How to add a new language to PMD using ANTLR grammar.
Table of Contents
Before you start…

This is really a big contribution and can’t be done with a drive by contribution. It requires dedicated passion and long commitment to implement support for a new language.

This step-by-step guide is just a small intro to get the basics started, and it’s also not necessarily up-to-date or complete. You have to be able to fill in the blanks.

Currently, the Antlr integration has some basic limitations compared to JavaCC: The output of the Antlr parser generator is not an abstract syntax tree (AST) but a parse tree (also known as CST, concrete syntax tree). As such, a parse tree is much more fine-grained than what a typical JavaCC grammar will produce. This means that the parse tree is much deeper and contains nodes down to the different token types.

The Antlr nodes are context objects and serve a different abstraction than nodes in an AST. These context objects themselves don’t have any attributes because they themselves represent the attributes (as nodes or leaves in the parse tree). As they don’t have attributes, there are no attributes that can be used in XPath based rules.

The current implementation of the languages using ANTLR use these context objects as nodes in PMD’s AST representation.

In order to overcome these limitations, one would need to implement a post-processing step that transforms a parse tree into an abstract syntax tree and introducing real nodes on a higher abstraction level. These real nodes can then have attributes which are available in XPath based rules. The transformation can happen with a visitor, but the implementation of the AST is a manual step. This step is not described in this guide.

After the basic support for a language is there, there are lots of missing features left. Typical features that can greatly improve rule writing are: symbol table, type resolution, call/data flow analysis.

Symbol table keeps track of variables and their usages. Type resolution tries to find the actual class type of each used type, following along method calls (including overloaded and overwritten methods), allowing to query subtypes and type hierarchy. This requires additional configuration of an auxiliary classpath. Call and data flow analysis keep track of the data as it is moving through different execution paths a program has.

These features are out of scope of this guide. Type resolution and data flow are features that definitely don’t come for free. It is much effort and requires perseverance to implement.

Steps

1. Start with a new sub-module

  • See pmd-swift for examples.
  • Make sure to add your new module to PMD’s parent pom as <module> entry, so that it is built alongside the other languages.
  • Also add your new module to the dependencies list in “pmd-languages-deps/pom.xml”, so that the new language is automatically available in the binary distribution (pmd-dist).

2. Implement an AST parser for your language

  • ANTLR will generate the parser for you based on the grammar file. The grammar file needs to be placed in the folder src/main/antlr4 in the appropriate sub package ast of the language. E.g. for swift, the grammar file is Swift.g4 and is placed in the package net.sourceforge.pmd.lang.swift.ast.
  • Configure the options “superClass” and “contextSuperClass”. These are the base classes for the generated classes.

3. Create AST node classes

  • The individual AST nodes are generated, but you need to define the common interface for them.
  • You need to define the supertype interface for all nodes of the language. For that, we provide AntlrNode.
  • See SwiftNode as an example.
  • Additionally, you need several base classes:
    • a language specific inner node - these nodes represent the production rules from the grammar. In Antlr, they are called “ParserRuleContext”. We call them “InnerNode”. Use the base class from pmd-core BaseAntlrInnerNode . And example is SwiftInnerNode. Note that this language specific inner node is package-private, as it is only the base class for the concrete nodes generated by ANLTR.
    • a language specific root node - this provides the root of the AST and our parser will return subtypes of this node. The root node itself is a “InnerNode”. See SwiftRootNode. Note that this language specific root node is package-private, as it is only the base class for the concrete node generated by ANLTR.
    • a language specific terminal node. See SwiftTerminalNode.
    • a language specific error node. See SwiftErrorNode.
    • a language name dictionary. This is used to convert ANTLR node names to useful XPath node names. See `SwiftNameDictionary’.
  • Once these base classes exist, you need to change the ANTLR grammar to add additional members via @parser::members
    • Define a package private field DICO which creates a new instance of your language name dictionary using the vocabulary from the generated parser (VOCABULARY).
    • Define two additional methods to help converting the ANTLR context objects into PMD AST nodes. The methods are abstract in AntlrGeneratedParserBase and need to be implemented here for the concrete language: createPmdTerminal() and createPmdError().
  • In order for the generated code to match and use our custom classes, we have a common ant script, that fiddles with the generated code. The ant script is antlr4-wrapper.xml and does not need to be adjusted - it has plenty of parameters that can be configured. The ant script is added in the language module’s pom.xml where the parameters are set (e.g. name of root name class). Have a look at Swift’s example: pmd-swift/pom.xml.
  • You can add additional methods in your “InnerNode” (e.g. SwiftInnerNode) that are available on all nodes. But on most cases you won’t need to do anything.

4. Generate your parser (using ANTLR)

  • Make sure, you have the property <antlr4.visitor>true</antlr4.visitor> in your pom.xml file.
  • This is just a matter of building the language module. ANTLR is called via ant, and this step is added to the phase generate-sources. So you can just call e.g. ./mvnw generate-sources -pl pmd-swift to have the parser generated.
  • The generated code will be placed under target/generated-sources/antlr4 and will not be committed to source control.
  • You should review pmd-swift/pom.xml.

5. Create a TokenManager

  • This is needed to support CPD (copy paste detection)
  • We provide a default implementation using AntlrTokenManager.
  • You must create your own “AntlrCpdLexer” such as we do with SwiftCpdLexer.
  • If you wish to filter specific tokens (e.g. comments to support CPD suppression via “CPD-OFF” and “CPD-ON”) you can create your own implementation of AntlrTokenFilter. You’ll need to override then the protected method getTokenFilter(AntlrTokenManager) and return your custom filter. See the CpdLexer for C# as an exmaple: CsCpdLexer.

