C#读取Excel几种方法的体会

OStack程序员社区-中国程序员成长平台 › 门户 › 编程› Node.js›Node.js教程

原作者: [db:作者] 来自: [db:来源] 收藏邀请

(1) OleDb：用这种方法读取Excel速度还是非常的快的，但这种方式读取数据的时候不太灵活，不过可以在 DataTable 中对数据进行一些删减修改

这种方式将Excel作为一个数据源，直接用Sql语句获取数据了。所以读取之前要知道此次要读取的Sheet（当然也可以用序号，类似dt.Row[0][0]。这样倒是不需要知道Sheet）

if (fileType == ".xls")

   connStr = "Provider=Microsoft.Jet.OLEDB.4.0;" + "Data Source=" + fileName + ";" + ";Extended Properties=\"Excel 8.0;HDR=YES;IMEX=1\"";
else

   connStr = "Provider=Microsoft.ACE.OLEDB.12.0;" + "Data Source=" + fileName + ";" + ";Extended Properties=\"Excel 12.0;HDR=YES;IMEX=1\"";
 
OleDbConnection conn new OleDbConnection(connStr);

DataTable dtSheetName = conn.GetOleDbSchemaTable(OleDbSchemaGuid.Tables, new object[] { null, null, null, "TABLE" });

以上是读取Excel的Sheet名，xls和xlsx的连接字符串也不一样的，可以根据文件的后缀来区别。这里需要注意的一点，Excel里面只有一个Sheet，但通过这种方式读取Sheet可能会大于一个。原因已经有人在别的网站说过了，偷一下懒O(∩_∩)O，下面文段来自【cdwolfling】

【在使用过程中发现取出的Sheet和实际excel不一致，会多出不少。目前总结后有两种情况：

1. 取出的名称中，包括了XL命名管理器中的名称（参见XL2007的公式--命名管理器，快捷键Crtl+F3）；

2. 取出的名称中，包括了FilterDatabase后缀的，这是XL用来记录Filter范围的，参见http://www.mrexcel.com/forum/showthread.php?t=27225；

对于第一点比较简单，删除已有命名管理器中的内容即可；第二点处理起来比较麻烦， Filter删除后这些名称依然保留着，简单的做法是新增sheet然后将原sheet Copy进去】

---------------------------------

但实际情况并不能为每个Excel做以上检查，【cdwolfling】也给出了过滤的方案，当时还是有点问题，本来补充了一点。总之先看代码吧

for (int i = 0; i < dtSheetName.Rows.Count; i++)

{

SheetName = (string)dtSheetName.Rows[i]["TABLE_NAME"];
 
if (SheetName .Contains("$") && !SheetName .Replace("'", "").EndsWith("$"))continue;//过滤无效SheetName完毕....

da.SelectCommand = new OleDbCommand(String.Format(sql_F, tblName), conn);
DataSet dsItem = new DataSet();
da.Fill(dsItem, tblName);

}

因为读取出来无效SheetName一般情况最后一个字符都不会是$。如果SheetName有一些特殊符号，读取出来的SheetName会自动加上单引号，比如在Excel中将SheetName编辑成：MySheet(1)，此时读取出来的SheetName就为：'MySheet(1)$'，所以判断最后一个字符是不是$之前最好过滤一下单引号。

优点：读取方式简单、读取速度快

缺点：除了读取过程不太灵活之外，这种读取方式还有个弊端就是，当Excel数据量很大时。会非常占用内存，当内存不够时会抛出内存溢出的异常。

不过一般情况下还是非常不错的

读取Excel完整代码：

/// <summary>

        /// 读取Excel文件到DataSet中

        /// </summary>

        /// <param name="filePath">文件路径</param>

        /// <returns></returns>

        public static DataSet ToDataTable(string filePath)

        {

            string connStr = "";            

            string fileType = System.IO.Path.GetExtension(fileName);

            if (string.IsNullOrEmpty(fileType)) return null;
 
            if (fileType == ".xls")

                connStr = "Provider=Microsoft.Jet.OLEDB.4.0;" + "Data Source=" + filePath+ ";" + ";Extended Properties=\"Excel 8.0;HDR=YES;IMEX=1\"";

            else

                connStr = "Provider=Microsoft.ACE.OLEDB.12.0;" + "Data Source=" + filePath+ ";" + ";Extended Properties=\"Excel 12.0;HDR=YES;IMEX=1\"";

            string sql_F = "Select * FROM [{0}]";
 
            OleDbConnection conn = null;

            OleDbDataAdapter da = null;

            DataTable dtSheetName= null;
 
            DataSet ds = new DataSet();

            try

            {

                // 初始化连接，并打开

                conn = new OleDbConnection(connStr);

                conn.Open();
 
                // 获取数据源的表定义元数据                        

                string SheetName = "";

                dtSheetName= conn.GetOleDbSchemaTable(OleDbSchemaGuid.Tables, new object[] { null, null, null, "TABLE" });
 
