode escape case, and occasionally even when actual
109Unicode characters are used, an explanatory comment can be very helpful.</p>
110
111<p>Examples:</p>
112
113<table>
114 <tbody><tr>
115 <th>Example</th>
116 <th>Discussion</th>
117 </tr>
118
119 <tr>
120 <td><code class="prettyprint lang-java">String unitAbbrev = "&#956;s";</code></td>
121 <td>Best: perfectly clear even without a comment.</td>
122 </tr>
123
124 <tr>
125 <td><code class="prettyprint lang-java">String unitAbbrev = "\u03bcs"; // "&#956;s"</code></td>
126 <td>Allowed, but there's no reason to do this.</td>
127 </tr>
128
129 <tr>
130 <td><code class="prettyprint lang-java">String unitAbbrev = "\u03bcs";
131 // Greek letter mu, "s"</code></td>
132 <td>Allowed, but awkward and prone to mistakes.</td>
133 </tr>
134
135 <tr>
136 <td><code class="badcode">String unitAbbrev = "\u03bcs";</code></td>
137 <td>Poor: the reader has no idea what this is.</td>
138 </tr>
139
140 <tr>
141 <td><code class="prettyprint lang-java">return '\ufeff' + content;
142 // byte order mark</code></td>
143 <td>Good: use escapes for non-printable characters, and comment if necessary.</td>
144 </tr>
145</tbody></table>
146
147<p class="tip"><strong>Tip:</strong> Never make your code less readable simply out of fear that
148some programs might not handle non-ASCII characters properly. If that should happen, those
149programs are <strong>broken</strong> and they must be <strong>fixed</strong>.</p>
150
151
152<a name="filestructure"></a>
153<h2 id="s3-source-file-structure">3 Source file structure</h2>
154
155<div>
156<p>A source file consists of, <strong>in order</strong>:</p>
157
158<ol>
159 <li>License or copyright information, if present</li>
160 <li>Package statement</li>
161 <li>Import statements</li>
162 <li>Exactly one top-level class</li>
163</ol>
164</div>
165
166<p><strong>Exactly one blank line</strong> separates each section that is present.</p>
167
168<h3 id="s3.1-copyright-statement">3.1 License or copyright information, if present</h3>
169
170<p>If license or copyright information belongs in a file, it belongs here.</p>
171
172
173
174<h3 id="s3.2-package-statement">3.2 Package statement</h3>
175
176<p>The package statement is <strong>not line-wrapped</strong>. The column limit (Section 4.4,
177<a href="#s4.4-column-limit">Column limit: 100</a>) does not apply to package statements.</p>
178
179<a name="imports"></a>
180<h3 id="s3.3-import-statements">3.3 Import statements</h3>
181
182<h4 id="s3.3.1-wildcard-imports">3.3.1 No wildcard imports</h4>
183
184<p><strong>Wildcard imports</strong>, static or otherwise, <strong>are not used</strong>.</p>
185
186<h4 id="s3.3.2-import-line-wrapping">3.3.2 No line-wrapping</h4>
187
188<p>Import statements are <strong>not line-wrapped</strong>. The column limit (Section 4.4,
189<a href="#s4.4-column-limit">Column limit: 100</a>) does not apply to import
190statements.</p>
191
192<h4 id="s3.3.3-import-ordering-and-spacing">3.3.3 Ordering and spacing</h4>
193
194<p>Imports are ordered as follows:</p>
195
196<ol>
197 <li>All static imports in a single block.</li>
198 <li>All non-static imports in a single block.</li>
199</ol>
200
201<p>If there are both static and non-static imports, a single blank line separates the two
202blocks. There are no other blank lines between import statements.</p>
203
204<p>Within each block the imported names appear in ASCII sort order. (<strong>Note:</strong>
205this is not the same as the import <em>statements</em> being in ASCII sort order, since '.'
206sorts before ';'.)</p>
207
208
209
210<h4 id="s3.3.4-import-class-not-static">3.3.4 No static import for classes</h4>
211
212<p>Static import is not used for static nested classes. They are imported with
213normal imports.</p>
214
215<h3 id="s3.4-class-declaration">3.4 Class declaration</h3>
216
217<a name="oneclassperfile"></a>
218<h4 id="s3.4.1-one-top-level-class">3.4.1 Exactly one top-level class declaration</h4>
219
220<p>Each top-level class resides in a source file of its own.</p>
221
222<a name="s3.4.2-class-member-ordering"></a>
223<h4 id="s3.4.2-ordering-class-contents">3.4.2 Ordering of class contents</h4>
224
225<p>The order you choose for the members and initializers of your class can have a great effect on
226learnability. However, there's no single correct recipe for how to do it; different classes may
227order their contents in different ways.</p>
228
229<p>What is important is that each class uses <strong><em>some</em> logical order</strong>, which its
230maintainer could explain if asked. For example, new methods are not just habitually added to the end
231of the class, as that would yield "chronological by date added" ordering, which is not a logical
232ordering.</p>
233
234
235
236<a name="overloads"></a>
237<h5 id="s3.4.2.1-overloads-never-split">3.4.2.1 Overloads: never split</h5>
238
239<p>When a class has multiple constructors, or multiple methods with the same name, these appear
240sequentially, with no other code in between (not even private members).</p>
241
242<h2 id="s4-formatting">4 Formatting</h2>
243
244<p class="terminology"><strong>Terminology Note:</strong> <em>block-like construct</em> refers to
245the body of a class, method or constructor. Note that, by Section 4.8.3.1 on
246<a href="#s4.8.3.1-array-initializers">array initializers</a>, any array initializer
247<em>may</em> optionally be treated as if it were a block-like construct.</p>
248
249<a name="braces"></a>
250<h3 id="s4.1-braces">4.1 Braces</h3>
251
252<h4 id="s4.1.1-braces-always-used">4.1.1 Braces are used where optional</h4>
253
254<p>Braces are used with
255<code class="prettyprint lang-java">if</code>,
256<code class="prettyprint lang-java">else</code>,
257<code class="prettyprint lang-java">for</code>,
258<code class="prettyprint lang-java">do</code> and
259<code class="prettyprint lang-java">while</code> statements, even when the
260body is empty or contains only a single statement.</p>
261
262<h4 id="s4.1.2-blocks-k-r-style">4.1.2 Nonempty blocks: K &amp; R style</h4>
263
264<p>Braces follow the Kernighan and Ritchie style
265("<a href="http://www.codinghorror.com/blog/2012/07/new-programming-jargon.html">Egyptian brackets</a>")
266for <em>nonempty</em> blocks and block-like constructs:</p>
267
268<ul>
269 <li>No line break before the opening brace.</li>
270
271 <li>Line break after the opening brace.</li>
272
273 <li>Line break before the closing brace.</li>
274
275 <li>Line break after the closing brace, <em>only if</em> that brace terminates a statement or
276 terminates the body of a method, constructor, or <em>named</em> class.
277 For example, there is <em>no</em> line break after the brace if it is followed by
278 <code class="prettyprint lang-java">else</code> or a comma.</li>
279</ul>
280
281<p>Examples:</p>
282
283<pre class="prettyprint lang-java">return () -&gt; {
284 while (condition()) {
285 method();
286 }
287};
288
289return new MyClass() {
290 @Override public void method() {
291 if (condition()) {
292 try {
293 something();
294 } catch (ProblemException e) {
295 recover();
296 }
297 } else if (otherCondition()) {
298 somethingElse();
299 } else {
300 lastThing();
301 }
302 }
303};
304</pre>
305
306<p>A few exceptions for enum classes are given in Section 4.8.1,
307<a href="#s4.8.1-enum-classes">Enum classes</a>.</p>
308
309<a name="emptyblocks"></a>
310<h4 id="s4.1.3-braces-empty-blocks">4.1.3 Empty blocks: may be concise</h4>
311
312<p>An empty block or block-like construct may be in K &amp; R style (as described in
313<a href="#s4.1.2-blocks-k-r-style">Section 4.1.2</a>). Alternatively, it may be closed immediately
314after it is opened, with no characters or line break in between
315(<code class="prettyprint lang-java">{}</code>), <strong>unless</strong> it is part of a
316<em>multi-block statement</em> (one that directly contains multiple blocks:
317<code class="prettyprint lang-java">if/else</code> or
318<code class="prettyprint lang-java">try/catch/finally</code>).</p>
319
320<p>Examples:</p>
321
322<pre class="prettyprint lang-java"> // This is acceptable
323 void doNothing() {}
324
325 // This is equally acceptable
326 void doNothingElse() {
327 }
328</pre>
329<pre class="prettyprint lang-java badcode"> // This is not acceptable: No concise empty blocks in a multi-block statement
330 try {
331 doSomething();
332 } catch (Exception e) {}
333</pre>
334
335<h3 id="s4.2-block-indentation">4.2 Block indentation: +2 spaces</h3>
336
337<p>Each time a new block or block-like construct is opened, the indent increases by two
338spaces. When the block ends, the indent returns to the previous indent level. The indent level
339applies to both code and comments throughout the block. (See the example in Section 4.1.2,
340<a href="#s4.1.2-blocks-k-r-style">Nonempty blocks: K &amp; R Style</a>.)</p>
341
342<h3 id="s4.3-one-statement-per-line">4.3 One statement per line</h3>
343
344<p>Each statement is followed by a line break.</p>
345
346<a name="columnlimit"></a>
347<h3 id="s4.4-column-limit">4.4 Column limit: 100</h3>
348
349<p>Java code has a column limit of 100 characters. A "character" means any Unicode code point.
350Except as noted below, any line that would exceed this limit must be line-wrapped, as explained in
351Section 4.5, <a href="#s4.5-line-wrapping">Line-wrapping</a>.
352</p>
353
354<p class="tip">Each Unicode code point counts as one character, even if its display width is
355greater or less. For example, if using
356<a href="https://en.wikipedia.org/wiki/Halfwidth_and_fullwidth_forms">fullwidth characters</a>,
357you may choose to wrap the line earlier than where this rule strictly requires.</p>
358
359<p><strong>Exceptions:</strong></p>
360
361<ol>
362 <li>Lines where obeying the column limit is not possible (for example, a long URL in Javadoc,
363 or a long JSNI method reference).</li>
364
365 <li><code class="prettyprint lang-java">package</code> and
366 <code class="prettyprint lang-java">import</code> statements (see Sections
367 3.2 <a href="#s3.2-package-statement">Package statement</a> and
368 3.3 <a href="#s3.3-import-statements">Import statements</a>).</li>
369
370 <li>Command lines in a comment that may be cut-and-pasted into a shell.</li>
371</ol>
372
373<h3 id="s4.5-line-wrapping">4.5 Line-wrapping</h3>
374
375<p class="terminology"><strong>Terminology Note:</strong> When code that might otherwise legally
376occupy a single line is divided into multiple lines, this activity is called
377<em>line-wrapping</em>.</p>
378
379<p>There is no comprehensive, deterministic formula showing <em>exactly</em> how to line-wrap in
380every situation. Very often there are several valid ways to line-wrap the same piece of code.</p>
381
382<p class="note"><strong>Note:</strong> While the typical reason for line-wrapping is to avoid
383overflowing the column limit, even code that would in fact fit within the column limit <em>may</em>
384be line-wrapped at the author's discretion.</p>
385
386<p class="tip"><strong>Tip:</strong> Extracting a method or local variable may solve the problem
387without the need to line-wrap.</p>
388
389<h4 id="s4.5.1-line-wrapping-where-to-break">4.5.1 Where to break</h4>
390
391<p>The prime directive of line-wrapping is: prefer to break at a
392<strong>higher syntactic level</strong>. Also:</p>
393
394<ol>
395 <li>When a line is broken at a <em>non-assignment</em> operator the break comes <em>before</em>
396 the symbol. (Note that this is not the same practice used in Google style for other languages,
397 such as C++ and JavaScript.)
398 <ul>
399 <li>This also applies to the following "operator-like" symbols:
400 <ul>
401 <li>the dot separator (<code class="prettyprint lang-java">.</code>)</li>
402 <li>the two colons of a method reference
403 (<code class="prettyprint lang-java">::</code>)</li>
404 <li>an ampersand in a type bound
405 (<code class="prettyprint lang-java">&lt;T extends Foo &amp; Bar&gt;</code>)</li>
406 <li>a pipe in a catch block
407 (<code class="prettyprint lang-java">catch (FooException | BarException e)</code>).</li>
408 </ul>
409 </li>
410 </ul>
411 </li>
412
413 <li>When a line is broken at an <em>assignment</em> operator the break typically comes
414 <em>after</em> the symbol, but either way is acceptable.
415 <ul>
416 <li>This also applies to the "assignment-operator-like" colon in an enhanced
417 <code class="prettyprint lang-java">for</code> ("foreach") statement.</li>
418 </ul>
419 </li>
420
421 <li>A method or constructor name stays attached to the open parenthesis
422 (<code class="prettyprint lang-java">(</code>) that follows it.</li>
423
424 <li>A comma (<code class="prettyprint lang-java">,</code>) stays attached to the token that
425 precedes it.</li>
426
427 <li>A line is never broken adjacent to the arrow in a lambda, except that a
428 break may come immediately after the arrow if the body of the lambda consists
429 of a single unbraced expression. Examples:
430<pre class="prettyprint lang-java">MyLambda&lt;String, Long, Object&gt; lambda =
431 (String label, Long value, Object obj) -&gt; {
432 ...
433 };
434
435Predicate&lt;String&gt; predicate = str -&gt;
436 longExpressionInvolving(str);
437</pre>
438 </li>
439</ol>
440
441<p class="note"><strong>Note:</strong> The primary goal for line wrapping is to have clear
442code, <em>not necessarily</em> code that fits in the smallest number of lines.</p>
443
444<a name="indentation"></a>
445<h4 id="s4.5.2-line-wrapping-indent">4.5.2 Indent continuation lines at least +4 spaces</h4>
446
447<p>When line-wrapping, each line after the first (each <em>continuation line</em>) is indented
448at least +4 from the original line.</p>
449
450<p>When there are multiple continuation lines, indentation may be varied beyond +4 as
451desired. In general, two continuation lines use the same indentation level if and only if they
452begin with syntactically parallel elements.</p>
453
454<p>Section 4.6.3 on <a href="#s4.6.3-horizontal-alignment">Horizontal alignment</a> addresses
455the discouraged practice of using a variable number of spaces to align certain tokens with
456previous lines.</p>
457
458<h3 id="s4.6-whitespace">4.6 Whitespace</h3>
459
460<h4 id="s4.6.1-vertical-whitespace">4.6.1 Vertical Whitespace</h4>
461
Liam Miller-Cushonf9347e12018-05-22 16:39:27 -0700[diff] [blame]462<p>A single blank line always appears:</p>
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]463
464<ol>
465 <li><em>Between</em> consecutive members or initializers of a class: fields, constructors,
466 methods, nested classes, static initializers, and instance initializers.
467 <ul>
468 <li><span class="exception"><strong>Exception:</strong> A blank line between two consecutive
469 fields (having no other code between them) is optional. Such blank lines are used as needed to
470 create <em>logical groupings</em> of fields.</span></li>
471 <li><span class="exception"><strong>Exception:</strong> Blank lines between enum constants are
472 covered in <a href="#s4.8.1-enum-classes">Section 4.8.1</a>.</span></li>
473 </ul>
474 </li>
475
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]476 <li>As required by other sections of this document (such as Section 3,
477 <a href="#s3-source-file-structure">Source file structure</a>, and Section 3.3,
478 <a href="#s3.3-import-statements">Import statements</a>).</li>
479</ol>
480
Liam Miller-Cushonf9347e12018-05-22 16:39:27 -0700[diff] [blame]481<p>A single blank line may also appear anywhere it improves readability, for example between
482statements to organize the code into logical subsections. A blank line before the first member or
483initializer, or after the last member or initializer of the class, is neither encouraged nor
484discouraged.
485
486</p><p><em>Multiple</em> consecutive blank lines are permitted, but never required (or encouraged).</p>
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]487
488<h4 id="s4.6.2-horizontal-whitespace">4.6.2 Horizontal whitespace</h4>
489
490<p>Beyond where required by the language or other style rules, and apart from literals, comments and
491Javadoc, a single ASCII space also appears in the following places <strong>only</strong>.</p>
492
493<ol>
494 <li>Separating any reserved word, such as
495 <code class="prettyprint lang-java">if</code>,
496 <code class="prettyprint lang-java">for</code> or
497 <code class="prettyprint lang-java">catch</code>, from an open parenthesis
498 (<code class="prettyprint lang-java">(</code>)
499 that follows it on that line</li>
500
501 <li>Separating any reserved word, such as
502 <code class="prettyprint lang-java">else</code> or
503 <code class="prettyprint lang-java">catch</code>, from a closing curly brace
504 (<code class="prettyprint lang-java">}</code>) that precedes it on that line</li>
505
506 <li>Before any open curly brace
507 (<code class="prettyprint lang-java">{</code>), with two exceptions:
508 <ul>
509 <li><code class="prettyprint lang-java">@SomeAnnotation({a, b})</code> (no space is used)</li>
510
511 <li><code class="prettyprint lang-java">String[][] x = {{"foo"}};</code> (no space is required
512 between <code class="prettyprint lang-java">{{</code>, by item 8 below)</li>
513 </ul>
514 </li>
515
516 <li>On both sides of any binary or ternary operator. This also applies to the following
517 "operator-like" symbols:
518 <ul>
519 <li>the ampersand in a conjunctive type bound:
520 <code class="prettyprint lang-java">&lt;T extends Foo &amp; Bar&gt;</code></li>
521
522 <li>the pipe for a catch block that handles multiple exceptions:
523 <code class="prettyprint lang-java">catch (FooException | BarException e)</code></li>
524
525 <li>the colon (<code class="prettyprint lang-java">:</code>) in an enhanced
526 <code class="prettyprint lang-java">for</code> ("foreach") statement</li>
527
528 <li>the arrow in a lambda expression:
529 <code class="prettyprint lang-java">(String str) -&gt; str.length()</code></li>
530 </ul>
531 but not
532
533 <ul>
534 <li>the two colons (<code class="prettyprint lang-java">::</code>) of a method reference, which
535 is written like <code class="prettyprint lang-java">Object::toString</code></li>
536 <li>the dot separator (<code class="prettyprint lang-java">.</code>), which is written like
537 <code class="prettyprint lang-java">object.toString()</code></li>
538 </ul>
539 </li>
540
541 <li>After <code class="prettyprint lang-java">,:;</code> or the closing parenthesis
542 (<code class="prettyprint lang-java">)</code>) of a cast</li>
543
544 <li>On both sides of the double slash (<code class="prettyprint lang-java">//</code>) that
545 begins an end-of-line comment. Here, multiple spaces are allowed, but not required.</li>
546
547 <li>Between the type and variable of a declaration:
548 <code class="prettyprint lang-java">List&lt;String&gt; list</code></li>
549
550 <li><em>Optional</em> just inside both braces of an array initializer
551 <ul>
552 <li><code class="prettyprint lang-java">new int[] {5, 6}</code> and
553 <code class="prettyprint lang-java">new int[] { 5, 6 }</code> are both valid</li>
554 </ul>
555 </li>
Liam Miller-Cushonf9347e12018-05-22 16:39:27 -0700[diff] [blame]556
557 <li>Between a type annotation and <code class="prettyprint lang-java">[]</code> or
558 <code class="prettyprint lang-java">...</code>.</li>
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]559</ol>
560
Liam Miller-Cushonf9347e12018-05-22 16:39:27 -0700[diff] [blame]561<p>This rule is never interpreted as requiring or forbidding additional space at the start or
562end of a line; it addresses only <em>interior</em> space.</p>
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]563
564<h4 id="s4.6.3-horizontal-alignment">4.6.3 Horizontal alignment: never required</h4>
565
566<p class="terminology"><strong>Terminology Note:</strong> <em>Horizontal alignment</em> is the
567practice of adding a variable number of additional spaces in your code with the goal of making
568certain tokens appear directly below certain other tokens on previous lines.</p>
569
570<p>This practice is permitted, but is <strong>never required</strong> by Google Style. It is not
571even required to <em>maintain</em> horizontal alignment in places where it was already used.</p>
572
573<p>Here is an example without alignment, then using alignment:</p>
574
575<pre class="prettyprint lang-java">private int x; // this is fine
576private Color color; // this too
577
578private int x; // permitted, but future edits
579private Color color; // may leave it unaligned
580</pre>
581
582<p class="tip"><strong>Tip:</strong> Alignment can aid readability, but it creates problems for
583future maintenance. Consider a future change that needs to touch just one line. This change may
584leave the formerly-pleasing formatting mangled, and that is <strong>allowed</strong>. More often
585it prompts the coder (perhaps you) to adjust whitespace on nearby lines as well, possibly
586triggering a cascading series of reformattings. That one-line change now has a "blast radius."
587This can at worst result in pointless busywork, but at best it still corrupts version history
588information, slows down reviewers and exacerbates merge conflicts.</p>
589
590<a name="parentheses"></a>
591<h3 id="s4.7-grouping-parentheses">4.7 Grouping parentheses: recommended</h3>
592
593<p>Optional grouping parentheses are omitted only when author and reviewer agree that there is no
594reasonable chance the code will be misinterpreted without them, nor would they have made the code
595easier to read. It is <em>not</em> reasonable to assume that every reader has the entire Java
596operator precedence table memorized.</p>
597
598<h3 id="s4.8-specific-constructs">4.8 Specific constructs</h3>
599
600<h4 id="s4.8.1-enum-classes">4.8.1 Enum classes</h4>
601
602<p>After each comma that follows an enum constant, a line break is optional. Additional blank
603lines (usually just one) are also allowed. This is one possibility:
604
605</p><pre class="prettyprint lang-java">private enum Answer {
606 YES {
607 @Override public String toString() {
608 return "yes";
609 }
610 },
611
612 NO,
613 MAYBE
614}
615</pre>
616
617<p>An enum class with no methods and no documentation on its constants may optionally be formatted
618as if it were an array initializer (see Section 4.8.3.1 on
619<a href="#s4.8.3.1-array-initializers">array initializers</a>).</p>
620
621<pre class="prettyprint lang-java">private enum Suit { CLUBS, HEARTS, SPADES, DIAMONDS }
622</pre>
623
624<p>Since enum classes <em>are classes</em>, all other rules for formatting classes apply.</p>
625
626<a name="localvariables"></a>
627<h4 id="s4.8.2-variable-declarations">4.8.2 Variable declarations</h4>
628
629<h5 id="s4.8.2.1-variables-per-declaration">4.8.2.1 One variable per declaration</h5>
630
631<p>Every variable declaration (field or local) declares only one variable: declarations such as
632<code class="badcode">int a, b;</code> are not used.</p>
633
634<p><strong>Exception:</strong> Multiple variable declarations are acceptable in the header of a
635<code class="prettyprint lang-java">for</code> loop.</p>
636
637<h5 id="s4.8.2.2-variables-limited-scope">4.8.2.2 Declared when needed</h5>
638
639<p>Local variables are <strong>not</strong> habitually declared at the start of their containing
640block or block-like construct. Instead, local variables are declared close to the point they are
641first used (within reason), to minimize their scope. Local variable declarations typically have
642initializers, or are initialized immediately after declaration.</p>
643
644<h4 id="s4.8.3-arrays">4.8.3 Arrays</h4>
645
646<h5 id="s4.8.3.1-array-initializers">4.8.3.1 Array initializers: can be "block-like"</h5>
647
648<p>Any array initializer may <em>optionally</em> be formatted as if it were a "block-like
649construct." For example, the following are all valid (<strong>not</strong> an exhaustive
650list):</p>
651
652<pre class="prettyprint lang-java">new int[] { new int[] {
653 0, 1, 2, 3 0,
654} 1,
655 2,
656new int[] { 3,
657 0, 1, }
658 2, 3
659} new int[]
660 {0, 1, 2, 3}
661</pre>
662
663<h5 id="s4.8.3.2-array-declarations">4.8.3.2 No C-style array declarations</h5>
664
665<p>The square brackets form a part of the <em>type</em>, not the variable:
666<code class="prettyprint lang-java">String[] args</code>, not
667<code class="badcode">String args[]</code>.</p>
668
669<h4 id="s4.8.4-switch">4.8.4 Switch statements</h4>
670
671
672
673<p class="terminology"><strong>Terminology Note:</strong> Inside the braces of a
674<em>switch block</em> are one or more <em>statement groups</em>. Each statement group consists of
