ai-LiaoWangweb-app/app/spider/engine.py

import requests
from bs4 import BeautifulSoup
import json
import urllib.parse
import re
import asyncio
import sys
from crawl4ai import AsyncWebCrawler, BrowserConfig, CrawlerRunConfig, CacheMode
import urllib3

# Disable SSL warnings
urllib3.disable_warnings(urllib3.exceptions.InsecureRequestWarning)

if sys.platform == 'win32':
    asyncio.set_event_loop_policy(asyncio.WindowsProactorEventLoopPolicy())

def clean_text(s):
    s = re.sub(r"\s+", " ", s or "")
    return s.strip()

def absolutize_url(base_url, u):
    if not u:
        return ""
    if u.startswith("//"):
        return "https:" + u
    if u.startswith("/"):
        parsed = urllib.parse.urlparse(base_url)
        return urllib.parse.urljoin(f"{parsed.scheme}://{parsed.netloc}", u)
    if not u.startswith("http"):
        return urllib.parse.urljoin(base_url, u)
    return u

class GenericSpiderEngine:
    def __init__(self, source_config):
        """
        :param source_config: SpiderSource model instance or dictionary
        """
        self.config = source_config

    async def _crawl_with_ai(self, url, wait_for=None):
        browser_config = BrowserConfig(
            headless=True,
            verbose=False,
            java_script_enabled=True,
            channel="chromium",
        )
        run_config = CrawlerRunConfig(
            cache_mode=CacheMode.BYPASS
        )
        if wait_for:
            run_config.wait_for = wait_for
            
        async with AsyncWebCrawler(config=browser_config) as crawler:
            try:
                result = await crawler.arun(
                    url=url,
                    config=run_config
                )
                return result.html
            except Exception as e:
                # If wait_for failed, retry without it but with a delay
                if wait_for and "Wait condition failed" in str(e):
                    print(f"Warning: Wait for selector '{wait_for}' timed out. Retrying with 5s delay...")
                    run_config.wait_for = None
                    # Use js_code to wait for 5 seconds as a fallback
                    run_config.js_code = "await new Promise(r => setTimeout(r, 5000));"
                    
                    result = await crawler.arun(
                        url=url,
                        config=run_config
                    )
                    return result.html
                else:
                    raise e

    def run(self, keyword, limit=10, pages=1):
        url = self.config.url
        method = self.config.method or 'GET'
        headers = json.loads(self.config.headers) if self.config.headers else {
            "User-Agent": "Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/116.0.5845.97 Safari/537.36"
        }
        base_params = json.loads(self.config.params) if self.config.params else {}
        search_key = self.config.search_param_key or 'q'
        
        # Add keyword to params
        base_params[search_key] = keyword

        # Pagination Config
        has_pagination = getattr(self.config, 'has_pagination', False)
        pagination_param = getattr(self.config, 'pagination_param', 'pn')
        pagination_step = getattr(self.config, 'pagination_step', 10)
        pagination_start = getattr(self.config, 'pagination_start', 0)
        
        pages_to_crawl = pages if has_pagination else 1
        all_results = []
        
        for page_idx in range(pages_to_crawl):
            # Calculate current pagination value
            current_pagination_val = pagination_start + (page_idx * pagination_step)
            
            # Update params
            current_params = base_params.copy()
            if has_pagination and pagination_param:
                current_params[pagination_param] = current_pagination_val
                
            print(f"Crawling page {page_idx+1}/{pages_to_crawl} with {pagination_param}={current_pagination_val}")

            # Construct full URL for logging and fallback
            full_url = url
            if current_params:
                query_string = urllib.parse.urlencode(current_params)
                if '?' in full_url:
                    full_url += '&' + query_string
                else:
                    full_url += '?' + query_string

            html_content = ""
            current_url = full_url

            try:
                if method.upper() == 'GET':
                    response = requests.get(url, headers=headers, params=current_params, timeout=15, verify=False)
                else:
                    response = requests.post(url, headers=headers, data=current_params, timeout=15, verify=False)
                
                response.raise_for_status()
                
                # Robust decoding strategy for Chinese websites (Baidu, Gov sites, etc.)
                try:
                    html_content = response.content.decode('utf-8')
                except UnicodeDecodeError:
                    try:
                        html_content = response.content.decode('gb18030')
                    except UnicodeDecodeError:
                        response.encoding = response.apparent_encoding
                        html_content = response.text
                
                current_url = response.url
                
            except Exception as e:
                print(f"Standard fetch failed ({e}). Activating intelligent crawler for {full_url}")
                try:
                    # Extract wait_for selector from config
                    wait_for = None
                    try:
                        if self.config.selectors:
                            sel_json = json.loads(self.config.selectors)
                            wait_for = sel_json.get('list')
                    except Exception as ex:
                        print(f"Error parsing selectors for wait_for: {ex}")

                    html_content = asyncio.run(self._crawl_with_ai(full_url, wait_for=wait_for))
                    current_url = full_url # Crawl4AI result url might be different but we start with full_url
                except Exception as ai_e:
                     print(f"Crawl4AI also failed: {ai_e}")
                     continue # Skip this page
            
            if not html_content:
                continue
                
            page_results = self.parse(html_content, current_url, limit)
            all_results.extend(page_results)
            
            if page_idx < pages_to_crawl - 1:
                import time
                time.sleep(1)

        return all_results

    def parse(self, html, current_url, limit):
        if not self.config.selectors:
            return []
            
        selectors = json.loads(self.config.selectors)
        list_selector = selectors.get('list')
        if not list_selector:
            return []

        soup = BeautifulSoup(html, 'lxml')
        items = []
        containers = soup.select(list_selector)
        
        for container in containers:
            if len(items) >= limit:
                break
            
            item = {}
            # Helper to extract text or attr
            def extract(field, sel_config):
                if not sel_config:
                    return ""
                # sel_config can be "selector" or {"selector": "...", "attr": "..."}
                selector = sel_config
                attr = None
                if isinstance(sel_config, dict):
                    selector = sel_config.get('selector')
                    attr = sel_config.get('attr')
                
                element = container.select_one(selector) if selector else container
                if not element:
                    return ""
                
                if attr:
                    return element.get(attr, "")
                return clean_text(element.get_text())

            item['title'] = extract('title', selectors.get('title'))
            item['link'] = absolutize_url(current_url, extract('link', selectors.get('link')))
            item['abstract'] = extract('abstract', selectors.get('abstract'))
            item['source'] = extract('source', selectors.get('source'))
            item['cover'] = extract('cover', selectors.get('cover'))
            item['date'] = extract('date', selectors.get('date'))
            
            if item['title']:
                items.append(item)
                
        return items

def run_generic_spider(source_model, keyword, limit=10, pages=1):
    engine = GenericSpiderEngine(source_model)
    return engine.run(keyword, limit, pages)