fix: 修复各个代码文件中对action参数的调用

- 将FLOW_GENERATION替换为IMAGE_GENERATION/VIDEO_GENERATION
- browser_captcha_personal.py: get_token/execute方法支持action参数
- flow_client.py: _get_api_captcha_token支持动态action
- 更新数据库和模型的默认值
- 添加GitHub Actions工作流用于构建ghcr.io镜像

Co-Authored-By: Warp <agent@warp.dev>

.github/workflows/docker-publish.yml

@@ -0,0 +1,64 @@
+name: Build and Push Docker Image
+
+on:
+  push:
+    branches:
+      - main
+    tags:
+      - 'v*'
+  pull_request:
+    branches:
+      - main
+  workflow_dispatch:
+
+env:
+  REGISTRY: ghcr.io
+  IMAGE_NAME: ${{ github.repository }}
+
+jobs:
+  build-and-push:
+    runs-on: ubuntu-latest
+    permissions:
+      contents: read
+      packages: write
+
+    steps:
+      - name: Checkout repository
+        uses: actions/checkout@v4
+
+      - name: Set up QEMU
+        uses: docker/setup-qemu-action@v3
+
+      - name: Set up Docker Buildx
+        uses: docker/setup-buildx-action@v3
+
+      - name: Log in to Container Registry
+        if: github.event_name != 'pull_request'
+        uses: docker/login-action@v3
+        with:
+          registry: ${{ env.REGISTRY }}
+          username: ${{ github.actor }}
+          password: ${{ secrets.GITHUB_TOKEN }}
+
+      - name: Extract metadata (tags, labels)
+        id: meta
+        uses: docker/metadata-action@v5
+        with:
+          images: ${{ env.REGISTRY }}/${{ env.IMAGE_NAME }}
+          tags: |
+            type=ref,event=branch
+            type=ref,event=pr
+            type=semver,pattern={{version}}
+            type=semver,pattern={{major}}.{{minor}}
+            type=raw,value=latest,enable={{is_default_branch}}
+
+      - name: Build and push Docker image
+        uses: docker/build-push-action@v5
+        with:
+          context: .
+          platforms: linux/amd64,linux/arm64
+          push: ${{ github.event_name != 'pull_request' }}
+          tags: ${{ steps.meta.outputs.tags }}
+          labels: ${{ steps.meta.outputs.labels }}
+          cache-from: type=gha
+          cache-to: type=gha,mode=max

.gitignore

@@ -57,4 +57,5 @@ browser_data
 
 data
 config/setting.toml
-config/setting_warp.toml
+config/setting_warp.toml
+config/setting_warp_example.toml

Dockerfile

@@ -2,40 +2,9 @@ FROM python:3.11-slim
 
 WORKDIR /app
 
-# 使用清华镜像源加速 apt (Debian bookworm)
-RUN sed -i 's|deb.debian.org|mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list.d/debian.sources \
-    && sed -i 's|security.debian.org|mirrors.tuna.tsinghua.edu.cn|g' /etc/apt/sources.list.d/debian.sources
-
-# 安装 Playwright 所需的系统依赖
-RUN apt-get update && apt-get install -y \
-    libnss3 \
-    libnspr4 \
-    libatk1.0-0 \
-    libatk-bridge2.0-0 \
-    libcups2 \
-    libdrm2 \
-    libxkbcommon0 \
-    libxcomposite1 \
-    libxdamage1 \
-    libxfixes3 \
-    libxrandr2 \
-    libgbm1 \
-    libasound2 \
-    libpango-1.0-0 \
-    libcairo2 \
-    && rm -rf /var/lib/apt/lists/*
-
-# 安装 Python 依赖（使用清华 PyPI 镜像）
+# 安装 Python 依赖
 COPY requirements.txt .
-RUN pip install --no-cache-dir -r requirements.txt \
-    -i https://pypi.tuna.tsinghua.edu.cn/simple/ \
-    --trusted-host pypi.tuna.tsinghua.edu.cn
-
-# 设置 Playwright 下载镜像（使用 npmmirror）
-ENV PLAYWRIGHT_DOWNLOAD_HOST=https://registry.npmmirror.com/-/binary/playwright
-
-# 安装 Playwright 浏览器
-RUN playwright install chromium
+RUN pip install --no-cache-dir --root-user-action=ignore -r requirements.txt
 
 COPY . .
 

config/setting.toml

@@ -12,7 +12,7 @@ max_poll_attempts = 200
 
 [server]
 host = "0.0.0.0"
-port = 18282
+port = 8000
 
 [debug]
 enabled = false
@@ -21,8 +21,8 @@ log_responses = true
 mask_token = true
 
