Dockerfile.bak.19

# OpenClaw on Hugging Face Spaces

FROM node:22-slim

# 1. 基础依赖
RUN apt-get update && apt-get install -y --no-install-recommends \
    git openssh-client build-essential python3 python3-pip \
    g++ make ca-certificates curl \
    && rm -rf /var/lib/apt/lists/*

# 2. 安装 Hugging Face Hub
RUN pip3 install --no-cache-dir huggingface_hub --break-system-packages

# 3. Git 配置
RUN update-ca-certificates && \
    git config --global http.sslVerify false && \
    git config --global url."https://github.com/".insteadOf ssh://git@github.com/

# 4. 安装 OpenClaw
RUN npm install -g openclaw@latest --unsafe-perm

# 5. 环境变量预设 - 所有可配置的环境变量，使用 ${VAR:-default} 格式
ENV \
    # 基础配置
    PORT=${PORT:-7860} \
    NODE_ENV=${NODE_ENV:-production} \
    HOME=${HOME:-/root} \
    \
    # OpenClaw 核心配置
    OPENCLAW_GATEWAY_MODE=${OPENCLAW_GATEWAY_MODE:-local} \
    OPENCLAW_GATEWAY_TOKEN=${OPENCLAW_GATEWAY_TOKEN:-} \
    \
    # 模型配置
    MODEL_PROVIDER=${MODEL_PROVIDER:-nvidia} \
    MODEL_ID=${MODEL_ID:-moonshotai/kimi-k2.5} \
    MODEL_CONTEXT_WINDOW=${MODEL_CONTEXT_WINDOW:-256000} \
    MODEL_DISPLAY_NAME=${MODEL_DISPLAY_NAME:-"Kimi K2.5"} \
    \
    # API 配置
    API_BASE_URL=${API_BASE_URL:-https://integrate.api.nvidia.com/v1} \
    API_TYPE=${API_TYPE:-openai-completions} \
    OPENAI_API_KEY=${OPENAI_API_KEY:-} \
    \
    # 网关配置
    GATEWAY_AUTH_MODE=${GATEWAY_AUTH_MODE:-token} \
    GATEWAY_BIND=${GATEWAY_BIND:-lan} \
    GATEWAY_TRUSTED_PROXIES=${GATEWAY_TRUSTED_PROXIES:-0.0.0.0/0,10.0.0.0/8,172.16.0.0/12,192.168.0.0/16} \
    GATEWAY_ALLOWED_ORIGINS=${GATEWAY_ALLOWED_ORIGINS:-https://control.example.com} \
    \
    # 控制UI配置
    CONTROLUI_ALLOW_INSECURE_AUTH=${CONTROLUI_ALLOW_INSECURE_AUTH:-true} \
    CONTROLUI_DANGEROUS_HOST_HEADER=${CONTROLUI_DANGEROUS_HOST_HEADER:-true} \
    CONTROLUI_DANGEROUS_DISABLE_DEVICE_AUTH=${CONTROLUI_DANGEROUS_DISABLE_DEVICE_AUTH:-true} \
    \
    # Telegram 配置
    TELEGRAM_ENABLED=${TELEGRAM_ENABLED:-false} \
    TELEGRAM_BOT_TOKEN=${TELEGRAM_BOT_TOKEN:-} \
    TELEGRAM_DM_POLICY=${TELEGRAM_DM_POLICY:-allowlist} \
    TELEGRAM_ALLOW_FROM=${TELEGRAM_ALLOW_FROM:-} \
    TELEGRAM_PROXY_HOST=${TELEGRAM_PROXY_HOST:-} \
    \
    # 备份配置
    BACKUP_ENABLED=${BACKUP_ENABLED:-true} \
    BACKUP_INTERVAL=${BACKUP_INTERVAL:-21600} \
    BACKUP_RETENTION_DAYS=${BACKUP_RETENTION_DAYS:-5} \
    HF_DATASET=${HF_DATASET:-} \
    HF_TOKEN=${HF_TOKEN:-}

# 6. 同步脚本 - 备份和恢复整个 .openclaw 目录
RUN cat > /usr/local/bin/sync.py << 'SYNC_EOF'
#!/usr/bin/env python3
import os
import sys
import tarfile
import shutil
from huggingface_hub import HfApi, hf_hub_download
from datetime import datetime, timedelta

api = HfApi()
repo_id = os.getenv("HF_DATASET", "")
token = os.getenv("HF_TOKEN", "")
retention_days = int(os.getenv("BACKUP_RETENTION_DAYS", "5"))

