scrape.py

from fastapi import FastAPI, HTTPException, Query
from pydantic import BaseModel
from playwright.async_api import async_playwright
import asyncio
import base64
import logging
from typing import List, Optional

# Set up logging
logging.basicConfig(level=logging.INFO)
logger = logging.getLogger(__name__)

app = FastAPI(title="Playwright Web Scraper", description="A simple web scraper using Playwright")

class LinkInfo(BaseModel):
    text: str
    href: str

class ScrapeResponse(BaseModel):
    body_content: Optional[str] = None
    screenshot: Optional[str] = None
    links: Optional[List[LinkInfo]] = None
    page_title: Optional[str] = None
    meta_description: Optional[str] = None

@app.get("/")
async def root():
    return {
        "message": "Playwright Web Scraper API - Body, Links & Images",
        "endpoints": {
            "/scrape": "Scrape webpage body content, links, and take screenshot",
            "/docs": "API documentation"
        },
        "example": "/scrape?url=https://example.com&screenshot=true&get_links=true&get_body=true",
        "features": [
            "Extract body tag content (clean text)",
            "Get all links with text and URLs",
            "Take full page screenshot",
            "Extract page title and meta description"
        ]
    }

@app.get("/scrape")
async def scrape_page(
    url: str = Query(..., description="URL to scrape"),
    screenshot: bool = Query(True, description="Take a full page screenshot"),
    get_links: bool = Query(True, description="Extract all links from the page"),
    get_body: bool = Query(True, description="Extract body tag content")
):
    logger.info(f"Starting scrape for URL: {url}")
    try:
        async with async_playwright() as p:
            logger.info("Launching browser...")
            browser = await p.chromium.launch(
                headless=True,
                args=[
                    '--no-sandbox',
                    '--disable-setuid-sandbox',
                    '--disable-dev-shm-usage',
                    '--disable-accelerated-2d-canvas',
                    '--no-first-run',
                    '--no-zygote',
                    '--disable-gpu'
                ]
            )
            page = await browser.new_page()

            try:
                logger.info(f"Navigating to {url}...")
                await page.goto(url, wait_until="networkidle")
                response = ScrapeResponse()

                # Always get page title and meta description
                logger.info("Getting page metadata...")
                response.page_title = await page.title()

                meta_desc = await page.evaluate("""
                    () => {
                        const meta = document.querySelector('meta[name="description"]');
                        return meta ? meta.getAttribute('content') : null;
                    }
                """)
                response.meta_description = meta_desc

                # Get body content (clean text)
                if get_body:
                    logger.info("Extracting body content...")
                    body_content = await page.evaluate("""
                        () => {
                            const body = document.querySelector('body');
                            if (!body) return null;

                            // Remove script and style elements
                            const scripts = body.querySelectorAll('script, style, noscript');
                            scripts.forEach(el => el.remove());

                            // Get clean text content
                            return body.innerText.trim();
                        }
                    """)
                    response.body_content = body_content

                # Get screenshot (full page)
                if screenshot:
                    logger.info("Taking full page screenshot...")
                    screenshot_bytes = await page.screenshot(full_page=True)
                    response.screenshot = base64.b64encode(screenshot_bytes).decode('utf-8')

                # Get links with better filtering
                if get_links:
                    logger.info("Extracting links...")
                    links = await page.evaluate("""
                        () => {
                            return Array.from(document.querySelectorAll('a[href]')).map(a => {
                                const text = a.innerText.trim();
                                const href = a.href;

                                // Only include links with meaningful text and valid URLs
                                if (text && href && href.startsWith('http')) {
                                    return {
                                        text: text.substring(0, 200), // Limit text length
                                        href: href
                                    }
                                }
                                return null;
                            }).filter(link => link !== null);
                        }
                    """)
                    response.links = [LinkInfo(**link) for link in links]

                await browser.close()
                logger.info("Scraping completed successfully")
                return response

            except Exception as e:
                logger.error(f"Error during scraping: {str(e)}")
                await browser.close()
                raise HTTPException(status_code=500, detail=f"Scraping error: {str(e)}")

    except Exception as e:
        logger.error(f"Error launching browser: {str(e)}")
        raise HTTPException(status_code=500, detail=f"Browser launch error: {str(e)}")