                // 初始化适配器

                da = new OleDbDataAdapter();

                for (int i = 0; i < dtSheetName.Rows.Count; i++)

                {

                    SheetName = (string)dtSheetName.Rows[i]["TABLE_NAME"];
 
                    if (SheetName .Contains("$") && !SheetName .Replace("'", "").EndsWith("$"))

                    {

                        continue;

                    }
 
                    da.SelectCommand = new OleDbCommand(String.Format(sql_F, SheetName ), conn);

                    DataSet dsItem = new DataSet();

                    da.Fill(dsItem, tblName);
 
                    ds.Tables.Add(dsItem.Tables[0].Copy());

                }

            }

            catch (Exception ex)

            {

            }

            finally

            {

                // 关闭连接

                if (conn.State == ConnectionState.Open)

                {

                    conn.Close();

                    da.Dispose();

                    conn.Dispose();

                }

            }

            return ds;

        }

(2)：Com组件的方式读取Excel

这种方式需要先引用 Microsoft.Office.Interop.Excel 。首选说下这种方式的优缺点

优点：可以非常灵活的读取Excel中的数据

缺点：如果是Web站点部署在IIS上时，还需要服务器机子已安装了Excel，有时候还需要为配置IIS权限。最重要的一点因为是基于单元格方式读取的，所以数据很慢（曾做过试验，直接读取千行、200多列的文件，直接读取耗时15分钟。即使采用多线程分段读取来提高CPU的利用率也需要8分钟。PS：CPU I3）

需要读取大文件的的童鞋们慎重。。。

附上单线程和多线程读取类：

public class ExcelOptions

    {

        private Stopwatch wath = new Stopwatch();
 
        /// <summary>

        /// 使用COM读取Excel

        /// </summary>

        /// <param name="excelFilePath">路径</param>

        /// <returns>DataTabel</returns>

        public System.Data.DataTable GetExcelData(string excelFilePath)

        {

            Excel.Application app = new Excel.Application();

            Excel.Sheets sheets;

            Excel.Workbook workbook = null;

            object oMissiong = System.Reflection.Missing.Value;

            System.Data.DataTable dt = new System.Data.DataTable();
 
            wath.Start();
 
            try

            {

                if (app == null)

                {

                    return null;

                }
 
                workbook = app.Workbooks.Open(excelFilePath, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong);
 
                //将数据读入到DataTable中——Start   
 
                sheets = workbook.Worksheets;

                Excel.Worksheet worksheet = (Excel.Worksheet)sheets.get_Item(1);//读取第一张表

                if (worksheet == null)

                    return null;
 
                string cellContent;

                int iRowCount = worksheet.UsedRange.Rows.Count;

                int iColCount = worksheet.UsedRange.Columns.Count;

                Excel.Range range;
 
                //负责列头Start

                DataColumn dc;

                int ColumnID = 1;

                range = (Excel.Range)worksheet.Cells[1, 1];

                while (range.Text.ToString().Trim() != "")

                {

                    dc = new DataColumn();

                    dc.DataType = System.Type.GetType("System.String");

                    dc.ColumnName = range.Text.ToString().Trim();

                    dt.Columns.Add(dc);
 
                    range = (Excel.Range)worksheet.Cells[1, ++ColumnID];

                }

                //End
 
                for (int iRow = 2; iRow <= iRowCount; iRow++)

                {

                    DataRow dr = dt.NewRow();
 
                    for (int iCol = 1; iCol <= iColCount; iCol++)

                    {

                        range = (Excel.Range)worksheet.Cells[iRow, iCol];
 
                        cellContent = (range.Value2 == null) ? "" : range.Text.ToString();
 
                        //if (iRow == 1)

                        //{

                        //    dt.Columns.Add(cellContent);

                        //}

                        //else

                        //{

                            dr[iCol - 1] = cellContent;

                        //}

                    }
 
                    //if (iRow != 1)

                    dt.Rows.Add(dr);

                }
 
                wath.Stop();

                TimeSpan ts = wath.Elapsed;
 
                //将数据读入到DataTable中——End

                return dt;

            }

            catch

            {

                return null;

            }

            finally

            {

                workbook.Close(false, oMissiong, oMissiong);

                System.Runtime.InteropServices.Marshal.ReleaseComObject(workbook);

                workbook = null;

                app.Workbooks.Close();

                app.Quit();

                System.Runtime.InteropServices.Marshal.ReleaseComObject(app);

                app = null;

                GC.Collect();

                GC.WaitForPendingFinalizers();

            }

        }
 
        /// <summary>

        /// 使用COM，多线程读取Excel（1 主线程、4 副线程）

        /// </summary>

        /// <param name="excelFilePath">路径</param>

        /// <returns>DataTabel</returns>

        public System.Data.DataTable ThreadReadExcel(string excelFilePath)

        {

            Excel.Application app = new Excel.Application();