OPENCLAW_DIR = "/root/.openclaw"
BACKUP_DIR = "/tmp/openclaw_backups"

def ensure_backup_dir():
    """确保备份临时目录存在"""
    os.makedirs(BACKUP_DIR, exist_ok=True)

def restore():
    """从 Hugging Face Dataset 恢复最新的 .openclaw 目录"""
    if not repo_id or not token:
        print("⚠️  HF_DATASET 或 HF_TOKEN 未设置，跳过恢复")
        return False
    
    if not os.path.exists(OPENCLAW_DIR):
        os.makedirs(OPENCLAW_DIR, mode=0o755, exist_ok=True)
    
    try:
        # 列出数据集中的所有文件
        files = api.list_repo_files(repo_id=repo_id, repo_type="dataset", token=token)
        backup_files = [f for f in files if f.startswith("backup_") and f.endswith(".tar.gz")]
        
        if not backup_files:
            print("ℹ️  没有找到备份文件")
            return False
        
        # 按日期排序，获取最新的备份
        backup_files.sort(reverse=True)
        latest_backup = backup_files[0]
        
        print(f"📥 发现最新备份: {latest_backup}")
        
        # 下载备份文件
        ensure_backup_dir()
        local_backup_path = os.path.join(BACKUP_DIR, latest_backup)
        
        print(f"⬇️  下载备份文件...")
        downloaded_path = hf_hub_download(
            repo_id=repo_id,
            filename=latest_backup,
            repo_type="dataset",
            token=token,
            local_dir=BACKUP_DIR,
            local_dir_use_symlinks=False
        )
        
        # 创建临时恢复目录
        restore_temp = os.path.join(BACKUP_DIR, "restore_temp")
        if os.path.exists(restore_temp):
            shutil.rmtree(restore_temp)
        os.makedirs(restore_temp)
        
        # 解压备份文件
        print(f"📦 解压备份文件...")
        with tarfile.open(downloaded_path, "r:gz") as tar:
            tar.extractall(path=restore_temp)
        
        # 检查解压后的内容
        extracted_items = os.listdir(restore_temp)
        print(f"解压内容: {extracted_items}")
        
        # 情况1: 解压后直接是 .openclaw 目录
        if ".openclaw" in extracted_items:
            source_dir = os.path.join(restore_temp, ".openclaw")
        # 情况2: 解压后是目录内容（sessions, openclaw.json 等）
        elif set(extracted_items) & {"sessions", "openclaw.json", "workspace"}:
            source_dir = restore_temp
        else:
            print(f"❌ 无法识别的备份格式: {extracted_items}")
            return False
        
        # 备份当前目录（如果需要）
        if os.path.exists(OPENCLAW_DIR) and os.listdir(OPENCLAW_DIR):
            backup_current = os.path.join(BACKUP_DIR, "current_before_restore")
            if os.path.exists(backup_current):
                shutil.rmtree(backup_current)
            shutil.copytree(OPENCLAW_DIR, backup_current)
            print(f"💾 已备份当前目录到: {backup_current}")
        
        # 清空并恢复目标目录
        print(f"🔄 恢复数据到 {OPENCLAW_DIR}...")
        if os.path.exists(OPENCLAW_DIR):
            # 删除所有内容但不删除目录本身
            for item in os.listdir(OPENCLAW_DIR):
                item_path = os.path.join(OPENCLAW_DIR, item)
                if os.path.isfile(item_path):
                    os.remove(item_path)
                elif os.path.isdir(item_path):
                    shutil.rmtree(item_path)
        
        # 复制所有文件
        for item in os.listdir(source_dir):
            src = os.path.join(source_dir, item)
            dst = os.path.join(OPENCLAW_DIR, item)
            if os.path.isdir(src):
                shutil.copytree(src, dst)
            else:
                shutil.copy2(src, dst)
        
        print(f"✅ 恢复完成！")
        
