66 lines
2.0 KiB
Python
66 lines
2.0 KiB
Python
"""Convert DOCX bytes to PDF via LibreOffice headless."""
|
|
|
|
from __future__ import annotations
|
|
|
|
import shutil
|
|
import subprocess
|
|
import tempfile
|
|
from pathlib import Path
|
|
|
|
|
|
class PdfConversionError(RuntimeError):
|
|
pass
|
|
|
|
|
|
def find_soffice() -> str | None:
|
|
for name in ("soffice", "libreoffice"):
|
|
path = shutil.which(name)
|
|
if path:
|
|
return path
|
|
# common Windows / Linux paths
|
|
for candidate in (
|
|
r"C:\Program Files\LibreOffice\program\soffice.exe",
|
|
"/usr/bin/soffice",
|
|
"/usr/bin/libreoffice",
|
|
):
|
|
if Path(candidate).exists():
|
|
return candidate
|
|
return None
|
|
|
|
|
|
def docx_bytes_to_pdf(docx_bytes: bytes, timeout_sec: int = 120) -> bytes:
|
|
soffice = find_soffice()
|
|
if not soffice:
|
|
raise PdfConversionError(
|
|
"LibreOffice (soffice) не найден. Установите libreoffice-writer в контейнере API."
|
|
)
|
|
with tempfile.TemporaryDirectory(prefix="espd_preview_") as tmp:
|
|
tmp_path = Path(tmp)
|
|
src = tmp_path / "document.docx"
|
|
src.write_bytes(docx_bytes)
|
|
try:
|
|
proc = subprocess.run(
|
|
[
|
|
soffice,
|
|
"--headless",
|
|
"--nologo",
|
|
"--nofirststartwizard",
|
|
"--convert-to",
|
|
"pdf",
|
|
"--outdir",
|
|
str(tmp_path),
|
|
str(src),
|
|
],
|
|
capture_output=True,
|
|
text=True,
|
|
timeout=timeout_sec,
|
|
check=False,
|
|
)
|
|
except subprocess.TimeoutExpired as e:
|
|
raise PdfConversionError("Таймаут конвертации DOCX→PDF") from e
|
|
pdf_path = tmp_path / "document.pdf"
|
|
if proc.returncode != 0 or not pdf_path.exists():
|
|
detail = (proc.stderr or proc.stdout or "").strip()[:500]
|
|
raise PdfConversionError(f"Ошибка LibreOffice: {detail or proc.returncode}")
|
|
return pdf_path.read_bytes()
|