sonarsource/slang

SLang

This is a developer documentation. If you want to analyze source code in SonarQube read one of the following documentation:

• Ruby language: analysis of Ruby documentation
• Scala language: analysis of Scala documentation
• Go language: analysis of Go documentation

SLang (SonarSource Language) is a framework to quickly develop code analyzers for SonarQube to help developers write Clean Code.

SLang defines language-agnostic AST. Using this AST we can develop simple syntax-based rules. Then we use a parser for real language to create this AST. Currently, Ruby and Scala analyzers use this approach.

Ruby

We use whitequark parser to parse the Ruby language by embedding it using JRuby runtime.

• AST documentation for the parser can be found here
• We use simple Ruby script to call the parser and invoke our visitor written in Java

Scala

We use Scalameta to parse the Scala language.

Scala coverage

For Scala files, we will import both Scoverage and JaCoCo coverage reports. Note that this will result in strange behavior since:

• Only line coverage will be used from the Scoverage report.
• JaCoCo can be imprecise when computing conditions coverage on Scala code, generating FP (typically on pattern matching).

This situation only applies to two Scala files, this current situation is acceptable.

Go

We use the native Go parser to parse the Go language.

Have questions or feedback?

To provide feedback (request a feature, report a bug, etc.) use the SonarQube Community Forum. Please do not forget to specify the language, plugin version, and SonarQube version.

Building

Setup

If you are on Windows, read the sonar-go-to-slang/README.md instructions.

SonarSource internal usage: Configure your gradle.properties - read the private/README.md instructions.

Build

Build and run Unit Tests:

./gradlew build

Integration Tests

By default, Integration Tests (ITs) are skipped during builds. If you want to run them, you need first to retrieve the related projects which are used as input:

git submodule update --init its/sources

Then build and run the Integration Tests using the its property:

./gradlew build -Pits --info --no-daemon

You can also build and run only Ruling Tests using the ruling property:

./gradlew build -Pruling --info --no-daemon

If you want to run ruling tests for a specific language, you can use ruling-{lang} property (ruling-scala, ruling-ruby, ruling-go). For example:

./gradlew build -Pruling-scala --info --no-daemon

License headers

License headers are automatically updated by the spotless plugin but only for Java files. Furthermore, there are files such as package-info.java and module-info.java that spotless ignores. Also, Scala and Go source files are not handled. For those files use a manual script like below to update the license. E.g., for Go files (on Mac):

`find . -type f -name "*.go" -exec sed -i '' 's/2018-2023/2018-2024/' "{}" \;`