        # 清理临时文件
        shutil.rmtree(restore_temp)
        os.remove(downloaded_path)
        
        return True
        
    except Exception as e:
        print(f"❌ 恢复失败: {e}")
        return False

def backup():
    """备份整个 .openclaw 目录到 Hugging Face Dataset"""
    if not repo_id or not token:
        print("⚠️  HF_DATASET 或 HF_TOKEN 未设置，跳过备份")
        return
    
    if not os.path.exists(OPENCLAW_DIR) or not os.listdir(OPENCLAW_DIR):
        print("ℹ️  .openclaw 目录为空，跳过备份")
        return
    
    try:
        # 生成备份文件名
        today = datetime.now().strftime("%Y-%m-%d")
        backup_filename = f"backup_{today}.tar.gz"
        backup_path = os.path.join(BACKUP_DIR, backup_filename)
        
        ensure_backup_dir()
        
        print(f"📦 创建备份: {backup_filename}")
        
        # 创建备份
        with tarfile.open(backup_path, "w:gz") as tar:
            tar.add(OPENCLAW_DIR, arcname=".openclaw")
        
        # 上传到 Hugging Face
        print(f"⬆️  上传到 Hugging Face Dataset: {repo_id}")
        api.upload_file(
            path_or_fileobj=backup_path,
            path_in_repo=backup_filename,
            repo_id=repo_id,
            repo_type="dataset",
            token=token
        )
        
        print(f"✅ 备份完成: {backup_filename}")
        
        # 清理旧备份（可选，保留最近 retention_days 天的备份）
        try:
            files = api.list_repo_files(repo_id=repo_id, repo_type="dataset", token=token)
            backup_files = [f for f in files if f.startswith("backup_") and f.endswith(".tar.gz")]
            backup_files.sort()
            
            # 如果备份文件数量超过保留天数，删除最旧的
            while len(backup_files) > retention_days:
                oldest = backup_files.pop(0)
                print(f"🗑️  删除旧备份: {oldest}")
                api.delete_file(
                    path_in_repo=oldest,
                    repo_id=repo_id,
                    repo_type="dataset",
                    token=token
                )
        except Exception as e:
            print(f"⚠️  清理旧备份失败: {e}")
        
        # 删除本地临时文件
        os.remove(backup_path)
        
    except Exception as e:
        print(f"❌ 备份失败: {e}")

def list_backups():
    """列出所有可用的备份"""
    if not repo_id or not token:
        print("⚠️  HF_DATASET 或 HF_TOKEN 未设置")
        return
    
    try:
        files = api.list_repo_files(repo_id=repo_id, repo_type="dataset", token=token)
        backup_files = [f for f in files if f.startswith("backup_") and f.endswith(".tar.gz")]
        backup_files.sort(reverse=True)
        
        if backup_files:
            print("📋 可用的备份:")
            for i, f in enumerate(backup_files, 1):
                print(f"  {i}. {f}")
        else:
            print("ℹ️  没有找到备份文件")
            
    except Exception as e:
        print(f"❌ 列出备份失败: {e}")

if __name__ == "__main__":
    if len(sys.argv) > 1:
        if sys.argv[1] == "backup":
            backup()
        elif sys.argv[1] == "restore":
            restore()
        elif sys.argv[1] == "list":
            list_backups()
        else:
            print(f"未知命令: {sys.argv[1]}")
            print("可用命令: backup, restore, list")
    else:
        # 默认执行恢复
        restore()
SYNC_EOF
RUN chmod +x /usr/local/bin/sync.py

# 7. Telegram API 替换脚本
RUN cat > /usr/local/bin/patch-telegram-api << 'PATCH_EOF'
#!/bin/bash

# 如果设置了 TELEGRAM_PROXY_HOST，则替换所有 Telegram API 地址
if [ -n "$TELEGRAM_PROXY_HOST" ]; then
    echo "🔧 检测到 TELEGRAM_PROXY_HOST 设置: $TELEGRAM_PROXY_HOST"
    echo "🔄 开始替换 OpenClaw 中的 Telegram API 地址..."
    
    OPENCLAW_DIR="/usr/local/lib/node_modules/openclaw"
    
    if [ -d "$OPENCLAW_DIR" ]; then
        # 统计替换前的匹配数量
        MATCH_COUNT=$(grep -r "api.telegram.org" "$OPENCLAW_DIR" 2>/dev/null | wc -l)
        echo "📊 找到 $MATCH_COUNT 处需要替换的地址"
        
        # 执行替换（使用 sed -i 进行原地替换）
        # 处理 .js 文件
        find "$OPENCLAW_DIR" -type f -name "*.js" -exec sed -i "s|api\\.telegram\\.org|$TELEGRAM_PROXY_HOST|g" {} +
        