 [proxy]
-proxy_enabled = true
-proxy_url = "http://localhost:7897"
+proxy_enabled = false
+proxy_url = ""
 
 [generation]
 image_timeout = 300

config/setting_example.toml

@@ -12,7 +12,7 @@ max_poll_attempts = 200
 
 [server]
 host = "0.0.0.0"
-port = 18282
+port = 8000
 
 [debug]
 enabled = false
@@ -21,8 +21,8 @@ log_responses = true
 mask_token = true
 
 [proxy]
-proxy_enabled = true
-proxy_url = "http://localhost:7897"
+proxy_enabled = false
+proxy_url = ""
 
 [generation]
 image_timeout = 300

config/setting_warp.toml

@@ -1,42 +0,0 @@
-[global]
-api_key = "han1234"
-admin_username = "admin"
-admin_password = "admin"
-
-[flow]
-labs_base_url = "https://labs.google/fx/api"
-api_base_url = "https://aisandbox-pa.googleapis.com/v1"
-timeout = 120
-poll_interval = 3.0
-max_poll_attempts = 200
-
-[server]
-host = "0.0.0.0"
-port = 8000
-
-[debug]
-enabled = false
-log_requests = true
-log_responses = true
-mask_token = true
-
-[proxy]
-proxy_enabled = true
-proxy_url = "socks5://warp:1080"
-
-[generation]
-image_timeout = 300
-video_timeout = 1500
-
-[admin]
-error_ban_threshold = 3
-
-[cache]
-enabled = false
-timeout = 7200  # 缓存超时时间(秒), 默认2小时
-base_url = ""   # 缓存文件访问的基础URL, 留空则使用服务器地址
-
-[captcha]
-captcha_method = "browser"  # 打码方式: yescaptcha 或 browser
-yescaptcha_api_key = ""  # YesCaptcha API密钥
-yescaptcha_base_url = "https://api.yescaptcha.com"

config/setting_warp_example.toml

@@ -1,42 +0,0 @@
-[global]
-api_key = "han1234"
-admin_username = "admin"
-admin_password = "admin"
-
-[flow]
-labs_base_url = "https://labs.google/fx/api"
-api_base_url = "https://aisandbox-pa.googleapis.com/v1"
-timeout = 120
-poll_interval = 3.0
-max_poll_attempts = 200
-
-[server]
-host = "0.0.0.0"
-port = 8000
-
-[debug]
-enabled = false
-log_requests = true
-log_responses = true
-mask_token = true
-
-[proxy]
-proxy_enabled = true
-proxy_url = "socks5://warp:1080"
-
-[generation]
-image_timeout = 300
-video_timeout = 1500
-
-[admin]
-error_ban_threshold = 3
-
-[cache]
-enabled = false
-timeout = 7200  # 缓存超时时间(秒), 默认2小时
-base_url = ""   # 缓存文件访问的基础URL, 留空则使用服务器地址
-
-[captcha]
-captcha_method = "browser"  # 打码方式: yescaptcha 或 browser
-yescaptcha_api_key = ""  # YesCaptcha API密钥
-yescaptcha_base_url = "https://api.yescaptcha.com"

