C#读取Excel的三种方式以及比较
（1）OleDB方式
优点：将Excel直接当做数据源处理，通过SQL直接读取内容，读取速度较快。
缺点：读取数据方式不够灵活，无法直接读取某一个单元格，只有将整个Sheet页读取出来后（结果为Datatable）再在Datatable中根据行列数来获取指定的值。
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。
读取代码如下：
1: public DataTable GetExcelTableByOleDB(string strExcelPath, string tableName)
2: {
3: try
4: {
5: DataTable dtExcel = new DataTable();
6: //数据表
7: DataSet ds = new DataSet();
8: //获取文件扩展名
9: string strExtension = System.IO.Path.GetExtension(strExcelPath);
10: string strFileName = System.IO.Path.GetFileName(strExcelPath);
11: //Excel的连接
12: OleDbConnection objConn = null;
13: switch (strExtension)
14: {
15: case ".xls":
16: objConn = new OleDbConnection("Provider=Microsoft.Jet.OLEDB.4.0;Data Source=" + strExcelPath + ";" + "Extended Properties=\"Excel 8.0;HDR=NO;IMEX=1;\"");
17: break;
18: case ".xlsx":
19: objConn = new OleDbConnection("Provider=Microsoft.ACE.OLEDB.12.0;Data Source=" + strExcelPath + ";" + "Extended Properties=\"Excel 12.0;HDR=NO;IMEX=1;\"");
20: break;
21: default:
22: objConn = null;
23: break;
24: }
25: if (objConn == null)
26: {
27: return null;
28: }
29: objConn.Open();
30: //获取Excel中所有Sheet表的信息
31: //System.Data.DataTable schemaTable = objConn.GetOleDbSchemaTable(System.Data.OleDb.OleDbSchemaGuid.Tables, null);
32: //获取Excel的第一个Sheet表名
33: //string tableName = schemaTable.Rows[0][2].ToString().Trim();
34: string strSql = "select * from [" + tableName + "]";
35: //获取Excel指定Sheet表中的信息
36: OleDbCommand objCmd = new OleDbCommand(strSql, objConn);
37: OleDbDataAdapter myData = new OleDbDataAdapter(strSql, objConn);
38: myData.Fill(ds, tableName);//填充数据
39: objConn.Close();
40: //dtExcel即为excel文件中指定表中存储的信息
41: dtExcel = ds.Tables[tableName];
42: return dtExcel;
43: }
44: catch
45: {
46: return null;
47: }
48: }
下面说明一下连接字符串
HDR=Yes，这代表第一行是标题，不做为数据使用（但是我在实际使用中，如果第一行存在复杂数值，那么读取得到的Datatable列标题会自动设置为F1、F2等方式命名，与实际应用不符，所以当时是通过HDR=No方式将所有内容读取到Datatable中，然后手动将第一行设置成标题的）；IMEX ( IMport EXport mode )设置
IMEX 有三种模式：
0 is Export mode
1 is Import mode
2 is Linked mode (full update capabilities)
我这里特别要说明的就是 IMEX 参数了，因为不同的模式代表著不同的读写行为：
当 IMEX=0 时为“汇出模式”，这个模式开启的 Excel 档案只能用来做“写入”用途。
当 IMEX=1 时为“汇入模式”，这个模式开启的 Excel 档案只能用来做“读取”用途。
当 IMEX=2 时为“链接模式”，这个模式开启的 Excel 档案可同时支援“读取”与“写入”用途。
---------------------------------
另外，读取Excel2007版本的文件时，版本应该从8.0改为12.0，同时驱动不能再用Jet，而应该用ACE。负责会造成“找不到可安装的 ISAM”的错误。
---------------------------------
在网上还发现采用这种方式存在取出的Sheet表的个数多于实际Excel表中的Sheet表个数的情况，其原因有二：
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。
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。下面给出了过滤的方案。（此问题我们有验证过，大家自己验证一下吧）
1: //objConn为读取Excel的链接，下面通过过滤来获取有效的Sheet页名称集合
2: System.Data.DataTable schemaTable = objConn.GetOleDbSchemaTable(System.Data.OleDb.OleDbSchemaGuid.Tables, null);
3: List<string> lstSheetNames = new List<string>();
4: for (int i = 0; i < schemaTable.Rows.Count; i++)
5: {
6: string strSheetName = (string)dtSheetName.Rows[i]["TABLE_NAME"];
7: if (strSheetName.Contains("$") && !strSheetName.Replace("'", "").EndsWith("$"))
8: {
9: //过滤无效SheetName完毕....
