Analyzer_Interface.md

docscanner/internal/analyzer/analyzer.go

Interface

type Analyzer interface {
	Supports(filename string) bool
	Analyze(filepath string, data []byte) (*model.ScanResult, error)
}

Design decisions

• Supports decides, based on filename/extension, whether this analyzer owns the file.
• Analyze receives the full file contents as a []byte plus the path.
• Each analyzer returns a *model.ScanResult describing its findings.

Extensibility

• New analyzers live in internal/analyzer/ and implement this interface.
• The worker pool only depends on the interface, so adding new analyzers does not change concurrency code.

We can plug in Excel, PowerPoint, archives, binaries—anything—behind this interface.

Where the interface is used

main.go
	└─ builds []Analyzer { WordAnalyzer, PDFAnalyzer, ... }
			 ↓
StartWorkerPool (workerpool.go)
	└─ receives []Analyzer and, for each file:
				├─ a.Supports(file)
				└─ a.Analyze(file, data) → *ScanResult

The rest of the system (CLI, worker pool, directory walker) only knows the Analyzer interface, not concrete implementations. This keeps code structure simple and makes it obvious where to plug in new behavior.