    If you don’t need a custom token filter, you don’t need to override the method. It returns the default AntlrTokenFilter which doesn’t filter anything.

6. Create a PMD parser “adapter”

  • Create your own parser, that adapts the ANLTR interface to PMD’s parser interface.
  • We provide a AntlrBaseParser implementation that you need to extend to create your own adapter as we do with PmdSwiftParser.

7. Create a language version handler

  • Now you need to create your version handler, as we did with SwiftHandler.
  • This class is sort of a gateway between PMD and all parsing logic specific to your language.
  • For a minimal implementation, it just needs to return a parser (see step #6).
  • It can be used to provide other features for your language like
    • violation suppression logic
    • ViolationDecorators, to add additional language specific information to the created violations. The Java language module uses this to provide the method name or class name, where the violation occurred.
    • metrics
    • custom XPath functions

8. Create a base visitor

  • A parser visitor adapter is not needed anymore with PMD 7. The visitor interface now provides a default implementation.
  • The visitor for ANTLR based AST is generated along the parser from the ANTLR grammar file. The base interface for a visitor is AstVisitor.
  • The generated visitor class for Swift is called SwiftVisitor.
  • In order to help use this visitor later on, a base visitor class should be created. See SwiftVisitorBase as an example.

9. Make PMD recognize your language

  • Create your own subclass of net.sourceforge.pmd.lang.impl.SimpleLanguageModuleBase, see Swift as an example: SwiftLanguageModule.
  • Add for each version of your language a call to addVersion in your language module’s constructor. Use addDefaultVersion for defining the default version.
  • You’ll need to refer the language version handler created in step #7.
  • Create the service registration via the text file src/main/resources/META-INF/services/net.sourceforge.pmd.lang.Language. Add your fully qualified class name as a single line into it.

10. Create an abstract rule class for the language

  • You need to create your own abstract rule class in order to interface your language with PMD’s generic rule execution.
  • See AbstractSwiftRule as an example.
  • The rule basically just extends AbstractVisitorRule and only redefines the abstract buildVisitor() method to return our own type of visitor. In this case our SwiftVisitor is used. While there is no real functionality added, every language should have its own base class for rules. This helps to organize the code.
  • All other rules for your language should extend this class. The purpose of this class is to provide a visitor via the method buildVisitor() for analyzing the AST. The provided visitor only implements the visit methods for specific AST nodes. The other node types use the default behavior, and you don’t need to care about them.
  • Note: This is different from how it was in PMD 6: Each rule in PMD 6 was itself a visitor (implementing the visitor interface of the specific language). Now the rule just provides a visitor, which can be hidden and potentially shared between rules.

11. Create rules

  • Creating rules is already pretty well documented in PMD - and it’s no different for a new language, except you may have different AST nodes.
  • PMD supports 2 types of rules, through visitors or XPath.
  • To add a visitor rule:
    • You need to extend the abstract rule you created on the previous step, you can use the swift rule UnavailableFunctionRule as an example. Note, that all rule classes should be suffixed with Rule and should be placed in a package the corresponds to their category.
  • To add an XPath rule you can follow our guide Writing XPath Rules.
  • When creating the category ruleset XML file, the XML can reference build properties that are replaced during the build. This is used for the externalInfoUrl attribute of a rule. E.g. we use ${pmd.website.baseurl} to point to the correct webpage (depending on the PMD version). In order for this to work, you need to add a resource filtering configuration in the language module’s pom.xml. Under <build> add the following lines:
        <resources>
            <resource>
                <directory>${project.basedir}/src/main/resources</directory>
                <filtering>true</filtering>
            </resource>
        </resources>
    

12. Test the rules

  • Testing rules is described in depth in Testing your rules.
    • Each rule has its own test class: Create a test class for your rule extending PmdRuleTst (see UnavailableFunctionTest for example)
    • Create a category rule set for your language (see pmd-swift/src/main/resources/bestpractices.xml for example)
    • Place the test XML file with the test cases in the correct location
    • When executing the test class
      • this triggers the unit test to read the corresponding XML file with the rule test data (see UnavailableFunction.xml for example)
      • This test XML file contains sample pieces of code which should trigger a specified number of violations of this rule. The unit test will execute the rule on this piece of code, and verify that the number of violations matches.
  • To verify the validity of all the created rulesets, create a subclass of AbstractRuleSetFactoryTest (see RuleSetFactoryTest in pmd-swift for example). This will load all rulesets and verify, that all required attributes are provided.

    Note: You’ll need to add your ruleset to categories.properties, so that it can be found.

13. Create documentation page

Finishing up your new language module by adding a page in the documentation. Create a new markdown file <langId>.md in docs/pages/pmd/languages/. This file should have the following frontmatter:

---
title: <Language Name>
permalink: pmd_languages_<langId>.html
last_updated: <Month> <Year> (<PMD Version>)
tags: [languages, PmdCapableLanguage, CpdCapableLanguage]
---

On this page, language specifics can be documented, e.g. when the language was first supported by PMD. There is also the following Jekyll Include, that creates summary box for the language:


{% include language_info.html name='<Language Name>' id='<langId>' implementation='<langId>::lang.<langId>.<langId>LanguageModule' supports_cpd=true supports_pmd=true %}

Optional features

See Optional features in JavaCC based languages.

In order to implement these, most likely an AST needs to be developed first. The parse tree (CST, concrete syntax tree) is not suitable to add methods such as getSymbol() to the node classes.