10: continue;
11: }
12: if (lstSheetNames != null && !lstSheetNames.Contains(strSheetName))
13: lstSheetNames.Add(strSheetName);
14: }
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。
---------------------------------
（2）Com组件的方式（通过添加 Microsoft.Office.Interop.Excel引用实现）
优点：能够非常灵活的读取Excel中的数据，用户可以灵活的调用各种函数进行处理。
缺点：基于单元格的处理，读取速度较慢，对于数据量较大的文件最好不要使用此种方式读取。
需要添加相应的DLL引用，必须存在此引用才可使用，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。
读取代码如下：
1: private Stopwatch wath = new Stopwatch();
2: /// <summary>
3: /// 使用COM读取Excel
4: /// </summary>
5: /// <param name="excelFilePath">路径</param>
6: /// <returns>DataTabel</returns>
7: public System.Data.DataTable GetExcelData(string excelFilePath)
8: {
9: Excel.Application app = new Excel.Application();
10: Excel.Sheets sheets;
11: Excel.Workbook workbook = null;
12: object oMissiong = System.Reflection.Missing.Value;
13: System.Data.DataTable dt = new System.Data.DataTable();
14: wath.Start();
15: try
16: {
17: if (app == null)
18: {
19: return null;
20: }
21: workbook = app.Workbooks.Open(excelFilePath, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong,
22: oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong);
23: //将数据读入到DataTable中——Start
24: sheets = workbook.Worksheets;
25: Excel.Worksheet worksheet = (Excel.Worksheet)sheets.get_Item(1);//读取第一张表
26: if (worksheet == null)
27: return null;
28: string cellContent;
29: int iRowCount = worksheet.UsedRange.Rows.Count;
30: int iColCount = worksheet.UsedRange.Columns.Count;
31: Excel.Range range;
32: //负责列头Start
33: DataColumn dc;
34: int ColumnID = 1;
35: range = (Excel.Range)worksheet.Cells[1, 1];
36: while (range.Text.ToString().Trim() != "")
37: {
38: dc = new DataColumn();
39: dc.DataType = System.Type.GetType("System.String");
40: dc.ColumnName = range.Text.ToString().Trim();
41: dt.Columns.Add(dc);
42:
43: range = (Excel.Range)worksheet.Cells[1, ++ColumnID];
44: }
45: //End
46: for (int iRow = 2; iRow <= iRowCount; iRow++)
47: {
48: DataRow dr = dt.NewRow();
49: for (int iCol = 1; iCol <= iColCount; iCol++)
50: {
51: range = (Excel.Range)worksheet.Cells[iRow, iCol];
52: cellContent = (range.Value2 == null) ? "" : range.Text.ToString();
53: dr[iCol - 1] = cellContent;
54: }
55: dt.Rows.Add(dr);
56: }
57: wath.Stop();
58: TimeSpan ts = wath.Elapsed;
59: //将数据读入到DataTable中——End
60: return dt;
61: }
62: catch
63: {
64: return null;
65: }
66: finally
67: {
68: workbook.Close(false, oMissiong, oMissiong);
69: System.Runtime.InteropServices.Marshal.ReleaseComObject(workbook);
70: workbook = null;
71: app.Workbooks.Close();
72: app.Quit();
73: System.Runtime.InteropServices.Marshal.ReleaseComObject(app);
74: app = null;
75: GC.Collect();
76: GC.WaitForPendingFinalizers();
77: }
78: }
79: /// <summary>
80: /// 使用COM，多线程读取Excel（1 主线程、4 副线程）
81: /// </summary>
82: /// <param name="excelFilePath">路径</param>
83: /// <returns>DataTabel</returns>
84: public System.Data.DataTable ThreadReadExcel(string excelFilePath)
85: {
86: Excel.Application app = new Excel.Application();
87: Excel.Sheets sheets = null;
88: Excel.Workbook workbook = null;
89: object oMissiong = System.Reflection.Missing.Value;
90: System.Data.DataTable dt = new System.Data.DataTable();
91: wath.Start();
