Background code analysis is one of the most popular plugins tasks. The analysis is performed by daemons created using a special IDaemonStage interface. Nevertheless, the easiest way to perform code analysis is to employ the ElementProblemAnalyzer<T> class: it creates the corresponding daemon and daemon stage by itself.
The analysis could be divided into two steps:
Actually, the analysis of code: performed by ElementProblemAnalyzer<T>, where T is the class of code elements you want to analyze.
Problematic code highlighting: performed by a class that implements the IHighlighting interface.
As an example, let's create a code analyzer that checks all variable declarations on whether they contain the word "Crap" (in Create a Quick-Fix, we implement a quick-fix that suggests replacing the "Crap" occurrence with "BadWord").
Problem Analyzer
[ElementProblemAnalyzer(typeof(IVariableDeclaration), HighlightingTypes =
new[] {typeof(BadWordNamingWarning)})]
public class BadWordNamingAnalyzer : ElementProblemAnalyzer<IVariableDeclaration>
{
protected override void Run(IVariableDeclaration element, ElementProblemAnalyzerData data,
IHighlightingConsumer consumer)
{
var nodeText = element.DeclaredName.ToLower();
if (!nodeText.Contains("crap"))
return;
consumer.AddHighlighting(new BadWordNamingWarning(element, element.NameIdentifier.GetDocumentRange()));
}
}
Highlighting
[StaticSeverityHighlighting(Severity.WARNING, HighlightingGroupIds.GutterMarksGroup)]
public class BadWordNamingWarning : IHighlighting
{
private readonly DocumentRange _documentRange;
public readonly IVariableDeclaration VariableDeclaration;
public BadWordNamingWarning(IVariableDeclaration variableDeclaration, DocumentRange documentRange)
{
VariableDeclaration = variableDeclaration;
_documentRange = documentRange;
}
public bool IsValid()
{
return VariableDeclaration.IsValid();
}
public DocumentRange CalculateRange()
{
return _documentRange;
}
public string ToolTip => "The name contains a bad word";
public string ErrorStripeToolTip { get; }
}