request.py

@@ -1,150 +0,0 @@
-import os
-import json
-import re
-import base64
-import aiohttp # Async test. Need to install
-import asyncio
-
-
-# --- 配置区域 ---
-BASE_URL = os.getenv('GEMINI_FLOW2API_URL', 'http://127.0.0.1:8000')
-BACKEND_URL = BASE_URL + "/v1/chat/completions"
-API_KEY = os.getenv('GEMINI_FLOW2API_APIKEY', 'Bearer han1234')
-if API_KEY is None:
-    raise ValueError('[gemini flow2api] api key not set')
-MODEL_LANDSCAPE = "gemini-3.0-pro-image-landscape"
-MODEL_PORTRAIT = "gemini-3.0-pro-image-portrait"
-
-# 修改: 增加 model 参数，默认为 None
-async def request_backend_generation(
-        prompt: str,
-        images: list[bytes] = None,
-        model: str = None) -> bytes | None:
-    """
-    请求后端生成图片。
-    :param prompt: 提示词
-    :param images: 图片二进制列表
-    :param model: 指定模型名称 (可选)
-    :return: 成功返回图片bytes，失败返回None
-    """
-    # 更新token
-    images = images or []
-    
-    # 逻辑: 如果未指定 model，默认使用 Landscape
-    use_model = model if model else MODEL_LANDSCAPE
-
-    # 1. 构造 Payload
-    if images:
-        content_payload = [{"type": "text", "text": prompt}]
-        print(f"[Backend] 正在处理 {len(images)} 张图片输入...")
-        for img_bytes in images:
-            b64_str = base64.b64encode(img_bytes).decode('utf-8')
-            content_payload.append({
-                "type": "image_url",
-                "image_url": {"url": f"data:image/jpeg;base64,{b64_str}"}
-            })
-    else:
-        content_payload = prompt
-
-    payload = {
-        "model": use_model,  # 使用选定的模型
-        "messages": [{"role": "user", "content": content_payload}],
-        "stream": True
-    }
-    
-    headers = {
-        "Authorization": API_KEY,
-        "Content-Type": "application/json"
-    }
-
-    image_url = None
-    print(f"[Backend] Model: {use_model} | 发起请求: {prompt[:20]}...") 
-    
-    try:
-        async with aiohttp.ClientSession() as session:
-            async with session.post(BACKEND_URL, json=payload, headers=headers, timeout=120) as response:
-                if response.status != 200:
-                    err_text = await response.text()
-                    content = response.content
-                    print(f"[Backend Error] Status {response.status}: {err_text} {content}")
-                    raise Exception(f"API Error: {response.status}: {err_text}")
-
-                async for line in response.content:
-                    line_str = line.decode('utf-8').strip()
-                    if line_str.startswith('{"error'):
-                        chunk = json.loads(data_str)
-                        delta = chunk.get("choices", [{}])[0].get("delta", {})
-                        msg = delta['reasoning_content']
-                        if '401' in msg:
-                            msg += '\nAccess Token 已失效，需重新配置。'
-                        elif '400' in msg:
-                            msg += '\n返回内容被拦截。'
-                        raise Exception(msg)
-
-                    if not line_str or not line_str.startswith('data: '):
-                        continue
-                    
-                    data_str = line_str[6:]
-                    if data_str == '[DONE]':
-                        break
-                    
-                    try:
-                        chunk = json.loads(data_str)
-                        delta = chunk.get("choices", [{}])[0].get("delta", {})
-                        
-                        # 打印思考过程
-                        if "reasoning_content" in delta:
-                            print(delta['reasoning_content'], end="", flush=True)
-
-                        # 提取内容中的图片链接
-                        if "content" in delta:
-                            content_text = delta["content"]
-                            img_match = re.search(r'!\[.*?\]\((.*?)\)', content_text)
-                            if img_match:
-                                image_url = img_match.group(1)
-                                print(f"\n[Backend] 捕获图片链接: {image_url}")
-                    except json.JSONDecodeError:
-                        continue
-            
-            # 3. 下载生成的图片
-            if image_url:
-                async with session.get(image_url) as img_resp:
-                    if img_resp.status == 200:
-                        image_bytes = await img_resp.read()
-                        return image_bytes
-                    else:
-                        print(f"[Backend Error] 图片下载失败: {img_resp.status}")
-    except Exception as e:
-        print(f"[Backend Exception] {e}")
-        raise e 
-        
-    return None
-
-if __name__ == '__main__':
-    async def main():
-        print("=== AI 绘图接口测试 ===")
-        user_prompt = input("请输入提示词 (例如 '一只猫'): ").strip()
-        if not user_prompt:
-            user_prompt = "A cute cat in the garden"
-        
-        print(f"正在请求: {user_prompt}")
-        
-        # 这里的 images 传空列表用于测试文生图
-        # 如果想测试图生图，你需要手动读取本地文件：
-        # with open("output_test.jpg", "rb") as f: img_data = f.read()
-        # result = await request_backend_generation(user_prompt, [img_data])
-        
-        result = await request_backend_generation(user_prompt)
-        
-        if result:
-            filename = "output_test.jpg"
-            with open(filename, "wb") as f:
-                f.write(result)
-            print(f"\n[Success] 图片已保存为 {filename}，大小: {len(result)} bytes")
-        else:
-            print("\n[Failed] 生成失败")
-
-    # 运行测试
-    if os.name == 'nt':  # Windows 兼容性
-        asyncio.set_event_loop_policy(asyncio.WindowsSelectorEventLoopPolicy())
-    asyncio.run(main())

src/core/database.py

@@ -223,7 +223,7 @@ class Database:
                         capsolver_api_key TEXT DEFAULT '',
                         capsolver_base_url TEXT DEFAULT 'https://api.capsolver.com',
                         website_key TEXT DEFAULT '6LdsFiUsAAAAAIjVDZcuLhaHiDn5nnHVXVRQGeMV',
-                        page_action TEXT DEFAULT 'FLOW_GENERATION',
+                        page_action TEXT DEFAULT 'IMAGE_GENERATION',
                         browser_proxy_enabled BOOLEAN DEFAULT 0,
                         browser_proxy_url TEXT,
                         created_at TIMESTAMP DEFAULT CURRENT_TIMESTAMP,
@@ -509,7 +509,8 @@ class Database:
                     capsolver_api_key TEXT DEFAULT '',
                     capsolver_base_url TEXT DEFAULT 'https://api.capsolver.com',
                     website_key TEXT DEFAULT '6LdsFiUsAAAAAIjVDZcuLhaHiDn5nnHVXVRQGeMV',
-                    page_action TEXT DEFAULT 'FLOW_GENERATION',
+                    page_action TEXT DEFAULT 'IMAGE_GENERATION',
+
                     browser_proxy_enabled BOOLEAN DEFAULT 0,
                     browser_proxy_url TEXT,
                     browser_count INTEGER DEFAULT 1,

src/core/models.py

@@ -157,7 +157,7 @@ class CaptchaConfig(BaseModel):
     capsolver_api_key: str = ""
     capsolver_base_url: str = "https://api.capsolver.com"
     website_key: str = "6LdsFiUsAAAAAIjVDZcuLhaHiDn5nnHVXVRQGeMV"
-    page_action: str = "FLOW_GENERATION"
+    page_action: str = "IMAGE_GENERATION"
     browser_proxy_enabled: bool = False  # 浏览器打码是否启用代理
     browser_proxy_url: Optional[str] = None  # 浏览器打码代理URL
     browser_count: int = 1  # 浏览器打码实例数量

src/services/browser_captcha_personal.py

@@ -251,11 +251,12 @@ class BrowserCaptchaService:
         debug_logger.log_warning("[BrowserCaptcha] reCAPTCHA 加载超时")
         return False
 
-    async def _execute_recaptcha_on_tab(self, tab) -> Optional[str]:
+    async def _execute_recaptcha_on_tab(self, tab, action: str = "IMAGE_GENERATION") -> Optional[str]:
         """在指定标签页执行 reCAPTCHA 获取 token
         
         Args:
             tab: nodriver 标签页对象
+            action: reCAPTCHA action类型 (IMAGE_GENERATION 或 VIDEO_GENERATION)
             
         Returns:
             reCAPTCHA token 或 None
@@ -272,7 +273,7 @@ class BrowserCaptchaService:
                 
                 try {{
                     grecaptcha.enterprise.ready(function() {{
-                        grecaptcha.enterprise.execute('{self.website_key}', {{action: 'FLOW_GENERATION'}})
+                        grecaptcha.enterprise.execute('{self.website_key}', {{action: '{action}'}})
                             .then(function(token) {{
                                 window.{token_var} = token;
                             }})
@@ -311,13 +312,16 @@ class BrowserCaptchaService:
 
     # ========== 主要 API ==========
 
-    async def get_token(self, project_id: str) -> Optional[str]:
+    async def get_token(self, project_id: str, action: str = "IMAGE_GENERATION") -> Optional[str]:
         """获取 reCAPTCHA token
         
         自动常驻模式：如果该 project_id 没有常驻标签页，则自动创建并常驻
         
         Args:
             project_id: Flow项目ID
+            action: reCAPTCHA action类型
+                - IMAGE_GENERATION: 图片生成和2K/4K图片放大 (默认)
+                - VIDEO_GENERATION: 视频生成和视频放大
 
         Returns:
             reCAPTCHA token字符串，如果获取失败返回None
@@ -335,16 +339,16 @@ class BrowserCaptchaService:
                 resident_info = await self._create_resident_tab(project_id)
                 if resident_info is None:
                     debug_logger.log_warning(f"[BrowserCaptcha] 无法为 project_id={project_id} 创建常驻标签页，fallback 到传统模式")
-                    return await self._get_token_legacy(project_id)
+                    return await self._get_token_legacy(project_id, action)
                 self._resident_tabs[project_id] = resident_info
                 debug_logger.log_info(f"[BrowserCaptcha] ✅ 已为 project_id={project_id} 创建常驻标签页 (当前共 {len(self._resident_tabs)} 个)")
         