        # 再次检查是否还有未替换的（可能是其他格式）
        REMAINING=$(grep -r "api.telegram.org" "$OPENCLAW_DIR" 2>/dev/null | wc -l)
        
        if [ "$REMAINING" -eq 0 ]; then
            echo "✅ Telegram API 地址替换完成！"
            echo "   原始地址: api.telegram.org"
            echo "   新地址: $TELEGRAM_PROXY_HOST"
        else
            echo "⚠️  仍有 $REMAINING 处未替换，尝试二次替换..."
            # 尝试更宽松的匹配（处理可能的转义或编码）
            find "$OPENCLAW_DIR" -type f \( -name "*.js" -o -name "*.json" -o -name "*.ts" \) -exec sed -i "s|api.telegram.org|$TELEGRAM_PROXY_HOST|g" {} +
            
            FINAL_REMAINING=$(grep -r "api.telegram.org" "$OPENCLAW_DIR" 2>/dev/null | wc -l)
            echo "📊 最终剩余未替换: $FINAL_REMAINING 处"
        fi
        
        # 验证替换结果（显示几处示例）
        echo "🔍 验证替换结果（前3处）:"
        grep -r "$TELEGRAM_PROXY_HOST" "$OPENCLAW_DIR" 2>/dev/null | head -3 | sed 's|.*|   &|'
    else
        echo "❌ OpenClaw 目录不存在: $OPENCLAW_DIR"
    fi
else
    echo "ℹ️  未设置 TELEGRAM_PROXY_HOST，跳过 Telegram API 替换"
fi
PATCH_EOF
RUN chmod +x /usr/local/bin/patch-telegram-api

# 8. 启动脚本
RUN cat > /usr/local/bin/start-openclaw << 'START_EOF'
#!/bin/bash
set -e

# 创建必要的目录
mkdir -p /root/.openclaw
mkdir -p /root/.openclaw/sessions
mkdir -p /root/.openclaw/workspace

echo "========================================"
echo "OpenClaw Gateway Starting..."
echo "========================================"

# 尝试恢复数据（如果配置了备份）
if [ -n "$HF_DATASET" ] && [ -n "$HF_TOKEN" ]; then
    echo "🔄 尝试从 Hugging Face 恢复数据..."
    python3 /usr/local/bin/sync.py restore || echo "⚠️  恢复失败，继续启动..."
fi

# 执行 Telegram API 地址替换（如果设置了代理）
/usr/local/bin/patch-telegram-api

# 准备 API 配置
CLEAN_BASE="${API_BASE_URL}"
CLEAN_BASE=$(echo "$CLEAN_BASE" | sed 's|/chat/completions||g' | sed 's|/v1/|/v1|g' | sed 's|/v1$|/v1|')

# 生成令牌（如果没设置）
if [ -z "$OPENCLAW_GATEWAY_TOKEN" ]; then
    OPENCLAW_GATEWAY_TOKEN=$(openssl rand -hex 16)
    echo "🔑 生成的网关令牌: $OPENCLAW_GATEWAY_TOKEN"
fi

# 转换可信代理列表为JSON数组
TRUSTED_PROXIES_JSON=$(echo "$GATEWAY_TRUSTED_PROXIES" | tr ',' '\n' | awk '{ printf "\"%s\",", $0 }' | sed 's/,$//' | sed 's/^/[/' | sed 's/$/]/')

# 转换允许的源列表为JSON数组
ALLOWED_ORIGINS_JSON=$(echo "$GATEWAY_ALLOWED_ORIGINS" | tr ',' '\n' | awk '{ printf "\"%s\",", $0 }' | sed 's/,$//' | sed 's/^/[/' | sed 's/$/]/')

# 转换Telegram允许列表为JSON数组
if [ -n "$TELEGRAM_ALLOW_FROM" ]; then
    # 规范化ID，添加tg:前缀（如果没有）
    TELEGRAM_ALLOW_JSON=$(echo "$TELEGRAM_ALLOW_FROM" | tr ',' '\n' | while read id; do
        id=$(echo "$id" | xargs)  # 去除空格
        if [[ "$id" =~ ^[0-9]+$ ]]; then
            echo "\"tg:$id\""
        elif [[ "$id" =~ ^tg: ]]; then
            echo "\"$id\""
        elif [[ "$id" =~ ^@ ]]; then
            echo "\"$id\""
        else
            echo "\"$id\""
        fi
    done | paste -sd ',' | sed 's/^/[/' | sed 's/$/]/')
else
    TELEGRAM_ALLOW_JSON="[]"
fi

