优化sehuat

utils/sehuatang/shehuatang_undetected.py

@@ -25,42 +25,16 @@ from PIL import Image as PILImage
 import re
 from PyPDF2 import PdfReader, PdfWriter
 from loguru import logger
-from urllib.parse import urlparse, urljoin
-from requests.adapters import HTTPAdapter
-from urllib3.util import Retry
-
-UNREACHABLE_HOSTS = set()
 
 
-def download_image(url, session, referer=None):
+def download_image(url, session):
     """使用同步的 session 下载图片，确保 Cookie 一致"""
     try:
-        parsed = urlparse(url)
+        if not url.lower().endswith(('.jpg', '.jpeg', '.png')):
-        host = parsed.netloc.lower()
-        if host in UNREACHABLE_HOSTS:
             return None
-        headers = {}
+        response = session.get(url, timeout=15)
-        if referer:
-            headers['Referer'] = referer
-        else:
-            headers['Referer'] = f'{parsed.scheme}://{host}/'
-        response = session.get(url, headers=headers, timeout=10)
         response.raise_for_status()
-        ctype = response.headers.get('Content-Type', '')
-        if 'image' not in ctype.lower() and not url.lower().endswith(('.jpg', '.jpeg', '.png', '.webp', '.gif')):
-            return None
         return BytesIO(response.content)
-    except requests.exceptions.RequestException as e:
-        try:
-            parsed = urlparse(url)
-            host = parsed.netloc.lower()
-            msg = str(e)
-            if 'Network is unreachable' in msg or 'Failed to establish a new connection' in msg:
-                UNREACHABLE_HOSTS.add(host)
-        except Exception:
-            pass
-        logger.warning(f"下载图片失败: {e}")
-        return None
     except Exception as e:
         logger.warning(f"下载图片失败: {e}")
         return None
@@ -159,16 +133,7 @@ def fetch_and_create_pdf(url):
         # 同步 Session
         session = requests.Session()
         ua = driver.execute_script("return navigator.userAgent")
-        session.headers.update({
+        session.headers.update({'User-Agent': ua, 'Referer': 'https://www.sehuatang.net/'})
-            'User-Agent': ua,
-            'Accept': 'image/avif,image/webp,image/apng,image/svg+xml,image/*;q=0.8,*/*;q=0.5',
-            'Accept-Language': 'zh-CN,zh;q=0.9',
-            'Connection': 'keep-alive'
-        })
-        retry = Retry(total=3, connect=3, read=3, backoff_factor=0.5, status_forcelist=[429, 500, 502, 503, 504], allowed_methods=frozenset(['GET']))
-        adapter = HTTPAdapter(max_retries=retry)
-        session.mount('https://', adapter)
-        session.mount('http://', adapter)
         for c in driver.get_cookies():
             session.cookies.set(c['name'], c['value'])
 
@@ -192,23 +157,16 @@ def fetch_and_create_pdf(url):
                     for m in magnets:
                         content.append(Paragraph(f"<b>{m}</b>", normal_style))
 
-                    page_url = resp.url
                     for img_tag in div.find_all('img'):
-                        src = img_tag.get('zoomfile') or img_tag.get('file') or img_tag.get('src') or img_tag.get('data-src')
+                        src = img_tag.get('zoomfile')
-                        if not src:
+                        if src and 'http' in src:
-                            continue
+                            img_io = download_image(src, session)
-                        if src.startswith('//'):
+                            if img_io:
-                            src = 'https:' + src
+                                with PILImage.open(img_io) as p_img:
-                        elif not src.startswith('http'):
+                                    iw, ih = p_img.size
-                            src = urljoin(page_url, src)
+                                    sc = min(max_w / iw, max_h / ih, 1.0)
-                        img_io = download_image(src, session, referer=page_url)
+                                    img_io.seek(0)
-                        if img_io:
+                                    content.append(Image(img_io, width=iw * sc, height=ih * sc))
-                            with PILImage.open(img_io) as p_img:
-                                iw, ih = p_img.size
-                                sc = min(max_w / iw, max_h / ih, 1.0)
-                                img_io.seek(0)
-                                content.append(Image(img_io, width=iw * sc, height=ih * sc))
-                            time.sleep(0.15)
 
                     if post != today_posts[-1]: content.append(PageBreak())
             except Exception as e: