sync_data.sh

#!/bin/bash

# 检查必要的环境变量
if [ -z "$WEBDAV_URL" ] || [ -z "$WEBDAV_USERNAME" ] || [ -z "$WEBDAV_PASSWORD" ]; then
    echo "缺少必要的环境变量 WEBDAV_URL、WEBDAV_USERNAME 或 WEBDAV_PASSWORD"
    exit 1
fi

if [ -z "$HF_TOKEN" ] || [ -z "$DATASET_ID" ]; then
    echo "缺少必要的环境变量 HF_TOKEN 或 DATASET_ID"
    exit 1
fi

# 创建数据目录
mkdir -p ./data

# 创建 HuggingFace 同步脚本
cat > /tmp/hf_sync.py << 'EOL'
from huggingface_hub import HfApi
import sys
import os

def manage_backups(api, repo_id, max_files=50):
    files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
    backup_files = [f for f in files if f.startswith('webui_backup_') and f.endswith('.db')]
    backup_files.sort()
    
    if len(backup_files) >= max_files:
        files_to_delete = backup_files[:(len(backup_files) - max_files + 1)]
        for file_to_delete in files_to_delete:
            try:
                api.delete_file(path_in_repo=file_to_delete, repo_id=repo_id, repo_type="dataset")
                print(f'已删除旧备份: {file_to_delete}')
            except Exception as e:
                print(f'删除 {file_to_delete} 时出错: {str(e)}')

def upload_backup(file_path, file_name, token, repo_id):
    api = HfApi(token=token)
    try:
        api.upload_file(
            path_or_fileobj=file_path,
            path_in_repo=file_name,
            repo_id=repo_id,
            repo_type="dataset"
        )
        print(f"成功上传 {file_name}")
        
        manage_backups(api, repo_id)
    except Exception as e:
        print(f"文件上传出错: {str(e)}")

# 下载最新备份
def download_latest_backup(token, repo_id):
    try:
        api = HfApi(token=token)
        files = api.list_repo_files(repo_id=repo_id, repo_type="dataset")
        backup_files = [f for f in files if f.startswith('webui_backup_') and f.endswith('.db')]
        
        if not backup_files:
            print("未找到备份文件")
            return False
            
        latest_backup = sorted(backup_files)[-1]
        
        filepath = api.hf_hub_download(
            repo_id=repo_id,
            filename=latest_backup,
            repo_type="dataset"
        )
        
        if filepath and os.path.exists(filepath):
            os.makedirs('./data', exist_ok=True)
            os.system(f'cp "{filepath}" ./data/webui.db')
            print(f"成功从 {latest_backup} 恢复备份")
            return True
        return False
                
    except Exception as e:
        print(f"下载备份时出错: {str(e)}")
        return False

if __name__ == "__main__":
    action = sys.argv[1]
    token = sys.argv[2]
    repo_id = sys.argv[3]
    
    if action == "upload":
        file_path = sys.argv[4]
        file_name = sys.argv[5]
        upload_backup(file_path, file_name, token, repo_id)
    elif action == "download":
        download_latest_backup(token, repo_id)
EOL

# 首次启动时的数据恢复策略
echo "开始初始化数据恢复..."

# 首先尝试从 WebDAV 恢复最新文件
echo "正在尝试从 WebDAV 获取文件列表..."

# 获取 WebDAV 目录中的文件列表并找出最新的备份文件
webdav_files=$(curl -s -X PROPFIND --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" -H "Depth: 1" "$WEBDAV_URL/openwebui/" | grep -o '<d:href>[^<]*webui_[0-9][0-9]-[0-9][0-9]\.db</d:href>' | sed 's/<d:href>//g' | sed 's/<\/d:href>//g')

if [ -n "$webdav_files" ]; then
    # 找出最新的文件 (按文件名排序，取最后一个)
    latest_file=$(echo "$webdav_files" | sort | tail -n 1)
    echo "找到最新的 WebDAV 备份文件: $latest_file"
    
    # 下载最新文件
    curl -L --fail --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL$latest_file" -o "./data/webui.db" && {
        echo "成功从 WebDAV 下载最新数据库文件: $latest_file"
        # 上传到 HuggingFace
        timestamp=$(date +%Y%m%d_%H%M%S)
        backup_file="webui_backup_${timestamp}.db"
        cp ./data/webui.db "/tmp/${backup_file}"
        echo "正在上传初始备份到 HuggingFace..."
        python3 /tmp/hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
        rm -f "/tmp/${backup_file}"
    } || {
        echo "从 WebDAV 下载最新文件失败，尝试下载 WebDAV 的 webui.db..."
        
