Add checkstyle rule for javadoc and update all javadocs to 80 cols

This commit is contained in:
Mumfrey committed 2015-01-20 13:49:14 +00:00
1 parent 116ecb0458
commit e87402bf70
49 files changed
+844 -432

No files matched your search

@@ -30,23 +30,31 @@ import java.lang.annotation.RetentionPolicy;
import java.lang.annotation.Target;
/**
* Used to indicate a Mixin class member which is acting as a placeholder for a method or field in the target class
* Used to indicate a Mixin class member which is acting as a placeholder for a
* method or field in the target class
*/
@Target({ ElementType.METHOD, ElementType.FIELD })
@Retention(RetentionPolicy.RUNTIME)
public @interface Shadow {
// CHECKSTYLE:OFF
/**
* <p>In general, shadow methods can be declared using their name in the target class as you would expect, however we run into a problem when we
* want to mix in a method with the same name and arguments, but a different return type to the shadow method. While the JVM itself will happily
* support methods with signatures that differ only on return type, the compiler itself does not. This poses a problem, since we have no way to
* <p>In general, shadow methods can be declared using their name in the
* target class as you would expect, however we run into a problem when we
* want to mix in a method with the same name and arguments, but a different
* return type to the shadow method. While the JVM itself will happily
* support methods with signatures that differ only on return type, the
* compiler itself does not. This poses a problem, since we have no way to
* leverage this behaviour since our mixin class will not compile.</p>
*
* <p>To circumvent this compiler limitation, the prefix option can be used. By specifying a prefix for the shadow method, it is subsequently
* possible to compile the mixin class, the specified prefix will then be stripped from the method name prior to applying the mixin, and
* everything will work as expected. You may either use the default prefix: "shadow$", or you may specify your own. It is good practice to
* specify the prefix if you are using it, regardless of whether you use the default or not. For example consider the intrinsic readability of
* the following snippets</p>:
* <p>To circumvent this compiler limitation, the prefix option can be used.
* By specifying a prefix for the shadow method, it is subsequently possible
* to compile the mixin class, the specified prefix will then be stripped
* from the method name prior to applying the mixin, and everything will
* work as expected. You may either use the default prefix: "shadow$", or
* you may specify your own. It is good practice to specify the prefix if
* you are using it, regardless of whether you use the default or not. For
* example consider the intrinsic readability of the following snippets</p>:
*
* <blockquote><pre>
* &#64;Shadow abstract void someMethod(int arg1, int arg2);
@@ -55,22 +63,34 @@ public @interface Shadow {
* &#64;Shadow(prefix = "foo$") abstract void foo$someMethod(int arg1, int arg2);
* </pre></blockquote>
*
* <p>All of these declarations are semantically equivalent, however the third and fourth are the most expressive in terms of making their
* intentions clear, and thus specifying prefix is recommended, since it aids readability and maintainability.</p>
* <p>All of these declarations are semantically equivalent, however the
* third and fourth are the most expressive in terms of making their
* intentions clear, and thus specifying prefix is recommended, since it
* aids readability and maintainability.</p>
*
* <p>Note that specifying a <em>prefix</em> does not <b>enforce</b> use of the prefix, the behaviour of <em>prefix</em> is such that the prefix
* will be stripped from the start of the method name <em>as long as the method name actually starts with the prefix</em>! This has important
* repercussions since if the annotation value does not match the method prefix then <em>no renaming will take place</em> likey resulting in
* a failure state indicated by an {@link InvalidMixinException} at run time.</p>
* <p>Note that specifying a <em>prefix</em> does not <b>enforce</b> use of
* the prefix, the behaviour of <em>prefix</em> is such that the prefix will
* be stripped from the start of the method name <em>as long as the method
* name actually starts with the prefix</em>! This has important
* repercussions since if the annotation value does not match the method
* prefix then <em>no renaming will take place</em> likey resulting in a
* failure state indicated by an {@link InvalidMixinException} at run
* time.</p>
*/
public String prefix() default "shadow$";
// CHECKSTYLE:ON
/**
* By default, the annotation processor will attempt to locate an obfuscation mapping for all {@link Shadow} methods since it is anticipated that
* in general the target of a {@link Shadow} annotation will be an obfuscated field or method in the target class. However since it is possible
* to also apply mixins to non-obfuscated targets (or non-obfuscated methods in obfuscated targets, such as methods added by Forge) it may be
* desirable to suppress the compiler warning which would otherwise be generated. Setting this value to <em>false</em> will cause the annotation
* processor to skip this annotation when attempting to build the obfuscation table for the mixin.
* By default, the annotation processor will attempt to locate an
* obfuscation mapping for all {@link Shadow} methods since it is
* anticipated that in general the target of a {@link Shadow} annotation
* will be an obfuscated field or method in the target class. However since
* it is possible to also apply mixins to non-obfuscated targets (or non-
* obfuscated methods in obfuscated targets, such as methods added by Forge)
* it may be desirable to suppress the compiler warning which would
* otherwise be generated. Setting this value to <em>false</em> will cause
* the annotation processor to skip this annotation when attempting to build
* the obfuscation table for the mixin.
*/
public boolean remap() default true;
}