pmd-rules.xml

<?xml version="1.0"?>
<!--

    Copyright (C) 2011-2018 ARM Limited. All rights reserved.
    SPDX-License-Identifier: Apache-2.0
    Licensed under the Apache License, Version 2.0 (the "License");
    you may not use this file except in compliance with the License.
    You may obtain a copy of the License at

    http://www.apache.org/licenses/LICENSE-2.0

    Unless required by applicable law or agreed to in writing, software
    distributed under the License is distributed on an "AS IS" BASIS,
    WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
    See the License for the specific language governing permissions and
    limitations under the License.

-->
<ruleset name="mbed ruleset" xmlns="http://pmd.sourceforge.net/ruleset/2.0.0"
	xmlns:xsi="http://www.w3.org/2001/XMLSchema-instance"
	xsi:schemaLocation="http://pmd.sourceforge.net/ruleset/2.0.0 http://pmd.sourceforge.net/ruleset_2_0_0.xsd">

	<description>pmd-rules</description>

	<rule ref="rulesets/java/basic.xml" />
	<rule ref="rulesets/java/braces.xml" />
	<rule ref="rulesets/java/clone.xml" />
    <rule ref="rulesets/java/unusedcode.xml"/>
    <rule ref="rulesets/java/finalizers.xml" />
    <rule ref="rulesets/java/imports.xml" >
        <exclude name="TooManyStaticImports" />
    </rule>


    <rule ref="rulesets/java/codesize.xml" >
        <!-- See the explanation for TooManyMethods.-->
        <exclude name="TooManyFields" />
        <!-- Design is very hard to measure by numbers like this.-->
        <!-- The number and quality of dependencies might be a better indicator, -->
        <!-- and that requires a different tool.-->
        <exclude name="TooManyMethods" />
        <!-- See the explanation for TooManyMethods.-->
        <exclude name="ExcessivePublicCount" />
        <!-- Needs better understanding and proper configuration-->
        <exclude name="StdCyclomaticComplexity" />
        <!-- Needs better understanding and proper configuration-->
        <exclude name="ModifiedCyclomaticComplexity" />
        <!-- See the explanation for TooManyMethods.-->
        <exclude name="ExcessiveParameterList" />
    </rule>
	<rule ref="rulesets/java/codesize.xml/CyclomaticComplexity">
		<properties>
			<property name="reportLevel" value="20" />
		</properties>
        <properties>
            <!-- ignore this run on methods: equals and hashCode-->
            <property name="violationSuppressXPath"
                      value="./ancestor::ClassOrInterfaceBodyDeclaration/MethodDeclaration[@Name='equals' or @Name='hashCode']"/>
        </properties>
    </rule>

    <rule ref="rulesets/java/codesize.xml/NPathComplexity">
        <properties>
            <property name="minimum" value="400" />
        </properties>
        <properties>
            <!-- ignore this run on methods: equals and hashCode-->
            <property name="violationSuppressXPath"
                      value="./ancestor::ClassOrInterfaceBodyDeclaration/MethodDeclaration[@Name='equals' or @Name='hashCode']"/>
        </properties>
    </rule>