# 创建 OpenClaw 配置（如果配置文件不存在）
if [ ! -f "/root/.openclaw/openclaw.json" ]; then
    echo "📝 创建 OpenClaw 配置文件..."
    cat > /root/.openclaw/openclaw.json <<OPENCLAW_CONFIG
{
  "models": {
    "providers": {
      "${MODEL_PROVIDER}": {
        "baseUrl": "${CLEAN_BASE}",
        "apiKey": "${OPENAI_API_KEY}",
        "api": "${API_TYPE}",
        "models": [{
          "id": "${MODEL_ID}",
          "name": "${MODEL_DISPLAY_NAME}",
          "contextWindow": ${MODEL_CONTEXT_WINDOW}
        }]
      }
    }
  },
  "agents": {
    "defaults": {
      "model": {
        "primary": "${MODEL_PROVIDER}/${MODEL_ID}"
      },
      "workspace": "/root/.openclaw/workspace"
    }
  },
  "gateway": {
    "mode": "${OPENCLAW_GATEWAY_MODE}",
    "bind": "${GATEWAY_BIND}",
    "port": ${PORT},
    "trustedProxies": ${TRUSTED_PROXIES_JSON},
    "auth": {
      "mode": "${GATEWAY_AUTH_MODE}",
      "token": "${OPENCLAW_GATEWAY_TOKEN}"
    },
    "remote": {
      "token": "${OPENCLAW_GATEWAY_TOKEN}"
    },
    "controlUi": {
      "allowInsecureAuth": ${CONTROLUI_ALLOW_INSECURE_AUTH},
      "dangerouslyAllowHostHeaderOriginFallback": ${CONTROLUI_DANGEROUS_HOST_HEADER},
      "dangerouslyDisableDeviceAuth": ${CONTROLUI_DANGEROUS_DISABLE_DEVICE_AUTH},
      "allowedOrigins": ${ALLOWED_ORIGINS_JSON}
    }
  },
  "session": {
    "store": "/root/.openclaw/sessions/sessions.json"
  },
  "channels": {
    "telegram": {
      "enabled": ${TELEGRAM_ENABLED},
      "botToken": "${TELEGRAM_BOT_TOKEN}",
      "dmPolicy": "${TELEGRAM_DM_POLICY}",
      "allowFrom": ${TELEGRAM_ALLOW_JSON}
    }
  }
}
OPENCLAW_CONFIG
else
    echo "📁 使用现有的配置文件"
fi

echo "========================================"
echo "📊 配置信息:"
echo "  • 模型: ${MODEL_ID}"
echo "  • Provider: ${MODEL_PROVIDER}"
echo "  • 端口: ${PORT}"
echo "  • 备份: ${BACKUP_ENABLED}"
echo "  • Allowed Origins: ${GATEWAY_ALLOWED_ORIGINS}"
echo "  • Telegram Enabled: ${TELEGRAM_ENABLED}"
if [ "${TELEGRAM_ENABLED}" = "true" ]; then
    echo "  • Telegram DM Policy: ${TELEGRAM_DM_POLICY}"
    echo "  • Telegram Allow From: ${TELEGRAM_ALLOW_FROM}"
    if [ -n "$TELEGRAM_PROXY_HOST" ]; then
        echo "  • Telegram Proxy: ${TELEGRAM_PROXY_HOST}"
    fi
fi
echo "========================================"

# 后台备份循环（如果启用）
if [ "${BACKUP_ENABLED}" = "true" ] && [ -n "$HF_DATASET" ] && [ -n "$HF_TOKEN" ]; then
    echo "🔄 启动自动备份 (间隔: ${BACKUP_INTERVAL}秒)"
    (
    while true; do
        sleep ${BACKUP_INTERVAL}
        echo "⏰ 执行定时备份..."
        python3 /usr/local/bin/sync.py backup || echo "⚠️  备份失败"
    done
    ) &
fi

# 运行诊断并启动网关
echo "🔧 运行 OpenClaw 诊断..."
openclaw doctor --fix || true

echo "🚀 启动 OpenClaw Gateway..."
echo "========================================"
exec openclaw gateway run --port $PORT
START_EOF
RUN chmod +x /usr/local/bin/start-openclaw

EXPOSE 7860

CMD ["/usr/local/bin/start-openclaw"]