reader-api / app.py
bcci's picture
Update app.py
5dc46f9 verified
raw
history blame
4.4 kB
from fastapi import FastAPI, Request, HTTPException
from fastapi.responses import PlainTextResponse
from urllib.parse import unquote
import uvicorn
from scrapling import AsyncFetcher, StealthyFetcher
from markitdown import MarkItDown
import tempfile
import os
import asyncio
import re
app = FastAPI()
async_fetcher = AsyncFetcher(auto_match=True) # Use AsyncFetcher
stealthy_fetcher = StealthyFetcher() # Keep StealthyFetcher (it handles its own async internally)
md = MarkItDown()
async def stealthy_scraper(url):
"""Fetches HTML content using StealthyFetcher (already async internally)."""
# Use await if async_fetch is available, otherwise keep .fetch
html = await stealthy_fetcher.async_fetch(url) #Corrected to async_fetch
return html.html_content
async def scraper(url):
"""Fetches HTML content using AsyncFetcher."""
html = await async_fetcher.get(url) # Use await for async operations
return html.html_content
async def text_scraper(url):
"""Fetches HTML content using AsyncFetcher and than extract text."""
html = await async_fetcher.get(url)
return re.sub(r'\s+', ' ', re.sub(r'\n+', ' ', html.get_all_text())).strip()
async def convert_html_to_md(html):
"""Converts HTML to Markdown using MarkItDown (assuming it's synchronous)."""
with tempfile.NamedTemporaryFile(suffix=".html", delete=False) as temp_file:
# Asynchronously write to the temporary file
await asyncio.to_thread(temp_file.write, html.encode('utf-8'))
await asyncio.to_thread(temp_file.flush)
temp_file_path = temp_file.name
# Use asyncio.to_thread for synchronous operations within async functions
md_text = await asyncio.to_thread(md.convert, temp_file_path)
x = md_text.text_content
# Asynchronously remove file
await asyncio.to_thread(os.remove, temp_file_path)
return x
@app.get("/read/{url:path}", response_class=PlainTextResponse)
async def get_markdown_get(request: Request, url: str):
"""Handles GET requests to /read/{url}, returning Markdown content."""
try:
full_url = str(request.url)
full_url = full_url.split("/read/")[1]
if not full_url.startswith(('http://', 'https://')):
full_url = f"http://{full_url}"
markdown_output = await convert_html_to_md(await scraper(full_url))
return PlainTextResponse(markdown_output)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Error processing URL: {e}")
@app.get("/read_text/{url:path}", response_class=PlainTextResponse)
async def get_text_get(request: Request, url: str):
"""Handles GET requests to /read/{url}, returning Markdown content."""
try:
full_url = str(request.url)
full_url = full_url.split("/read_text/")[1]
if not full_url.startswith(('http://', 'https://')):
full_url = f"http://{full_url}"
text_output = await text_scraper(full_url)
return PlainTextResponse(text_output)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Error processing URL: {e}")
@app.get("/reader/{url:path}", response_class=PlainTextResponse)
async def get_markdown_get_stealthy(request: Request, url: str): # Renamed for clarity
"""Handles GET requests to /reader/{url}, using StealthyFetcher."""
try:
full_url = str(request.url)
full_url = full_url.split("/reader/")[1]
if not full_url.startswith(('http://', 'https://')):
full_url = f"http://{full_url}"
markdown_output = await convert_html_to_md(await stealthy_scraper(full_url))
return PlainTextResponse(markdown_output)
except Exception as e:
raise HTTPException(status_code=500, detail=f"Error processing URL: {e}")
if __name__ == "__main__":
async def run_app():
#This part is only needed for StealthyFetcher to work.
try:
process = await asyncio.create_subprocess_exec('camoufox', 'fetch')
await process.wait() #Wait for camoufox to initialize
print("Camoufox initialized successfully!")
except Exception as e:
print(f"An unexpected error occurred starting camoufox: {e}")
config = uvicorn.Config(app, host="0.0.0.0", port=7860)
server = uvicorn.Server(config)
await server.serve()
asyncio.run(run_app())