92: try
93: {
94: if (app == null)
95: {
96: return null;
97: }
98: workbook = app.Workbooks.Open(excelFilePath, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong,
99: oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong, oMissiong);
100: //将数据读入到DataTable中——Start
101: sheets = workbook.Worksheets;
102: Excel.Worksheet worksheet = (Excel.Worksheet)sheets.get_Item(1);//读取第一张表
103: if (worksheet == null)
104: return null;
105: string cellContent;
106: int iRowCount = worksheet.UsedRange.Rows.Count;
107: int iColCount = worksheet.UsedRange.Columns.Count;
108: Excel.Range range;
109: //负责列头Start
110: DataColumn dc;
111: int ColumnID = 1;
112: range = (Excel.Range)worksheet.Cells[1, 1];
113: while (iColCount >= ColumnID)
114: {
115: dc = new DataColumn();
116: dc.DataType = System.Type.GetType("System.String");
117: string strNewColumnName = range.Text.ToString().Trim();
118: if (strNewColumnName.Length == 0) strNewColumnName = "_1";
119: //判断列名是否重复
120: for (int i = 1; i < ColumnID; i++)
121: {
122: if (dt.Columns[i - 1].ColumnName == strNewColumnName)
123: strNewColumnName = strNewColumnName + "_1";
124: }
125: dc.ColumnName = strNewColumnName;
126: dt.Columns.Add(dc);
127: range = (Excel.Range)worksheet.Cells[1, ++ColumnID];
128: }
129: //End
130: //数据大于500条，使用多进程进行读取数据
131: if (iRowCount - 1 > 500)
132: {
133: //开始多线程读取数据
134: //新建线程
135: int b2 = (iRowCount - 1) / 10;
136: DataTable dt1 = new DataTable("dt1");
137: dt1 = dt.Clone();
138: SheetOptions sheet1thread = new SheetOptions(worksheet, iColCount, 2, b2 + 1, dt1);
139: Thread othread1 = new Thread(new ThreadStart(sheet1thread.SheetToDataTable));
140: othread1.Start();
141: //阻塞 1 毫秒，保证第一个读取 dt1
142: Thread.Sleep(1);
143: DataTable dt2 = new DataTable("dt2");
144: dt2 = dt.Clone();
145: SheetOptions sheet2thread = new SheetOptions(worksheet, iColCount, b2 + 2, b2 * 2 + 1, dt2);
146: Thread othread2 = new Thread(new ThreadStart(sheet2thread.SheetToDataTable));
147: othread2.Start();
148: DataTable dt3 = new DataTable("dt3");
149: dt3 = dt.Clone();
150: SheetOptions sheet3thread = new SheetOptions(worksheet, iColCount, b2 * 2 + 2, b2 * 3 + 1, dt3);
151: Thread othread3 = new Thread(new ThreadStart(sheet3thread.SheetToDataTable));
152: othread3.Start();
153: DataTable dt4 = new DataTable("dt4");
154: dt4 = dt.Clone();
155: SheetOptions sheet4thread = new SheetOptions(worksheet, iColCount, b2 * 3 + 2, b2 * 4 + 1, dt4);
156: Thread othread4 = new Thread(new ThreadStart(sheet4thread.SheetToDataTable));
157: othread4.Start();
158: //主线程读取剩余数据
159: for (int iRow = b2 * 4 + 2; iRow <= iRowCount; iRow++)
160: {
161: DataRow dr = dt.NewRow();
162: for (int iCol = 1; iCol <= iColCount; iCol++)
163: {
164: range = (Excel.Range)worksheet.Cells[iRow, iCol];
165: cellContent = (range.Value2 == null) ? "" : range.Text.ToString();
166: dr[iCol - 1] = cellContent;
167: }
168: dt.Rows.Add(dr);
169: }
170: othread1.Join();
171: othread2.Join();
172: othread3.Join();
173: othread4.Join();
174: //将多个线程读取出来的数据追加至 dt1 后面
175: foreach (DataRow dr in dt.Rows)
176: dt1.Rows.Add(dr.ItemArray);
177: dt.Clear();
178: dt.Dispose();
179: foreach (DataRow dr in dt2.Rows)
180: dt1.Rows.Add(dr.ItemArray);
181: dt2.Clear();
