DataTable将行转成列

　　以前一直觉得泛型比较好用，一直没使用DataTable，最近发现泛型也不是万能的，比如将行转成列。

　　虽然这种在前台JS转比较方便，但可能也需要这样的接口，提供别人调用，还是记录一下。

/// <summary>
        /// 将行转成列
        /// DataTable dtResult = Utility.ConvertDataTableToVertical(dt, "ID", new string[]{"Code","StationName"}, new string[] { "Value", "Name", "IsExcessive" });
        /// </summary>
        /// <param name="dt">数据源</param>
        /// <param name="rowKeyArray">唯一标示列</param>
        /// <param name="columnKeyArray">要转成表头的列</param>
        /// <param name="valueKeyArray">数据行</param>
        /// <param name="preserveColumnArray">剩下的要显示的行</param>
        /// <returns></returns>
public static DataTable ConvertDataTableToVertical(DataTable dt, string[] rowKeyArray, string[] columnKeyArray, string[] valueKeyArray, List<string> preserveColumnArray)
        {
            DataTable dtResult = new DataTable();
            if (preserveColumnArray == null)
            {
                preserveColumnArray = new List<string>();
                foreach (var column in dt.Columns)
                {
                    if (columnKeyArray.Contains(column.ToString()) || valueKeyArray.Contains(column.ToString()))
                    {
                        continue;
                    }
                    preserveColumnArray.Add(column.ToString());
                }
            }
            var columns = (from r in dt.AsEnumerable() select GetStringFromDataRow(r, columnKeyArray)).Where(r => !string.IsNullOrEmpty(r)).Distinct().OrderBy(r => r).ToArray();
            var columnCount = columns.Length * valueKeyArray.Length + preserveColumnArray.Count;
            Dictionary<object, object[]> rows = new Dictionary<object, object[]>();
            foreach (DataRow dr in dt.Rows)
            {
                var rowKey = GetStringFromDataRow(dr, rowKeyArray);
                if (rows.Keys.Contains(rowKey) == false)
                {
                    rows.Add(rowKey, new object[columnCount]);
                    for (var j = 0; j < preserveColumnArray.Count; j++)
                    {
                        rows[rowKey][j] = dr[preserveColumnArray[j]];
                    }
                }
                for (var i = 0; i < columns.Length; i++)
                {
                    if (columns[i] == GetStringFromDataRow(dr, columnKeyArray))
                    {
                        var j = i * valueKeyArray.Length + preserveColumnArray.Count;
                        foreach (var valueKey in valueKeyArray)
                        {
                            rows[rowKey][j] = dr[valueKey];
                            j++;
                        }
                        break;
                    }
                }
            }
            foreach (var column in preserveColumnArray)
            {
                dtResult.Columns.Add(column, dt.Columns[column].DataType);
            }
            foreach (var column in columns)
            {
                foreach (var valueKey in valueKeyArray)
                {
                    if (column.IndexOf('_') == -1)
                    {
                        dtResult.Columns.Add(column + '_' + valueKey, dt.Columns[valueKey].DataType);
                    }
                    else
                    {
                        dtResult.Columns.Add(column + '_' + valueKey, dt.Columns[valueKey].DataType);
                        //dtResult.Columns.Add(column.Substring(0, column.IndexOf('_')) + '_' + valueKey + column.Substring(column.IndexOf('_')), dt.Columns[valueKey].DataType);
                    }
                }
            }
            foreach (var keyValuePair in rows)
            {
                dtResult.Rows.Add(keyValuePair.Value);
            }
            return dtResult;
        }

public static string GetStringFromDataRow(DataRow dr, string[] keys)
        {
            string strResult = string.Empty;
            foreach (string key in keys)
            {
                Type ObjectValueType = dr[key].GetType();
                switch (ObjectValueType.FullName)
                {
                    case "System.DateTime":
                        strResult += "_" + DateTimeHelper.DateTimeToUTCTime((DateTime)dr[key]);
                        break;
                    default:
                        strResult += "_" + dr[key];
                    break;
                }                
            }
            if (strResult.Length > 0)
            {
                strResult = strResult.Remove(0, 1);
            }
            return strResult;
        }

 　　经常得到的是List，发现不会转成行显示（后台），只能先转成DataTable了。。。

/// <summary>
        /// 将泛类型集合List类转换成DataTable
        /// </summary>
        /// <param name="list">泛类型集合</param>
        /// <returns></returns>
        public static DataTable ListToDataTable<T>(List<T> entitys)
        {
            //检查实体集合不能为空
            if (entitys == null || entitys.Count < 1)
            {
                throw new Exception("需转换的集合为空");
            }
            //取出第一个实体的所有Propertie
            Type entityType = entitys[0].GetType();
            PropertyInfo[] entityProperties = entityType.GetProperties();

            //生成DataTable的structure
            //生产代码中，应将生成的DataTable结构Cache起来，此处略
            DataTable dt = new DataTable();
            for (int i = 0; i < entityProperties.Length; i++)
            {
                //dt.Columns.Add(entityProperties[i].Name, entityProperties[i].PropertyType);
                dt.Columns.Add(entityProperties[i].Name);
            }
            //将所有entity添加到DataTable中
            foreach (object entity in entitys)
            {
                //检查所有的的实体都为同一类型
                if (entity.GetType() != entityType)
                {
                    throw new Exception("要转换的集合元素类型不一致");
                }
                object[] entityValues = new object[entityProperties.Length];
                for (int i = 0; i < entityProperties.Length; i++)
                {
                    entityValues[i] = entityProperties[i].GetValue(entity, null);
                }
                dt.Rows.Add(entityValues);
            }
            return dt;
        }