mirror of
https://github.com/kennethreitz/kjvstudy.org.git
synced 2026-06-05 23:00:16 +00:00
kennethreitz
9ec7885dce
Convert all PDF generation endpoints from synchronous to async to prevent blocking FastAPI's event loop during CPU-intensive operations. Changes: - Add render_html_to_pdf_async() using ThreadPoolExecutor (2 workers) - Convert all PDF endpoints to async def - Use await render_html_to_pdf_async() instead of blocking calls - Keep render_html_to_pdf() for backward compatibility Performance impact: - Prevents event loop blocking during PDF generation - Allows other requests to be processed while PDFs are rendering - Limits concurrent PDF generation to 2 workers to control CPU usage Files updated: - kjvstudy_org/utils/pdf.py (new async implementation) - kjvstudy_org/server.py (5 PDF endpoints) - kjvstudy_org/routes/api.py (4 PDF endpoints) - kjvstudy_org/routes/resources.py (7 PDF endpoints) - kjvstudy_org/routes/stories.py (2 PDF endpoints) - kjvstudy_org/routes/study_guides.py (1 PDF endpoint) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude <noreply@anthropic.com>
50 lines
1.7 KiB
Python
50 lines
1.7 KiB
Python
"""Utility helpers for HTML-to-PDF generation."""
|
|
import io
|
|
from typing import BinaryIO
|
|
from concurrent.futures import ThreadPoolExecutor
|
|
import asyncio
|
|
|
|
try: # pragma: no cover - optional dependency
|
|
from weasyprint import HTML # type: ignore
|
|
WEASYPRINT_AVAILABLE = True
|
|
except (ImportError, OSError): # pragma: no cover - handled gracefully elsewhere
|
|
HTML = None
|
|
WEASYPRINT_AVAILABLE = False
|
|
|
|
# Thread pool for CPU-intensive PDF generation
|
|
_pdf_executor = ThreadPoolExecutor(max_workers=2, thread_name_prefix="pdf_worker")
|
|
|
|
|
|
def _render_pdf_sync(html_content: str) -> BinaryIO:
|
|
"""Internal synchronous PDF rendering function."""
|
|
if not WEASYPRINT_AVAILABLE or HTML is None:
|
|
raise RuntimeError("WeasyPrint is not available for PDF generation")
|
|
|
|
pdf_buffer = io.BytesIO()
|
|
HTML(string=html_content).write_pdf(pdf_buffer)
|
|
pdf_buffer.seek(0)
|
|
return pdf_buffer
|
|
|
|
|
|
def render_html_to_pdf(html_content: str) -> BinaryIO:
|
|
"""Synchronous wrapper for backward compatibility.
|
|
|
|
NOTE: Use render_html_to_pdf_async() in async contexts to avoid blocking.
|
|
|
|
Returns a BytesIO instance positioned at the beginning of the generated PDF.
|
|
Raises RuntimeError if WeasyPrint isn't available at runtime.
|
|
"""
|
|
return _render_pdf_sync(html_content)
|
|
|
|
|
|
async def render_html_to_pdf_async(html_content: str) -> BinaryIO:
|
|
"""Async-compatible PDF rendering that won't block the event loop.
|
|
|
|
Runs PDF generation in a thread pool to prevent blocking FastAPI.
|
|
|
|
Returns a BytesIO instance positioned at the beginning of the generated PDF.
|
|
Raises RuntimeError if WeasyPrint isn't available at runtime.
|
|
"""
|
|
loop = asyncio.get_event_loop()
|
|
return await loop.run_in_executor(_pdf_executor, _render_pdf_sync, html_content)
|