#!/usr/bin/env python3
"""Konvertiert PDF, DOCX, TXT und MD in durchsuchbare Markdown-Dateien.

Beispiel:
  /srv/knp-ai/venv/bin/python 03-prepare-documents.py ./Originale /srv/knp-ai/testwissen
"""

from __future__ import annotations
import argparse
import hashlib
from pathlib import Path
import shutil

try:
    import fitz  # PyMuPDF
    from docx import Document
except ImportError as exc:
    raise SystemExit(
        "Abhängigkeiten fehlen. Zuerst 03-install-document-tools.sh ausführen."
    ) from exc


def safe_name(path: Path) -> str:
    digest = hashlib.sha1(str(path).encode("utf-8")).hexdigest()[:8]
    stem = "".join(c if c.isalnum() or c in "-_." else "_" for c in path.stem)
    return f"{stem}_{digest}.md"


def frontmatter(source: Path, kind: str) -> str:
    src = str(source).replace('"', "'")
    return f'---\nsource: "{src}"\ntype: "{kind}"\n---\n\n'


def convert_pdf(src: Path) -> str:
    doc = fitz.open(src)
    parts = [frontmatter(src, "pdf")]
    for page_no, page in enumerate(doc, start=1):
        text = page.get_text("text").strip()
        parts.append(f"## Seite {page_no}\n\n{text}\n")
    return "\n".join(parts)


def convert_docx(src: Path) -> str:
    doc = Document(src)
    parts = [frontmatter(src, "docx")]
    for p in doc.paragraphs:
        text = p.text.strip()
        if text:
            parts.append(text)
    for table_no, table in enumerate(doc.tables, start=1):
        parts.append(f"\n## Tabelle {table_no}\n")
        for row in table.rows:
            parts.append(" | ".join(cell.text.strip() for cell in row.cells))
    return "\n\n".join(parts)


def convert_text(src: Path) -> str:
    return frontmatter(src, src.suffix.lower().lstrip(".")) + src.read_text(
        encoding="utf-8", errors="replace"
    )


def main() -> None:
    ap = argparse.ArgumentParser()
    ap.add_argument("source", type=Path, help="Ordner mit Originaldokumenten")
    ap.add_argument("target", type=Path, help="Zielordner für Markdown")
    args = ap.parse_args()

    args.target.mkdir(parents=True, exist_ok=True)
    supported = {".pdf", ".docx", ".txt", ".md"}
    converted = skipped = failed = 0

    for src in args.source.rglob("*"):
        if not src.is_file() or src.suffix.lower() not in supported:
            continue
        out = args.target / safe_name(src)
        try:
            ext = src.suffix.lower()
            if ext == ".pdf":
                content = convert_pdf(src)
            elif ext == ".docx":
                content = convert_docx(src)
            elif ext == ".md":
                # Bestehendes Markdown nur mit Metadaten ergänzen.
                content = convert_text(src)
            else:
                content = convert_text(src)
            out.write_text(content, encoding="utf-8")
            converted += 1
            print(f"OK   {src} -> {out.name}")
        except Exception as exc:
            failed += 1
            print(f"FEHLER {src}: {exc}")

    print(f"\nFertig: {converted} konvertiert, {skipped} übersprungen, {failed} Fehler.")


if __name__ == "__main__":
    main()
