excel-author

1#!/usr/bin/env python32"""Recalculate an .xlsx file's formulas using LibreOffice headless.3 4Usage: python recalc.py <path.xlsx> [timeout_seconds]5 6openpyxl writes formula strings but does not compute them. Downstream scripts7that open the file with data_only=True get None for every formula cell until8something has actually calculated the workbook. Excel does this on open;9headless pipelines need LibreOffice (or similar) to do it explicitly.10 11Exits 0 on success (workbook recomputed and resaved in place), non-zero on12failure. Writes status JSON to stdout either way.13"""14 15import json16import shutil17import subprocess18import sys19import tempfile20from pathlib import Path21 22 23def find_libreoffice() -> str | None:24    for cmd in ("libreoffice", "soffice"):25        path = shutil.which(cmd)26        if path:27            return path28    return None29 30 31def recalc(xlsx_path: str, timeout: int = 60) -> dict:32    src = Path(xlsx_path).resolve()33    if not src.exists():34        return {"status": "error", "error": f"File not found: {src}"}35 36    lo = find_libreoffice()37    if lo is None:38        return {39            "status": "error",40            "error": "libreoffice not found on PATH — install it or recalc in a real Excel session",41        }42 43    with tempfile.TemporaryDirectory() as td:44        try:45            subprocess.run(46                [47                    lo,48                    "--headless",49                    "--calc",50                    "--convert-to",51                    "xlsx",52                    str(src),53                    "--outdir",54                    td,55                ],56                check=True,57                capture_output=True,58                timeout=timeout,59            )60        except subprocess.TimeoutExpired:61            return {"status": "error", "error": f"libreoffice timed out after {timeout}s"}62        except subprocess.CalledProcessError as e:63            return {64                "status": "error",65                "error": f"libreoffice exited {e.returncode}: {e.stderr.decode(errors='replace')[:500]}",66            }67 68        produced = Path(td) / src.name69        if not produced.exists():70            return {"status": "error", "error": "libreoffice did not produce output file"}71 72        shutil.copy(produced, src)73 74    return {"status": "success", "file": str(src)}75 76 77def main():78    if len(sys.argv) < 2:79        print("Usage: python recalc.py <path.xlsx> [timeout_seconds]", file=sys.stderr)80        sys.exit(2)81    timeout = int(sys.argv[2]) if len(sys.argv) > 2 else 6082    result = recalc(sys.argv[1], timeout=timeout)83    print(json.dumps(result, indent=2))84    sys.exit(0 if result["status"] == "success" else 1)85 86 87if __name__ == "__main__":88    main()89