675one or more <em>switch labels</em> (either <code class="prettyprint lang-java">case FOO:</code> or
676<code class="prettyprint lang-java">default:</code>), followed by one or more statements (or, for
677the <em>last</em> statement group, <em>zero</em> or more statements).</p>
678
679<h5 id="s4.8.4.1-switch-indentation">4.8.4.1 Indentation</h5>
680
681<p>As with any other block, the contents of a switch block are indented +2.</p>
682
683<p>After a switch label, there is a line break, and the indentation level is increased +2, exactly
684as if a block were being opened. The following switch label returns to the previous indentation
685level, as if a block had been closed.</p>
686
687<a name="fallthrough"></a>
688<h5 id="s4.8.4.2-switch-fall-through">4.8.4.2 Fall-through: commented</h5>
689
690<p>Within a switch block, each statement group either terminates abruptly (with a
691<code class="prettyprint lang-java">break</code>,
692<code class="prettyprint lang-java">continue</code>,
693<code class="prettyprint lang-java">return</code> or thrown exception), or is marked with a comment
694to indicate that execution will or <em>might</em> continue into the next statement group. Any
695comment that communicates the idea of fall-through is sufficient (typically
696<code class="prettyprint lang-java">// fall through</code>). This special comment is not required in
697the last statement group of the switch block. Example:</p>
698
699<pre class="prettyprint lang-java">switch (input) {
700 case 1:
701 case 2:
702 prepareOneOrTwo();
703 // fall through
704 case 3:
705 handleOneTwoOrThree();
706 break;
707 default:
708 handleLargeNumber(input);
709}
710</pre>
711
712<p>Notice that no comment is needed after <code class="prettyprint lang-java">case 1:</code>, only
713at the end of the statement group.</p>
714
715<h5 id="s4.8.4.3-switch-default">4.8.4.3 The <code>default</code> case is present</h5>
716
717<p>Each switch statement includes a <code class="prettyprint lang-java">default</code> statement
718group, even if it contains no code.</p>
719
720<p><strong>Exception:</strong> A switch statement for an <code>enum</code> type <em>may</em> omit
721the <code class="prettyprint lang-java">default</code> statement group, <em>if</em> it includes
722explicit cases covering <em>all</em> possible values of that type. This enables IDEs or other static
723analysis tools to issue a warning if any cases were missed.
724
725</p>
726
727<a name="annotations"></a>
728<h4 id="s4.8.5-annotations">4.8.5 Annotations</h4>
729
730<p>Annotations applying to a class, method or constructor appear immediately after the
731documentation block, and each annotation is listed on a line of its own (that is, one annotation
732per line). These line breaks do not constitute line-wrapping (Section
7334.5, <a href="#s4.5-line-wrapping">Line-wrapping</a>), so the indentation level is not
734increased. Example:</p>
735
736<pre class="prettyprint lang-java">@Override
737@Nullable
738public String getNameIfPresent() { ... }
739</pre>
740
741<p class="exception"><strong>Exception:</strong> A <em>single</em> parameterless annotation
742<em>may</em> instead appear together with the first line of the signature, for example:</p>
743
744<pre class="prettyprint lang-java">@Override public int hashCode() { ... }
745</pre>
746
747<p>Annotations applying to a field also appear immediately after the documentation block, but in
748this case, <em>multiple</em> annotations (possibly parameterized) may be listed on the same line;
749for example:</p>
750
751<pre class="prettyprint lang-java">@Partial @Mock DataLoader loader;
752</pre>
753
754<p>There are no specific rules for formatting annotations on parameters, local variables, or types.
755</p>
756
757<a name="comments"></a>
758<h4 id="s4.8.6-comments">4.8.6 Comments</h4>
759
760<p>This section addresses <em>implementation comments</em>. Javadoc is addressed separately in
761Section 7, <a href="#s7-javadoc">Javadoc</a>.</p>
762
763<p>Any line break may be preceded by arbitrary whitespace followed by an implementation comment.
764Such a comment renders the line non-blank.</p>
765
766<h5 id="s4.8.6.1-block-comment-style">4.8.6.1 Block comment style</h5>
767
768<p>Block comments are indented at the same level as the surrounding code. They may be in
769<code class="prettyprint lang-java">/* ... */</code> style or
770<code class="prettyprint lang-java">// ...</code> style. For multi-line
771<code class="prettyprint lang-java">/* ... */</code> comments, subsequent lines must start with
772<code>*</code> aligned with the <code>*</code> on the previous line.</p>
773
774<pre class="prettyprint lang-java">/*
775 * This is // And so /* Or you can
776 * okay. // is this. * even do this. */
777 */
778</pre>
779
780
781<p>Comments are not enclosed in boxes drawn with asterisks or other characters.</p>
782
783<p class="tip"><strong>Tip:</strong> When writing multi-line comments, use the
784<code class="prettyprint lang-java">/* ... */</code> style if you want automatic code formatters to
785re-wrap the lines when necessary (paragraph-style). Most formatters don't re-wrap lines in
786<code class="prettyprint lang-java">// ...</code> style comment blocks.</p>
787
788
789
790<a name="modifiers"></a>
791<h4 id="s4.8.7-modifiers">4.8.7 Modifiers</h4>
792
793<p>Class and member modifiers, when present, appear in the order
794recommended by the Java Language Specification:
795</p>
796
797<pre>public protected private abstract default static final transient volatile synchronized native strictfp
798</pre>
799
800<h4 id="s4.8.8-numeric-literals">4.8.8 Numeric Literals</h4>
801
802<p><code>long</code>-valued integer literals use an uppercase <code>L</code> suffix, never
803lowercase (to avoid confusion with the digit <code>1</code>). For example, <code>3000000000L</code>
804rather than <code class="badcode">3000000000l</code>.</p>
805
806<a name="naming"></a>
807<h2 id="s5-naming">5 Naming</h2>
808
809<h3 id="s5.1-identifier-names">5.1 Rules common to all identifiers</h3>
810
811<p>Identifiers use only ASCII letters and digits, and, in a small number of cases noted below,
812underscores. Thus each valid identifier name is matched by the regular expression
813<code>\w+</code> .</p>
814
Liam Miller-Cushonf9347e12018-05-22 16:39:27 -0700[diff] [blame]815<p>In Google Style, special prefixes or suffixes are <strong>not</strong> used. For example, these
816names are not Google Style: <code class="badcode">name_</code>, <code class="badcode">mName</code>,
817<code class="badcode">s_name</code> and <code class="badcode">kName</code>.</p>