        curl -L --fail --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL/openwebui/webui.db" -o "./data/webui.db" && {
            echo "成功从 WebDAV 下载主数据库文件"
            # 上传到 HuggingFace
            timestamp=$(date +%Y%m%d_%H%M%S)
            backup_file="webui_backup_${timestamp}.db"
            cp ./data/webui.db "/tmp/${backup_file}"
            echo "正在上传初始备份到 HuggingFace..."
            python3 /tmp/hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
            rm -f "/tmp/${backup_file}"
        } || {
            echo "从 WebDAV 下载失败，尝试从 HuggingFace 恢复..."
            python3 /tmp/hf_sync.py download "${HF_TOKEN}" "${DATASET_ID}" || {
                echo "所有恢复方式均失败，将使用空数据库开始"
            }
        }
    }
else
    echo "WebDAV 中没有找到符合格式的备份文件，尝试下载 webui.db..."
    curl -L --fail --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL/openwebui/webui.db" -o "./data/webui.db" && {
        echo "成功从 WebDAV 下载主数据库文件"
        # 上传到 HuggingFace
        timestamp=$(date +%Y%m%d_%H%M%S)
        backup_file="webui_backup_${timestamp}.db"
        cp ./data/webui.db "/tmp/${backup_file}"
        echo "正在上传初始备份到 HuggingFace..."
        python3 /tmp/hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
        rm -f "/tmp/${backup_file}"
    } || {
        echo "从 WebDAV 下载失败，尝试从 HuggingFace 恢复..."
        python3 /tmp/hf_sync.py download "${HF_TOKEN}" "${DATASET_ID}" || {
            echo "所有恢复方式均失败，将使用空数据库开始"
        }
    }
fi

# 定义 WebDAV 同步函数
webdav_sync() {
    SYNC_INTERVAL=${SYNC_INTERVAL:-7200}  # 默认间隔时间为 7200 秒
    echo "WebDAV 同步进程启动，等待 ${SYNC_INTERVAL} 秒后开始同步..."
    sleep $SYNC_INTERVAL

    while true; do
        echo "开始 WebDAV 同步 $(date)"
        
        # 检查数据库文件是否存在
        if [ -f "./data/webui.db" ]; then
            # 生成当前时间的文件名
            FILENAME="webui_$(date +'%m-%d_%H%M').db"
            
            echo "同步到 WebDAV..."

            # 上传以日期命名的数据库文件
            curl -L -T "./data/webui.db" --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL/openwebui/$FILENAME" && {
                echo "WebDAV 上传成功: $FILENAME"
                
                # 覆盖Webdav目录下默认的webui.db文件
                curl -L -T "./data/webui.db" --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL/openwebui/webui.db" && {
                    echo "WebDAV 更新主文件成功"
                } || {
                    echo "WebDAV 更新主文件失败"
                }
            } || {
                echo "WebDAV 上传失败，等待重试..."
                sleep 10
                curl -L -T "./data/webui.db" --user "$WEBDAV_USERNAME:$WEBDAV_PASSWORD" "$WEBDAV_URL/openwebui/$FILENAME" || {
                    echo "重试失败，放弃上传。"
                }
            }
        else
            echo "未找到 webui.db 文件，跳过 WebDAV 同步"
        fi

        # 等待下一次同步间隔
        echo "WebDAV 同步完成，下次同步将在 ${SYNC_INTERVAL} 秒后进行..."
        sleep $SYNC_INTERVAL
    done
}

# 定义 HuggingFace 同步函数
hf_sync() {
    SYNC_INTERVAL=${SYNC_INTERVAL:-7200}
    echo "HuggingFace 同步进程启动，等待 ${SYNC_INTERVAL} 秒后开始同步..."
    sleep $(($SYNC_INTERVAL / 2))  # 错开与 WebDAV 同步的时间

    while true; do
        echo "开始 HuggingFace 同步 $(date)"
        
        if [ -f "./data/webui.db" ]; then
            timestamp=$(date +%Y%m%d_%H%M%S)
            backup_file="webui_backup_${timestamp}.db"
            
            # 复制数据库文件
            cp ./data/webui.db "/tmp/${backup_file}"
            
            echo "正在上传备份到 HuggingFace..."
            python3 /tmp/hf_sync.py upload "${HF_TOKEN}" "${DATASET_ID}" "/tmp/${backup_file}" "${backup_file}"
            
            rm -f "/tmp/${backup_file}"
        else
            echo "数据库文件不存在，跳过 HuggingFace 同步"
        fi
        
        echo "HuggingFace 同步完成，下次同步将在 ${SYNC_INTERVAL} 秒后进行..."
        sleep $SYNC_INTERVAL
    done
}

# 后台启动同步进程
webdav_sync &
hf_sync &