            Excel.Sheets sheets = null;

            Excel.Workbook workbook = null;

            object oMissiong = System.Reflection.Missing.Value;

            System.Data.DataTable dt = new System.Data.DataTable();
 
            wath.Start();
 
            try

            {

                if (app == null)

                {

                    return null;

                }
 
                workbook = app.Workbooks.Open(excelFilePath, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong);
 
                //将数据读入到DataTable中——Start   

                sheets = workbook.Worksheets;

                Excel.Worksheet worksheet = (Excel.Worksheet)sheets.get_Item(1);//读取第一张表

                if (worksheet == null)

                    return null;
 
                string cellContent;

                int iRowCount = worksheet.UsedRange.Rows.Count;

                int iColCount = worksheet.UsedRange.Columns.Count;

                Excel.Range range;
 
                //负责列头Start

                DataColumn dc;

                int ColumnID = 1;

                range = (Excel.Range)worksheet.Cells[1, 1];

                //while (range.Text.ToString().Trim() != "")

                while (iColCount >= ColumnID)

                {

                    dc = new DataColumn();

                    dc.DataType = System.Type.GetType("System.String");
 
                    string strNewColumnName = range.Text.ToString().Trim();

                    if (strNewColumnName.Length == 0) strNewColumnName = "_1";

                    //判断列名是否重复

                    for (int i = 1; i < ColumnID; i++)

                    {

                        if (dt.Columns[i - 1].ColumnName == strNewColumnName)

                            strNewColumnName = strNewColumnName + "_1";

                    }
 
                    dc.ColumnName = strNewColumnName;

                    dt.Columns.Add(dc);
 
                    range = (Excel.Range)worksheet.Cells[1, ++ColumnID];

                }

                //End
 
                //数据大于500条，使用多进程进行读取数据

                if (iRowCount - 1 > 500)

                {

                    //开始多线程读取数据

                    //新建线程

                    int b2 = (iRowCount - 1) / 10;

                    DataTable dt1 = new DataTable("dt1");

                    dt1 = dt.Clone();

                    SheetOptions sheet1thread = new SheetOptions(worksheet, iColCount, 2, b2 + 1, dt1);

                    Thread othread1 = new Thread(new ThreadStart(sheet1thread.SheetToDataTable));

                    othread1.Start();
 
                    //阻塞 1 毫秒，保证第一个读取 dt1

                    Thread.Sleep(1);
 
                    DataTable dt2 = new DataTable("dt2");

                    dt2 = dt.Clone();

                    SheetOptions sheet2thread = new SheetOptions(worksheet, iColCount, b2 + 2, b2 * 2 + 1, dt2);

                    Thread othread2 = new Thread(new ThreadStart(sheet2thread.SheetToDataTable));

                    othread2.Start();
 
                    DataTable dt3 = new DataTable("dt3");

                    dt3 = dt.Clone();

                    SheetOptions sheet3thread = new SheetOptions(worksheet, iColCount, b2 * 2 + 2, b2 * 3 + 1, dt3);

                    Thread othread3 = new Thread(new ThreadStart(sheet3thread.SheetToDataTable));

                    othread3.Start();
 
                    DataTable dt4 = new DataTable("dt4");

                    dt4 = dt.Clone();

                    SheetOptions sheet4thread = new SheetOptions(worksheet, iColCount, b2 * 3 + 2, b2 * 4 + 1, dt4);

                    Thread othread4 = new Thread(new ThreadStart(sheet4thread.SheetToDataTable));

                    othread4.Start();
 
                    //主线程读取剩余数据

                    for (int iRow = b2 * 4 + 2; iRow <= iRowCount; iRow++)

                    {

                        DataRow dr = dt.NewRow();

                        for (int iCol = 1; iCol <= iColCount; iCol++)

                        {

                            range = (Excel.Range)worksheet.Cells[iRow, iCol];

                            cellContent = (range.Value2 == null) ? "" : range.Text.ToString();

                            dr[iCol - 1] = cellContent;

                        }

                        dt.Rows.Add(dr);

                    }
 
                    othread1.Join();

                    othread2.Join();

                    othread3.Join();

                    othread4.Join();
 
                    //将多个线程读取出来的数据追加至 dt1 后面

                    foreach (DataRow dr in dt.Rows)

                        dt1.Rows.Add(dr.ItemArray);

                    dt.Clear();

                    dt.Dispose();
 
                    foreach (DataRow dr in dt2.Rows)

                        dt1.Rows.Add(dr.ItemArray);

                    dt2.Clear();

                    dt2.Dispose();
 
                    foreach (DataRow dr in dt3.Rows)

                        dt1.Rows.Add(dr.ItemArray);

                    dt3.Clear();

                    dt3.Dispose();
 
                    foreach (DataRow dr in dt4.Rows)

                        dt1.Rows.Add(dr.ItemArray);

                    dt4.Clear();

                    dt4.Dispose();