web_scraper.py

from flask import Flask, jsonify, request
import requests
from bs4 import BeautifulSoup
import os
import re
import urllib.parse
import time
import random
import base64
from io import BytesIO
from googlesearch import search
import json

app = Flask(__name__)

def search_images(query, num_images=5):
    # Headers to mimic a browser request
    headers = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
        'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
        'Accept-Language': 'en-US,en;q=0.5',
        'Accept-Encoding': 'gzip, deflate',
        'DNT': '1',
        'Connection': 'keep-alive',
    }

    # Format the query for URL
    formatted_query = urllib.parse.quote(query + " high quality")

    # Google Images URL
    url = f"https://www.google.com/search?q={formatted_query}&tbm=isch&safe=active"

    try:
        # Get the HTML content
        response = requests.get(url, headers=headers, timeout=30)
        response.raise_for_status()

        # Find all image URLs using regex
        image_urls = re.findall(r'https?://[^"\']*?(?:jpg|jpeg|png|gif)', response.text)

        # Remove duplicates while preserving order
        image_urls = list(dict.fromkeys(image_urls))

        # Filter and clean results
        results = []
        for img_url in image_urls:
            if len(results) >= num_images:
                break

            # Skip small thumbnails, icons, and low-quality images
            if ('gstatic.com' in img_url or 
                'google.com' in img_url or 
                'icon' in img_url.lower() or 
                'thumb' in img_url.lower() or
                'small' in img_url.lower()):
                continue

            try:
                # Verify the image URL is valid
                img_response = requests.head(img_url, headers=headers, timeout=5)
                if img_response.status_code == 200:
                    content_type = img_response.headers.get('Content-Type', '')
                    if content_type.startswith('image/'):
                        results.append({
                            'url': img_url,
                            'content_type': content_type
                        })

            except Exception as e:
                print(f"Error checking image URL: {str(e)}")
                continue

            # Add a small delay between checks
            time.sleep(random.uniform(0.2, 0.5))

        return results

    except Exception as e:
        print(f"An error occurred: {str(e)}")
        return []

def get_cover_image(query):
    """Get a high-quality cover image URL for a given query"""
    try:
        # Search for images
        images = search_images(query, num_images=3)  # Get top 3 images to choose from
        
        if not images:
            return None
            
        # Return the first valid image URL
        return images[0]['url']
        
    except Exception as e:
        print(f"Error getting cover image: {str(e)}")
        return None

@app.route('/search_images', methods=['GET'])
def api_search_images():
    try:
        # Get query parameters
        query = request.args.get('query', '')
        num_images = int(request.args.get('num_images', 5))

        if not query:
            return jsonify({'error': 'Query parameter is required'}), 400

        if num_images < 1 or num_images > 20:
            return jsonify({'error': 'Number of images must be between 1 and 20'}), 400

        # Search for images
        results = search_images(query, num_images)

        return jsonify({
            'success': True,
            'query': query,
            'results': results
        })

    except Exception as e:
        return jsonify({
            'success': False,
            'error': str(e)
        }), 500

def scrape_site_content(query, num_sites=5):
    headers = {
        'User-Agent': 'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/91.0.4472.124 Safari/537.36',
        'Accept': 'text/html,application/xhtml+xml,application/xml;q=0.9,*/*;q=0.8',
        'Accept-Language': 'en-US,en;q=0.5',
        'Accept-Encoding': 'gzip, deflate',
        'DNT': '1',
        'Connection': 'keep-alive',
    }

    results = []
    scraped = 0
    retries = 2  # Number of retries per URL
    timeout = 5  # Reduced timeout to 5 seconds

    try:
        # Get more URLs than needed to account for failures
        search_results = list(search(query, num_results=num_sites * 2))
        
        # Process each found URL
        for url in search_results:
            if scraped >= num_sites:
                break

            success = False
            for attempt in range(retries):
                try:
                    # Get the HTML content
                    print(f"Trying {url} (attempt {attempt + 1}/{retries})")
                    response = requests.get(
                        url, 
                        headers=headers, 
                        timeout=timeout,
                        verify=False  # Skip SSL verification
                    )
                    response.raise_for_status()

                    # Verify it's HTML content
                    content_type = response.headers.get('Content-Type', '').lower()
                    if 'text/html' not in content_type:
                        print(f"Skipping {url} - not HTML content")
                        break

                    # Parse the HTML content
                    soup = BeautifulSoup(response.text, 'html.parser')

                    # Remove script and style elements
                    for script in soup(["script", "style"]):
                        script.decompose()

                    # Extract text content (limit to first 10000 characters)
                    text_content = soup.get_text(separator='\n', strip=True)[:10000]
                    
                    # Skip if not enough content
                    if len(text_content.split()) < 100:  # Skip if less than 100 words
                        print(f"Skipping {url} - not enough content")
                        break

                    # Extract all links (limit to first 10)
                    links = []
                    for link in soup.find_all('a', href=True)[:10]:
                        href = link['href']
                        if href.startswith('http'):
                            links.append({
                                'text': link.get_text(strip=True),
                                'url': href
                            })