Ted Osborne505ba682018-01-30 12:36:50 -0500[diff] [blame]818
819<h3 id="s5.2-specific-identifier-names">5.2 Rules by identifier type</h3>
820
821<h4 id="s5.2.1-package-names">5.2.1 Package names</h4>
822
823<p>Package names are all lowercase, with consecutive words simply concatenated together (no
824underscores). For example, <code>com.example.deepspace</code>, not
825<code class="badcode">com.example.deepSpace</code> or
826<code class="badcode">com.example.deep_space</code>.</p>
827
828<h4 id="s5.2.2-class-names">5.2.2 Class names</h4>
829
830<p>Class names are written in <a href="#s5.3-camel-case">UpperCamelCase</a>.</p>
831
832<p>Class names are typically nouns or noun phrases. For example,
833<code class="prettyprint lang-java">Character</code> or
834<code class="prettyprint lang-java">ImmutableList</code>. Interface names may also be nouns or
835noun phrases (for example, <code class="prettyprint lang-java">List</code>), but may sometimes be
836adjectives or adjective phrases instead (for example,
837<code class="prettyprint lang-java">Readable</code>).</p>
838
839<p>There are no specific rules or even well-established conventions for naming annotation types.</p>
840
841<p><em>Test</em> classes are named starting with the name of the class they are testing, and ending
842with <code class="prettyprint lang-java">Test</code>. For example,
843<code class="prettyprint lang-java">HashTest</code> or
844<code class="prettyprint lang-java">HashIntegrationTest</code>.</p>
845
846<h4 id="s5.2.3-method-names">5.2.3 Method names</h4>
847
848<p>Method names are written in <a href="#s5.3-camel-case">lowerCamelCase</a>.</p>
849
850<p>Method names are typically verbs or verb phrases. For example,
851<code class="prettyprint lang-java">sendMessage</code> or
852<code class="prettyprint lang-java">stop</code>.</p>
853
854<p>Underscores may appear in JUnit <em>test</em> method names to separate logical components of the
855name, with <em>each</em> component written in <a href="#s5.3-camel-case">lowerCamelCase</a>.
856One typical pattern is <code><i>&lt;methodUnderTest&gt;</i>_<i>&lt;state&gt;</i></code>,
857for example <code class="prettyprint lang-java">pop_emptyStack</code>. There is no One Correct
858Way to name test methods.</p>
859
860<a name="constants"></a>
861<h4 id="s5.2.4-constant-names">5.2.4 Constant names</h4>
862
863<p>Constant names use <code class="prettyprint lang-java">CONSTANT_CASE</code>: all uppercase
864letters, with each word separated from the next by a single underscore. But what <em>is</em> a
865constant, exactly?</p>
866
867<p>Constants are static final fields whose contents are deeply immutable and whose methods have no
868detectable side effects. This includes primitives, Strings, immutable types, and immutable
869collections of immutable types. If any of the instance's observable state can change, it is not a
870constant. Merely <em>intending</em> to never mutate the object is not enough. Examples:</p>
871
872<pre class="prettyprint lang-java">// Constants
873static final int NUMBER = 5;
874static final ImmutableList&lt;String&gt; NAMES = ImmutableList.of("Ed", "Ann");
875static final ImmutableMap&lt;String, Integer&gt; AGES = ImmutableMap.of("Ed", 35, "Ann", 32);
876static final Joiner COMMA_JOINER = Joiner.on(','); // because Joiner is immutable
877static final SomeMutableType[] EMPTY_ARRAY = {};
878enum SomeEnum { ENUM_CONSTANT }
879
880// Not constants
881static String nonFinal = "non-final";
882final String nonStatic = "non-static";
883static final Set&lt;String&gt; mutableCollection = new HashSet&lt;String&gt;();
884static final ImmutableSet&lt;SomeMutableType&gt; mutableElements = ImmutableSet.of(mutable);
885static final ImmutableMap&lt;String, SomeMutableType&gt; mutableValues =
886 ImmutableMap.of("Ed", mutableInstance, "Ann", mutableInstance2);
887static final Logger logger = Logger.getLogger(MyClass.getName());
888static final String[] nonEmptyArray = {"these", "can", "change"};
889</pre>
890
891<p>These names are typically nouns or noun phrases.</p>
892
893<h4 id="s5.2.5-non-constant-field-names">5.2.5 Non-constant field names</h4>
894
895<p>Non-constant field names (static or otherwise) are written
896in <a href="#s5.3-camel-case">lowerCamelCase</a>.</p>
897
898<p>These names are typically nouns or noun phrases. For example,
899<code class="prettyprint lang-java">computedValues</code> or
900<code class="prettyprint lang-java">index</code>.</p>
901
902<h4 id="s5.2.6-parameter-names">5.2.6 Parameter names</h4>
903
904<p>Parameter names are written in <a href="#s5.3-camel-case">lowerCamelCase</a>.</p>
905
906<p>One-character parameter names in public methods should be avoided.</p>
907
908<h4 id="s5.2.7-local-variable-names">5.2.7 Local variable names</h4>
909
910<p>Local variable names are written in <a href="#s5.3-camel-case">lowerCamelCase</a>.</p>
911
912<p>Even when final and immutable, local variables are not considered to be constants, and should not
913be styled as constants.</p>
914
915<h4 id="s5.2.8-type-variable-names">5.2.8 Type variable names</h4>
916
917<p>Each type variable is named in one of two styles:</p>
918
919<ul>
920 <li>A single capital letter, optionally followed by a single numeral (such as
921 <code class="prettyprint lang-java">E</code>, <code class="prettyprint lang-java">T</code>,
922 <code class="prettyprint lang-java">X</code>, <code class="prettyprint lang-java">T2</code>)
923 </li>
924
925 <li>A name in the form used for classes (see Section 5.2.2,
926 <a href="#s5.2.2-class-names">Class names</a>), followed by the capital letter
927 <code class="prettyprint lang-java">T</code> (examples:
928 <code class="prettyprint lang-java">RequestT</code>,
929 <code class="prettyprint lang-java">FooBarT</code>).</li>
930</ul>
931
932<a name="acronyms"></a>
933<a name="camelcase"></a>
934<h3 id="s5.3-camel-case">5.3 Camel case: defined</h3>
935
936<p>Sometimes there is more than one reasonable way to convert an English phrase into camel case,
937such as when acronyms or unusual constructs like "IPv6" or "iOS" are present. To improve
938predictability, Google Style specifies the following (nearly) deterministic scheme.</p>
939
940<p>Beginning with the prose form of the name:</p>
941
942<ol>
943 <li>Convert the phrase to plain ASCII and remove any apostrophes. For example, "M&#252;ller's
944 algorithm" might become "Muellers algorithm".</li>
945
946 <li>Divide this result into words, splitting on spaces and any remaining punctuation (typically
947 hyphens).