182: dt2.Dispose();
183: foreach (DataRow dr in dt3.Rows)
184: dt1.Rows.Add(dr.ItemArray);
185: dt3.Clear();
186: dt3.Dispose();
187: foreach (DataRow dr in dt4.Rows)
188: dt1.Rows.Add(dr.ItemArray);
189: dt4.Clear();
190: dt4.Dispose();
191: return dt1;
192: }
193: else
194: {
195: for (int iRow = 2; iRow <= iRowCount; iRow++)
196: {
197: DataRow dr = dt.NewRow();
198: for (int iCol = 1; iCol <= iColCount; iCol++)
199: {
200: range = (Excel.Range)worksheet.Cells[iRow, iCol];
201: cellContent = (range.Value2 == null) ? "" : range.Text.ToString();
202: dr[iCol - 1] = cellContent;
203: }
204: dt.Rows.Add(dr);
205: }
206: }
207: wath.Stop();
208: TimeSpan ts = wath.Elapsed;
209: //将数据读入到DataTable中——End
210: return dt;
211: }
212: catch
213: {
214: return null;
215: }
216: finally
217: {
218: workbook.Close(false, oMissiong, oMissiong);
219: System.Runtime.InteropServices.Marshal.ReleaseComObject(workbook);
220: System.Runtime.InteropServices.Marshal.ReleaseComObject(sheets);
221: workbook = null;
222: app.Workbooks.Close();
223: app.Quit();
224: System.Runtime.InteropServices.Marshal.ReleaseComObject(app);
225: app = null;
226: GC.Collect();
227: GC.WaitForPendingFinalizers();
228: }
229: }
---------------------------------
（3）NPOI方式读取Excel（此方法未经过测试）
NPOI 是 POI 项目的 .NET 版本。POI是一个开源的Java读写Excel、WORD等微软OLE2组件文档的项目。使用 NPOI 你就可以在没有安装 Office 或者相应环境的机器上对 WORD/EXCEL 文档进行读写。
优点：读取Excel速度较快，读取方式操作灵活性
缺点：需要下载相应的插件并添加到系统引用当中。
1: /// <summary>
2: /// 将excel中的数据导入到DataTable中
3: /// </summary>
4: /// <param name="sheetName">excel工作薄sheet的名称</param>
5: /// <param name="isFirstRowColumn">第一行是否是DataTable的列名</param>
6: /// <returns>返回的DataTable</returns>
7: public DataTable ExcelToDataTable(string sheetName, bool isFirstRowColumn)
8: {
9: ISheet sheet = null;
10: DataTable data = new DataTable();
11: int startRow = 0;
12: try
13: {
14: fs = new FileStream(fileName, FileMode.Open, FileAccess.Read);
15: if (fileName.IndexOf(".xlsx") > 0) // 2007版本
16: workbook = new XSSFWorkbook(fs);
17: else if (fileName.IndexOf(".xls") > 0) // 2003版本
18: workbook = new HSSFWorkbook(fs);
19: if (sheetName != null)
20: {
21: sheet = workbook.GetSheet(sheetName);
22: }
23: else
24: {
25: sheet = workbook.GetSheetAt(0);
26: }
27: if (sheet != null)
28: {
29: IRow firstRow = sheet.GetRow(0);
30: int cellCount = firstRow.LastCellNum; //一行最后一个cell的编号 即总的列数
31: if (isFirstRowColumn)
32: {
33: for (int i = firstRow.FirstCellNum; i < cellCount; ++i)
34: {
35: DataColumn column = new DataColumn(firstRow.GetCell(i).StringCellValue);
36: data.Columns.Add(column);
37: }
38: startRow = sheet.FirstRowNum + 1;
39: }
40: else
41: {
42: startRow = sheet.FirstRowNum;
43: }
44: //最后一列的标号
45: int rowCount = sheet.LastRowNum;
46: for (int i = startRow; i <= rowCount; ++i)
47: {
48: IRow row = sheet.GetRow(i);
49: if (row == null) continue; //没有数据的行默认是null
50:
51: DataRow dataRow = data.NewRow();
52: for (int j = row.FirstCellNum; j < cellCount; ++j)
53: {
54: if (row.GetCell(j) != null) //同理，没有数据的单元格都默认是null
55: dataRow[j] = row.GetCell(j).ToString();
56: }
57: data.Rows.Add(dataRow);
58: }
59: }
60: return data;
61: }
62: catch (Exception ex)
63: {
64: Console.WriteLine("Exception: " + ex.Message);
65: return null;
66: }
67: }