對于某些項目，開發(fā)人員需要從Word文檔中提取數(shù)據(jù)并導出到數(shù)據(jù)庫。最大的挑戰(zhàn)是必須支持現(xiàn)有Word文檔。

相同格式且?guī)Ф鄠€數(shù)據(jù)塊的Word文檔有成千上萬。該文檔格式并不是設計來被另一個系統(tǒng)所讀取的。這意味著，沒有書簽、合并字段、從標準指令識別實際數(shù)據(jù)的方式等。還好，所有輸入字段都在表格內(nèi)，但這些表格也是不同的格式，一些是單行/單元格，另一些則變化多端。

我們可以用Aspose.Words來創(chuàng)建和操作Word文檔。

以C#創(chuàng)建一個類似的表格模型從而稍后當讀取文檔的時候我們可以用上它。

如下所示，你可以看到創(chuàng)建的名為WordDocumentTable的類，帶有三個屬性，如TableID,RowID和ColumnID，如之前所說的，我們沒有支持TableID/RowIDs，這些屬性僅僅暗示著Word文檔的位置。開始索引假定為0。

public class WordDocumentTable
{ 
	public WordDocumentTable(int PiTableID) 
	{  
		MiTableID = PiTableID; 
	}

	public WordDocumentTable(int PiTableID, int PiColumnID) 
	{  
		MiTableID = PiTableID;  
		MiColumnID = PiColumnID; 
	}

	public WordDocumentTable(int PiTableID, int PiColumnID, int PiRowID) 
	{  
		MiTableID = PiTableID;  
		MiColumnID = PiColumnID;  
		MiRowID = PiRowID; 
	}

	private int MiTableID = 0;

	public int TableID 
	{  
		get { return MiTableID; }  
		set { MiTableID = value; } 
	}        

	private int MiRowID = 0;    
	public int RowID 
	{  
		get { return MiRowID; }  
		set { MiRowID = value; } 
	}

	private int MiColumnID = 0;    
	public int ColumnID 
	{  
		get { return MiColumnID; }  
		set { MiColumnID = value; } 
	}
}

現(xiàn)在來到提取環(huán)節(jié)。如下所示，你將看到我想要從文檔中讀取的表格單元格的連接。

private List<WordDocumentTable> WordDocumentTables
{  
	get  
	{    
		List<WordDocumentTable> wordDocTable = new List<WordDocumentTable>();      
		//Reads the data from the first Table of the document.    
		wordDocTable.Add(new WordDocumentTable(0));      
		//Reads the data from the second table and its second column. 
		//This table has only one row.    
		wordDocTable.Add(new WordDocumentTable(1, 1));      
		//Reads the data from third table, second row and second cell.    
		wordDocTable.Add(new WordDocumentTable(2, 1, 1));  
		return wordDocTable;  
	}
}

下面是從基于表格的Aspose.Words文檔提取數(shù)據(jù)，行和單元格。

public void ExtractTableData(byte[] PobjData)
{          
	using (MemoryStream LobjStream = new MemoryStream(PobjData)) 
	{  
		Document LobjAsposeDocument = new Document(LobjStream);     
		foreach(WordDocumentTable wordDocTable in WordDocumentTables)  
		{   
			Aspose.Words.Tables.Table table = (Aspose.Words.Tables.Table)
			LobjAsposeDocument.GetChild
			(NodeType.Table, wordDocTable.TableID, true);   
			string cellData = table.Range.Text;

			if (wordDocTable.ColumnID > 0)   
			{    
				if (wordDocTable.RowID == 0)    
				{     
					NodeCollection LobjCells = 
					table.GetChildNodes(NodeType.Cell, true);     
					cellData = LobjCells[wordDocTable.ColumnID].ToTxt();
				}    
				else    
				{     
					NodeCollection LobjRows = 
					table.GetChildNodes(NodeType.Row, true);     
					cellData = ((Row)(LobjRows[wordDocTable.RowID])).
					Cells[wordDocTable.ColumnID].ToTxt();    
				}   
			}

			Console.WriteLine(String.Format("Data in Table {0}, 
					Row {1}, Column {2} : {3}",           
									wordDocTable.TableID,          
									wordDocTable.RowID,          
									wordDocTable.ColumnID,          
									cellData);              
		} 
	}
}