948
949 <ul>
950 <li><em>Recommended:</em> if any word already has a conventional camel-case appearance in common
951 usage, split this into its constituent parts (e.g., "AdWords" becomes "ad&#160;words"). Note
952 that a word such as "iOS" is not really in camel case <em>per se</em>; it defies <em>any</em>
953 convention, so this recommendation does not apply.</li>
954 </ul>
955 </li>
956
957 <li>Now lowercase <em>everything</em> (including acronyms), then uppercase only the first
958 character of:
959 <ul>
960 <li>... each word, to yield <em>upper camel case</em>, or</li>
961 <li>... each word except the first, to yield <em>lower camel case</em></li>
962 </ul>
963 </li>
964
965 <li>Finally, join all the words into a single identifier.</li>
966</ol>
967
968<p>Note that the casing of the original words is almost entirely disregarded. Examples:</p>
969
970<table>
971 <tbody><tr>
972 <th>Prose form</th>
973 <th>Correct</th>
974 <th>Incorrect</th>
975 </tr>
976 <tr>
977 <td>"XML HTTP request"</td>
978 <td><code class="prettyprint lang-java">XmlHttpRequest</code></td>
979 <td><code class="badcode">XMLHTTPRequest</code></td>
980 </tr>
981 <tr>
982 <td>"new customer ID"</td>
983 <td><code class="prettyprint lang-java">newCustomerId</code></td>
984 <td><code class="badcode">newCustomerID</code></td>
985 </tr>
986 <tr>
987 <td>"inner stopwatch"</td>
988 <td><code class="prettyprint lang-java">innerStopwatch</code></td>
989 <td><code class="badcode">innerStopWatch</code></td>
990 </tr>
991 <tr>
992 <td>"supports IPv6 on iOS?"</td>
993 <td><code class="prettyprint lang-java">supportsIpv6OnIos</code></td>
994 <td><code class="badcode">supportsIPv6OnIOS</code></td>
995 </tr>
996 <tr>
997 <td>"YouTube importer"</td>
998 <td><code class="prettyprint lang-java">YouTubeImporter</code><br>
999 <code class="prettyprint lang-java">YoutubeImporter</code>*</td>
1000 <td></td>
1001 </tr>
1002</tbody></table>
1003
1004<p>*Acceptable, but not recommended.</p>
1005
1006<p class="note"><strong>Note:</strong> Some words are ambiguously hyphenated in the English
1007language: for example "nonempty" and "non-empty" are both correct, so the method names
1008<code class="prettyprint lang-java">checkNonempty</code> and
1009<code class="prettyprint lang-java">checkNonEmpty</code> are likewise both correct.</p>
1010
1011
1012<h2 id="s6-programming-practices">6 Programming Practices</h2>
1013
1014<h3 id="s6.1-override-annotation">6.1 <code>@Override</code>: always used</h3>
1015
1016<p>A method is marked with the <code class="prettyprint lang-java">@Override</code> annotation
1017whenever it is legal. This includes a class method overriding a superclass method, a class method
1018implementing an interface method, and an interface method respecifying a superinterface
1019method.</p>
1020
1021<p class="exception"><strong>Exception:</strong>
1022<code class="prettyprint lang-java">@Override</code> may be omitted when the parent method is
1023<code class="prettyprint lang-java">@Deprecated</code>.</p>
1024
1025<a name="caughtexceptions"></a>
1026<h3 id="s6.2-caught-exceptions">6.2 Caught exceptions: not ignored</h3>
1027
1028<p>Except as noted below, it is very rarely correct to do nothing in response to a caught
1029exception. (Typical responses are to log it, or if it is considered "impossible", rethrow it as an
1030<code class="prettyprint lang-java">AssertionError</code>.)</p>
1031
1032<p>When it truly is appropriate to take no action whatsoever in a catch block, the reason this is
1033justified is explained in a comment.</p>
1034
1035<pre class="prettyprint lang-java">try {
1036 int i = Integer.parseInt(response);
1037 return handleNumericResponse(i);
1038} catch (NumberFormatException ok) {
1039 // it's not numeric; that's fine, just continue
1040}
1041return handleTextResponse(response);
1042</pre>
1043
1044<p class="exception"><strong>Exception:</strong> In tests, a caught exception may be ignored
1045without comment <em>if</em> its name is or begins with <code class="prettyprint lang-java">expected</code>. The
1046following is a very common idiom for ensuring that the code under test <em>does</em> throw an
1047exception of the expected type, so a comment is unnecessary here.</p>
1048
1049<pre class="prettyprint lang-java">try {
1050 emptyStack.pop();
1051 fail();
1052} catch (NoSuchElementException expected) {
1053}
1054</pre>
1055
1056<h3 id="s6.3-static-members">6.3 Static members: qualified using class</h3>
1057
1058<p>When a reference to a static class member must be qualified, it is qualified with that class's
1059name, not with a reference or expression of that class's type.</p>
1060
1061<pre class="prettyprint lang-java">Foo aFoo = ...;
1062Foo.aStaticMethod(); // good
1063<span class="badcode">aFoo.aStaticMethod();</span> // bad
1064<span class="badcode">somethingThatYieldsAFoo().aStaticMethod();</span> // very bad
1065</pre>
1066
1067<a name="finalizers"></a>
1068<h3 id="s6.4-finalizers">6.4 Finalizers: not used</h3>
1069
1070<p>It is <strong>extremely rare</strong> to override <code class="prettyprint
1071lang-java">Object.finalize</code>.</p>
1072
1073<p class="tip"><strong>Tip:</strong> Don't do it. If you absolutely must, first read and understand
1074
1075
1076 <a href="http://books.google.com/books?isbn=8131726592"><em>Effective Java</em> Item 7,</a>
1077
1078"Avoid Finalizers," very carefully, and <em>then</em> don't do it.</p>
1079
1080
1081<a name="javadoc"></a>
1082<h2 id="s7-javadoc">7 Javadoc</h2>
1083
1084
1085
1086<h3 id="s7.1-javadoc-formatting">7.1 Formatting</h3>
1087
1088<h4 id="s7.1.1-javadoc-multi-line">7.1.1 General form</h4>
1089
1090<p>The <em>basic</em> formatting of Javadoc blocks is as seen in this example:</p>
1091
1092<pre class="prettyprint lang-java">/**
1093 * Multiple lines of Javadoc text are written here,
1094 * wrapped normally...
