| """pdfsys-parser-pipeline — OCR-pipeline backend. | |
| Handles the "needs-ocr AND no complex content" branch. Reads the cached | |
| LayoutDocument produced by pdfsys-layout-analyser, renders each region via | |
| PyMuPDF, runs line-level OCR (RapidOCR / PaddleOCR-classic, selectable via | |
| config), and assembles the Markdown output. CPU-friendly. | |
| """ | |
| __version__ = "0.0.1" | |