[FEAT] Optical Character Recognition (OCR) Support #7
Description
Implement screen.ocr and screen.findText capabilities to enable semantic observation of applications that lack standard UIA/Win32 metadata.
Architectural Guidance:
- Placement: This should reside in the Client Layer (GUI/CLI) or as a separate Plugin Module rather than the wininspect_core.
- Reasoning: Core should remain lightweight and injection-free. OCR engines add significant binary size and external dependencies.
- Workflow: Client captures a screen region using the existing core methods and performs OCR locally.