1095 */
1096public int method(String p1) { ... }
1097</pre>
1098
1099<p>... or in this single-line example:</p>
1100
1101<pre class="prettyprint lang-java">/** An especially short bit of Javadoc. */
1102</pre>
1103
1104<p>The basic form is always acceptable. The single-line form may be substituted when the entirety
1105of the Javadoc block (including comment markers) can fit on a single line. Note that this only
1106applies when there are no block tags such as <code>@return</code>.
1107
1108</p><h4 id="s7.1.2-javadoc-paragraphs">7.1.2 Paragraphs</h4>
1109
1110<p>One blank line&#8212;that is, a line containing only the aligned leading asterisk
1111(<code>*</code>)&#8212;appears between paragraphs, and before the group of block tags if
1112present. Each paragraph but the first has <code>&lt;p&gt;</code> immediately before the first word,
1113with no space after.</p>
1114
1115<a name="s7.1.3-javadoc-at-clauses"></a>
1116
1117<h4 id="s7.1.3-javadoc-block-tags">7.1.3 Block tags</h4>
1118
1119<p>Any of the standard "block tags" that are used appear in the order <code>@param</code>,
1120<code>@return</code>, <code>@throws</code>, <code>@deprecated</code>, and these four types never
1121appear with an empty description. When a block tag doesn't fit on a single line, continuation lines
1122are indented four (or more) spaces from the position of the <code>@</code>.
1123</p>
1124
1125<h3 id="s7.2-summary-fragment">7.2 The summary fragment</h3>
1126
1127<p>Each Javadoc block begins with a brief <strong>summary fragment</strong>. This
1128fragment is very important: it is the only part of the text that appears in certain contexts such as
1129class and method indexes.</p>
1130
1131<p>This is a fragment&#8212;a noun phrase or verb phrase, not a complete sentence. It does
1132<strong>not</strong> begin with <code class="badcode">A {@code Foo} is a...</code>, or
1133<code class="badcode">This method returns...</code>, nor does it form a complete imperative sentence
1134like <code class="badcode">Save the record.</code>. However, the fragment is capitalized and
1135punctuated as if it were a complete sentence.</p>
1136
1137<p class="tip"><strong>Tip:</strong> A common mistake is to write simple Javadoc in the form
1138<code class="badcode">/** @return the customer ID */</code>. This is incorrect, and should be
1139changed to <code class="prettyprint lang-java">/** Returns the customer ID. */</code>.</p>
1140
1141<a name="s7.3.3-javadoc-optional"></a>
1142<h3 id="s7.3-javadoc-where-required">7.3 Where Javadoc is used</h3>
1143
1144<p>At the <em>minimum</em>, Javadoc is present for every
1145<code class="prettyprint lang-java">public</code> class, and every
1146<code class="prettyprint lang-java">public</code> or
1147<code class="prettyprint lang-java">protected</code> member of such a class, with a few exceptions
1148noted below.</p>
1149
1150<p>Additional Javadoc content may also be present, as explained in Section 7.3.4,
1151<a href="#s7.3.4-javadoc-non-required">Non-required Javadoc</a>.</p>
1152
1153<h4 id="s7.3.1-javadoc-exception-self-explanatory">7.3.1 Exception: self-explanatory methods</h4>
1154
1155<p>Javadoc is optional for "simple, obvious" methods like
1156<code class="prettyprint lang-java">getFoo</code>, in cases where there <em>really and truly</em> is
1157nothing else worthwhile to say but "Returns the foo".</p>
1158
1159<p class="note"><strong>Important:</strong> it is not appropriate to cite this exception to justify
1160omitting relevant information that a typical reader might need to know. For example, for a method
1161named <code class="prettyprint lang-java">getCanonicalName</code>, don't omit its documentation
1162(with the rationale that it would say only
1163<code class="badcode">/** Returns the canonical name. */</code>) if a typical reader may have no idea
1164what the term "canonical name" means!</p>
1165
1166<h4 id="s7.3.2-javadoc-exception-overrides">7.3.2 Exception: overrides</h4>
1167
1168<p>Javadoc is not always present on a method that overrides a supertype method.
1169
1170</p>
1171
1172
1173
1174<h4 id="s7.3.4-javadoc-non-required">7.3.4 Non-required Javadoc</h4>
1175
1176<p>Other classes and members have Javadoc <em>as needed or desired</em>.
1177
1178</p><p>Whenever an implementation comment would be used to define the overall purpose or behavior of a
1179class or member, that comment is written as Javadoc instead (using <code>/**</code>).</p>
1180
1181<p>Non-required Javadoc is not strictly required to follow the formatting rules of Sections
11827.1.2, 7.1.3, and 7.2, though it is of course recommended.</p>
1183
1184
1185
1186</div>
1187</div>
1188</body>
1189</html>