将Sqlite数据写入Sqlserver

Posted 2020-09-25 Coder_小菜

       /// <summary>
        /// 上传本地信息至服务器
        /// </summary>
        public static void UploadDataToServer()
        {
            try
            {
                #region 1.读取本地数据
                StringBuilder sbSelect = new StringBuilder();
                foreach (string tblName in CMMConstants.UPLOADTABLENAMES)
                {
                    // 查询未上传的数据
                    sbSelect.Append(string.Format(" SELECT * FROM {0} WHERE IsUpload=‘0‘ OR IsUpload IS NULL;", tblName));
                }
                SQLiteDBManager _SQLiteDBManager = new SQLiteDBManager();
                DataSet ds = _SQLiteDBManager.GetDataSet(sbSelect.ToString(), null);
                for (int i = 0; i < CMMConstants.UPLOADTABLENAMES.Length; i++)
                {
                    // 设置DataTableName
                    ds.Tables[i].TableName = CMMConstants.UPLOADTABLENAMES[i];
                }
                #endregion
                #region 2.上传至服务器
                // 服务器数据访问
                DBManager _DBManager = new DBManager();

                // 获取服务器表的字段列表
                DataSet dsColumns = _DBManager.GetColumns(CMMConstants.UPLOADTABLENAMES.ToList());
                List<SqlBulkCopyColumnMapping> mappingCollection = null;
                DataTable dtTemp = null;
                DataTable dtData = null;
                DataTable dtCopy = null;
                string sColumnName = null;

                // 转换guid列类型及复制数据表数据
                Func<DataTable> setType = () =>
                {
                    // 需要设置成guid类型的列名集合
                    string[] columnNames = new string[] { "ShiYongID", "ShiYanShiID", "ZuoWeiID", "ShiYongMingXiID", "ShuKaJiLuID" };
                    DataTable dtSetType = dtData.Clone();
                    foreach (string columnName in columnNames)
                    {
                        if (dtSetType.Columns.Contains(columnName))
                        {
                            // 设置列类型
                            dtSetType.Columns[columnName].DataType = typeof(Guid);
                        }
                    }

                    // 填充数据
                    dtSetType.Load(dtData.CreateDataReader());
                    return dtSetType;
                };

                // 循环多张表
                foreach (string tblName in CMMConstants.UPLOADTABLENAMES)
                {
                    dtData = ds.Tables[tblName];
                    dtTemp = dsColumns.Tables[tblName];
                    mappingCollection = new List<SqlBulkCopyColumnMapping>();

                    if (dtData.Rows.Count > 0)
                    {
                        // 循环表中的列集合
                        foreach (DataRow row in dtTemp.Rows)
                        {
                            sColumnName = row["name"].ToString();
                            if (dtData.Columns.Contains(sColumnName))
                            {
                                // 添加映射
                                mappingCollection.Add(new SqlBulkCopyColumnMapping(sColumnName, sColumnName));
                            }
                        }

                        // 获取转换列类型后的数据表
                        dtCopy = setType();

                        // 将数据更新到服务器
                        _DBManager.BulkCopy(tblName, dtCopy, mappingCollection);

                        // 更新本地数据上传状态
                        string sKeyName = CMMConstants.UPLOADTABLENAMES_PrimaryKey[tblName];
                        string sInStr = InStr(dtCopy, sKeyName);
                        string sUpdateSql = string.Format("UPDATE {0} SET IsUpload=‘{1}‘ where {2} IN({3})", tblName, "1", sKeyName, sInStr);
                        int iResult = _SQLiteDBManager.ExeSql(sUpdateSql);
                    }
                }

                #endregion
            }
            catch (Exception ex)
            {
                throw ex;
            }
        }

ps:
         未加func处理datatable前，报string转guid异常，直接修改datatable列类型报异常有数据时无法更改列类型，
    func处理也许并不是最好的，将bulkcopy改为使用自定义表类型，将数据传递到服务器再进行写入也许更好。