Upload hf_backend/test_same_context.py with huggingface_hub

hf_backend/test_same_context.py

@@ -0,0 +1,181 @@
+#!/usr/bin/env python3
+"""
+Test: Use same browser context for search + slow_download to preserve cookies.
+Also try checking if we can extract final URL from search results directly.
+"""
+
+from playwright.sync_api import sync_playwright
+import time
+import re
+
+MD5 = "d94c20d1364af9b484949659398c4062"
+SLOW_URL = f"https://annas-archive.gl/slow_download/{MD5}/0/3"
+
+def test_same_context_cookies():
+    """Try using cookies from search page to access slow_download."""
+    print("[测试] 在同一浏览器上下文中访问搜索页和slow_download...")
+
+    with sync_playwright() as p:
+        browser = p.chromium.launch(headless=True)
+        context = browser.new_context(
+            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36",
+        )
+        page = context.new_page()
+
+        # First, visit the search page to get cookies
+        print("   1. 访问搜索页面获取cookies...")
+        try:
+            page.goto("https://annas-archive.gl/search?q=Capitalism+A+Global+History", timeout=60000, wait_until="networkidle")
+            print(f"   搜索页标题: {page.title()}")
+            cookies = context.cookies()
+            print(f"   获取到 {len(cookies)} 个 cookies")
+            for c in cookies:
+                print(f"      {c['name']}: {c['value'][:30]}...")
+        except Exception as e:
+            print(f"   搜索页错误: {e}")
+
+        # Now try slow_download in same context
+        print("\n   2. 在同一上下文中访问slow_download...")
+        try:
+            response = page.goto(SLOW_URL, timeout=120000, wait_until="domcontentloaded")
+            print(f"   状态码: {response.status if response else 'None'}")
+
+            for i in range(45):
+                time.sleep(1)
+                url = page.url
+                title = page.title()
+
+                if url.lower().endswith('.epub'):
+                    print(f"   成功! 在第 {i+1} 秒 URL 变成 EPUB")
+                    browser.close()
+                    return url
+
+                if title != "DDoS-Guard":
+                    print(f"   {i+1}秒: 标题={title}, URL={url[:60]}...")
+
+                if (i + 1) % 15 == 0:
+                    print(f"   {i+1}秒...")
+
+        except Exception as e:
+            print(f"   slow_download 错误: {e}")
+
+        browser.close()
+    return None
+
+def test_extract_from_search_md5():
+    """
+    Check if there's a way to get the final URL directly from the MD5.
+    Maybe Anna's Archive has an API or we can construct the URL.
+    """
+    print("\n[测试] 分析MD5哈希看是否能直接构造最终URL...")
+
+    # The MD5 is: d94c20d1364af9b484949659398c4062
+    # This is 32 hex characters = 128 bits
+    # Anna's Archive file IDs might be based on this
+
+    # Try to check what fast_redirect or other endpoints exist
+    print(f"   MD5: {MD5}")
+    print(f"   可能的文件标识符: {MD5}")
+
+    # Try some alternative URL patterns
+    test_urls = [
+        f"https://annas-archive.gl/fast_download/{MD5}/0/3",
+        f"https://annas-archive.gl/download/{MD5}",
+        f"https://annas-archive.gl/get/{MD5}",
+    ]
+
+    for url in test_urls:
+        print(f"\n   测试: {url}")
+        with sync_playwright() as p:
+            browser = p.chromium.launch(headless=True)
+            page = browser.new_page()
+            try:
+                response = page.goto(url, timeout=15000, wait_until="domcontentloaded")
+                print(f"   状态码: {response.status if response else 'None'}")
+                print(f"   最终URL: {page.url}")
+            except Exception as e:
+                print(f"   错误: {str(e)[:50]}")
+            browser.close()
+        time.sleep(1)
+
+    return None
+
+def test_page_source_for_redirect():
+    """Check the actual page source of slow_download to understand redirect mechanism."""
+    print("\n[测试] 检查slow_download页面源码中的重定向机制...")
+
+    with sync_playwright() as p:
+        browser = p.chromium.launch(headless=True)
+        context = browser.new_context(
+            user_agent="Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/123.0.0.0 Safari/537.36",
+        )
+        page = context.new_page()
+
+        try:
+            response = page.goto(SLOW_URL, timeout=30000, wait_until="domcontentloaded")
+            print(f"   状态码: {response.status if response else 'None'}")
+
+            # Get page content
+            content = page.content()
+            print(f"   页面内容长度: {len(content)} 字符")
+
+            # Look for meta refresh, javascript redirects, etc.
+            if 'meta' in content.lower():
+                meta_refresh = re.findall(r'<meta[^>]*refresh[^>]*content=["\']([^"\']*)["\']', content, re.IGNORECASE)
+                if meta_refresh:
+                    print(f"   Meta Refresh: {meta_refresh}")
+
+            if 'window.location' in content.lower():
+                print("   发现 window.location 重定向")
+
+            if 'location.href' in content.lower():
+                print("   发现 location.href 重定向")
+
+            if 'setTimeout' in content.lower():
+                print("   发现 setTimeout 定时器")
+
+            # Look for any URLs in the page
+            urls = re.findall(r'https?://[^\s"\'<>]+\.epub[^\s"\'<>]*', content, re.IGNORECASE)
+            if urls:
+                print(f"   找到EPUB URLs: {urls}")
+
+            # Print relevant parts of the page
+            print("\n   页面内容片段:")
+            lines = content.split('\n')
+            for i, line in enumerate(lines):
+                if any(keyword in line.lower() for keyword in ['refresh', 'location', 'timeout', 'redirect', 'epub', 'download']):
+                    print(f"      {line.strip()[:100]}")
+
+        except Exception as e:
+            print(f"   错误: {e}")
+
+        browser.close()
+
+    return None
+
+def main():
+    print("=" * 60)
+    print("Anna's Archive slow_download 深度测试")
+    print("=" * 60)
+
+    print("\n目标 MD5:", MD5)
+    print("目标 URL:", SLOW_URL)
+
+    # Test 1: Same context cookies
+    result1 = test_same_context_cookies()
+    if result1:
+        print(f"\n*** 成功! 链接: {result1} ***")
+        return result1
+
+    # Test 2: Extract from MD5
+    test_extract_from_search_md5()
+
+    # Test 3: Check page source
+    test_page_source_for_redirect()
+
+    print("\n" + "=" * 60)
+    print("所有测试完成")
+    print("=" * 60)
+
+if __name__ == "__main__":
+    main()