"""Extract QUY_TRINH docx text for manual workflow modeling."""
from pathlib import Path
import docx

SRC = Path("D:/Dropbox/CONG_VIEC/SOLUTION/QUY_TRINH/QT TRINH KY HOP DONG TP-NCC.docx")
OUT = Path(__file__).parent.parent / "docs" / "workflow-raw.md"

d = docx.Document(str(SRC))
parts = []
for p in d.paragraphs:
    if p.text.strip():
        parts.append(p.text)
for i, tbl in enumerate(d.tables):
    parts.append(f"\n--- TABLE {i+1} ({len(tbl.rows)}r x {len(tbl.columns)}c) ---")
    for row in tbl.rows:
        cells = [c.text.strip().replace("\n", " | ") for c in row.cells]
        parts.append(" || ".join(cells))

OUT.parent.mkdir(parents=True, exist_ok=True)
OUT.write_text("\n".join(parts), encoding="utf-8")
print(f"OK -> {OUT} ({OUT.stat().st_size} bytes)")