	<rule ref="rulesets/java/design.xml">
        <!-- Sometimes important to avoid "DOS attack" but not as a general rule-->
        <exclude name="AvoidSynchronizedAtMethodLevel" />
        <!-- It's just extra effort to write and read the final keyword-->
        <exclude name="ClassWithOnlyPrivateConstructorsShouldBeFinal" />
        <!-- Maybe good idea if PMD could exclude null checks from this-->
        <exclude name="ConfusingTernary" />
        <!-- Statistical analysis is prone to givin false positives. Potential god classes-->
        <!-- most probably violate something else, too.-->
        <!-- And dependency analysis tools also help here.-->
        <exclude name="GodClass" />
        <!-- Switch is sometimes very readable-->
        <exclude name="TooFewBranchesForASwitchStatement"/>
        <!-- A static utility is a static utility even if it masquerades as something-->
        <!-- else by using the Singleton pattern-->
        <exclude name="UseUtilityClass" />
        <!-- This is good advice, but since it's violated so much in this project-->
        <!-- and the problem is not big (especially with good syntax colouring),-->
        <!-- we'll keep this ignored for now.-->
        <exclude name="AvoidReassigningParameters" />
        <!-- Good idea almost always, but not quite.-->
        <exclude name="ReturnEmptyArrayRatherThanNull" />
        <!-- Sometimes one step at a time makes clearer code.-->
        <!-- Debugging is also easier if the return value is in a variable.-->
        <exclude name="UnnecessaryLocalBeforeReturn" />
        <!-- There are valid reasons for passing arrays (making it nullable for example)-->
        <exclude name="UseVarargs" />
        <!-- TODO explain what false positives this gives-->
        <exclude name="MissingBreakInSwitch" />
        <!-- TODO enable when all findings have been fixed-->
        <exclude name="UseLocaleWithCaseConversions" />
        <!-- It gives a lot of warnings on 'equals' method, fixing would decrease readability-->
        <exclude name="SimplifyBooleanReturns"/>
        <!--Gives false positive-->
        <exclude name="FieldDeclarationsShouldBeAtStartOfClass"/>
        <!-- Good rule but in practice to often suppressed -->
        <exclude name="PreserveStackTrace"/>

        <exclude name="AccessorMethodGeneration"/>
	</rule>
	<rule ref="rulesets/java/migrating.xml">
        <!-- The annotation is not as readable and there is no way to state which-->
        <!-- line should throw the exception and with what message-->
        <exclude name="JUnitUseExpected" />
        <!-- Main code is not junit code-->
        <exclude name="JUnit4TestShouldUseTestAnnotation" />
	</rule>
	<rule ref="rulesets/java/naming.xml">
        <!-- Often good to start name with Abstract, but on the other hand this-->
        <!-- rule is a bit too much like Hungarian notation and I in interface names.-->
        <!-- Who cares if it's abstract or not when you are using it?-->
        <exclude name="AbstractNaming" />
        <!-- Opinion, for me a getter is not a command, it's a declarative-->
        <!-- data reference-->
        <exclude name="AvoidFieldNameMatchingMethodName" />
        <!-- Why should generics not be named properly, like all other things-->
        <!-- (well, except Windows filesystem roots)?-->
        <exclude name="GenericsNaming" />
        <!-- It can be long if it's the only way to make it good-->
        <exclude name="LongVariable" />
        <!-- It can be short if it's good-->
        <exclude name="ShortVariable" />
        <!-- TODO explain why.-->
        <exclude name="BooleanGetMethodName" />
        <!-- It can be short if it's good-->
        <exclude name="ShortClassName" />
        <!-- It can be short if it's good-->
        <exclude name="ShortMethodName" />
	</rule>
	<rule ref="rulesets/java/optimizations.xml">
		<!-- Too many false hits. Optimization can't be done with static analysis. 
			Besides, following this may encourage the antipattern of using too broad 
			scope for stuff: -->
		<exclude name="AvoidInstantiatingObjectsInLoops" />
		<!-- Good principle but too verbose in practice: -->
		<exclude name="MethodArgumentCouldBeFinal" />
		<!-- Good principle and maybe sometimes even practical but not in this 
			project: -->
		<exclude name="LocalVariableCouldBeFinal" />
	</rule>
	<rule ref="rulesets/java/strictexception.xml" >
        <!-- NPE communicates very cleary what is happening, it is not-->
        <!-- interesting who reports it (jvm or user code)-->
        <exclude name="AvoidThrowingNullPointerException" />
        <!-- TODO explain why-->
        <exclude name="AvoidCatchingGenericException" />
        <!-- TODO explain why-->
        <exclude name="AvoidThrowingRawExceptionTypes" />
        <!-- One valid case is to catch runtime, throw as such and after that-->
        <!-- catch Exception and wrap as runtime.-->
        <!-- Without the first all runtimes are unnecessarily wrapped.-->
        <exclude name="AvoidRethrowingException" />
	</rule>
	<rule ref="rulesets/java/strings.xml" >
        <!-- Splitting to multiple lines is sometimes more readable.-->
        <!-- Besides, where's the proof that it affects performance?-->
        <exclude name="ConsecutiveAppendsShouldReuse" />
	</rule>

</ruleset>