                    # Extract meta information
                    title = soup.title.string if soup.title else ''
                    meta_description = ''
                    meta_keywords = ''

                    meta_desc_tag = soup.find('meta', attrs={'name': 'description'})
                    if meta_desc_tag:
                        meta_description = meta_desc_tag.get('content', '')

                    meta_keywords_tag = soup.find('meta', attrs={'name': 'keywords'})
                    if meta_keywords_tag:
                        meta_keywords = meta_keywords_tag.get('content', '')

                    results.append({
                        'url': url,
                        'title': title,
                        'meta_description': meta_description,
                        'meta_keywords': meta_keywords,
                        'text_content': text_content,
                        'links': links
                    })

                    scraped += 1
                    success = True
                    # Add a random delay between scrapes
                    time.sleep(random.uniform(0.5, 1))
                    break  # Break retry loop on success

                except requests.Timeout:
                    print(f"Timeout on {url} (attempt {attempt + 1}/{retries})")
                    if attempt == retries - 1:  # Last attempt
                        print(f"Skipping {url} after {retries} timeout attempts")
                except requests.RequestException as e:
                    print(f"Error scraping {url} (attempt {attempt + 1}/{retries}): {str(e)}")
                    if attempt == retries - 1:  # Last attempt
                        print(f"Skipping {url} after {retries} failed attempts")
                
                # Add a longer delay between retries
                if not success and attempt < retries - 1:
                    time.sleep(random.uniform(1, 2))

            # If we haven't found enough valid content and have more URLs, continue
            if scraped < num_sites and len(results) < len(search_results):
                continue

        return results

    except Exception as e:
        print(f"Error in search/scraping process: {str(e)}")
        # Return whatever results we've managed to gather
        return results

@app.route('/scrape_sites', methods=['GET'])
def api_scrape_sites():
    try:
        # Get query parameters
        query = request.args.get('query', '')
        num_sites = int(request.args.get('num_sites', 10))

        if not query:
            return jsonify({'error': 'Query parameter is required'}), 400

        if num_sites < 1 or num_sites > 20:
            return jsonify({'error': 'Number of sites must be between 1 and 20'}), 400

        # Scrape the websites
        results = scrape_site_content(query, num_sites)

        return jsonify({
            'success': True,
            'query': query,
            'results': results
        })

    except Exception as e:
        return jsonify({
            'success': False,
            'error': str(e)
        }), 500

def analyze_with_gpt(scraped_content, research_query):
    """Analyze scraped content using OpenRouter's Gemini model"""
    try:
        headers = {
            'Authorization': f'Bearer {os.getenv("OPENROUTER_API_KEY")}',
            'HTTP-Referer': 'http://localhost:5001',
            'X-Title': 'Research Assistant'
        }

        # Prepare the prompt
        prompt = f"""You are a research assistant analyzing web content to provide comprehensive research.

Research Query: {research_query}

Below is content scraped from various web sources. Analyze this content and provide a detailed, well-structured research response.
Make sure to cite sources when making specific claims.

Scraped Content:
{json.dumps(scraped_content, indent=2)}

Please provide:
1. A comprehensive analysis of the topic
2. Key findings and insights
3. Supporting evidence from the sources
4. Any additional considerations or caveats

Format your response in markdown with proper headings and citations."""

        response = requests.post(
            'https://openrouter.ai/api/v1/chat/completions',
            headers=headers,
            json={
                'model': 'google/gemini-2.0-flash-thinking-exp:free',
                'messages': [{
                    'role': 'user',
                    'content': prompt
                }]
            },
            timeout=60
        )

        if response.status_code != 200:
            raise Exception(f"OpenRouter API error: {response.text}")

        return response.json()['choices'][0]['message']['content']
    except Exception as e:
        print(f"Error in analyze_with_gpt: {str(e)}")
        return f"Error analyzing content: {str(e)}"

def research_topic(query, num_sites=5):
    """Research a topic using web scraping and GPT analysis"""
    try:
        # First get web content using existing scrape_site_content function
        scraped_results = scrape_site_content(query, num_sites)
        
        # Format scraped content for analysis
        formatted_content = []
        for result in scraped_results:
            formatted_content.append({
                'source': result['url'],
                'title': result['title'],
                'content': result['text_content'][:2000],  # Limit content length for GPT
                'meta_info': {
                    'description': result['meta_description'],
                    'keywords': result['meta_keywords']
                }
            })
        
        # Get AI analysis of the scraped content
        analysis = analyze_with_gpt(formatted_content, query)
        
        return {
            'success': True,
            'query': query,
            'analysis': analysis,
            'sources': formatted_content
        }
    except Exception as e:
        return {
            'success': False,
            'error': str(e)
        }

@app.route('/research', methods=['GET'])
def api_research():
    try:
        query = request.args.get('query', '')
        # Always use 5 sites for consistency
        num_sites = 5

        if not query:
            return jsonify({'error': 'Query parameter is required'}), 400

        results = research_topic(query, num_sites)
        return jsonify(results)

    except Exception as e:
        return jsonify({
            'success': False,
            'error': str(e)
        }), 500

if __name__ == '__main__':
    app.run(host='0.0.0.0', port=5000)