         # 使用常驻标签页生成 token
         if resident_info and resident_info.recaptcha_ready and resident_info.tab:
             start_time = time.time()
-            debug_logger.log_info(f"[BrowserCaptcha] 从常驻标签页即时生成 token (project: {project_id})...")
+            debug_logger.log_info(f"[BrowserCaptcha] 从常驻标签页即时生成 token (project: {project_id}, action: {action})...")
             try:
-                token = await self._execute_recaptcha_on_tab(resident_info.tab)
+                token = await self._execute_recaptcha_on_tab(resident_info.tab, action)
                 duration_ms = (time.time() - start_time) * 1000
                 if token:
                     debug_logger.log_info(f"[BrowserCaptcha] ✅ Token生成成功（耗时 {duration_ms:.0f}ms）")
@@ -362,7 +366,7 @@ class BrowserCaptchaService:
                     self._resident_tabs[project_id] = resident_info
                     # 重建后立即尝试生成
                     try:
-                        token = await self._execute_recaptcha_on_tab(resident_info.tab)
+                        token = await self._execute_recaptcha_on_tab(resident_info.tab, action)
                         if token:
                             debug_logger.log_info(f"[BrowserCaptcha] ✅ 重建后 Token生成成功")
                             return token
@@ -371,7 +375,7 @@ class BrowserCaptchaService:
         
         # 最终 Fallback: 使用传统模式
         debug_logger.log_warning(f"[BrowserCaptcha] 所有常驻方式失败，fallback 到传统模式 (project: {project_id})")
-        return await self._get_token_legacy(project_id)
+        return await self._get_token_legacy(project_id, action)
 
     async def _create_resident_tab(self, project_id: str) -> Optional[ResidentTabInfo]:
         """为指定 project_id 创建常驻标签页
@@ -449,11 +453,12 @@ class BrowserCaptchaService:
             except Exception as e:
                 debug_logger.log_warning(f"[BrowserCaptcha] 关闭标签页时异常: {e}")
 
-    async def _get_token_legacy(self, project_id: str) -> Optional[str]:
+    async def _get_token_legacy(self, project_id: str, action: str = "IMAGE_GENERATION") -> Optional[str]:
         """传统模式获取 reCAPTCHA token（每次创建新标签页）
 
         Args:
             project_id: Flow项目ID
+            action: reCAPTCHA action类型 (IMAGE_GENERATION 或 VIDEO_GENERATION)
 
         Returns:
             reCAPTCHA token字符串，如果获取失败返回None
@@ -491,8 +496,8 @@ class BrowserCaptchaService:
                 return None
 
             # 执行 reCAPTCHA
-            debug_logger.log_info("[BrowserCaptcha] [Legacy] 执行 reCAPTCHA 验证...")
-            token = await self._execute_recaptcha_on_tab(tab)
+            debug_logger.log_info(f"[BrowserCaptcha] [Legacy] 执行 reCAPTCHA 验证 (action: {action})...")
+            token = await self._execute_recaptcha_on_tab(tab, action)
 
             duration_ms = (time.time() - start_time) * 1000
 

src/services/flow_client.py

@@ -1177,14 +1177,20 @@ class FlowClient:
                 return None, None
         # API打码服务
         elif captcha_method in ["yescaptcha", "capmonster", "ezcaptcha", "capsolver"]:
-            token = await self._get_api_captcha_token(captcha_method, project_id)
+            token = await self._get_api_captcha_token(captcha_method, project_id, action)
             return token, None
         else:
             debug_logger.log_info(f"[reCAPTCHA] 未知的打码方式: {captcha_method}")
             return None, None
 
-    async def _get_api_captcha_token(self, method: str, project_id: str) -> Optional[str]:
-        """通用API打码服务"""
+    async def _get_api_captcha_token(self, method: str, project_id: str, action: str = "IMAGE_GENERATION") -> Optional[str]:
+        """通用API打码服务
+        
+        Args:
+            method: 打码服务类型
+            project_id: 项目ID
+            action: reCAPTCHA action类型 (IMAGE_GENERATION 或 VIDEO_GENERATION)
+        """
         # 获取配置
         if method == "yescaptcha":
             client_key = config.yescaptcha_api_key
@@ -1212,7 +1218,7 @@ class FlowClient:
 
         website_key = "6LdsFiUsAAAAAIjVDZcuLhaHiDn5nnHVXVRQGeMV"
         website_url = f"https://labs.google/fx/tools/flow/project/{project_id}"
-        page_action = "FLOW_GENERATION"
+        page_action = action
 
         try:
             async with AsyncSession() as session: