/// <summary>
/// 将句子解析出来
/// </summary>
/// <returns></returns>
public List<string> Parser()
{
items = new List<string>();
Analyzer analyzer = new Lucene.China.ChineseAnalyzer(Server.MapPath("dll"));
StringReader sr = new StringReader(Chinese);
TokenStream stream = analyzer.TokenStream(null, sr);
Token t = stream.Next();
string t1 = "";
while (t != null)
{
t1 = t.ToString(); //显示格式： (关键词,0,2) ，需要处理
t1 = t1.Replace("(", "");
char[] separator = { ',' };
t1 = t1.Split(separator)[0];
items.Add(t1);
t = stream.Next();
}
return items;
}
/// <summary>
/// 将句子解析出来
/// </summary>
/// <returns></returns>
public List <string> Parser()
{
items = new List <string>();
Analyzer analyzer = new Lucene.China.ChineseAnalyzer(Server.MapPath("dll"));
StringReader sr = new StringReader(Chinese);
TokenStream stream = analyzer.TokenStream(null, sr);
Token t = stream.Next();
string t1 = "";
while (t != null)
{
t1 = t.ToString(); //显示格式： (关键词,0,2) ，需要处理
t1 = t1.Replace("(", "");
char[] separator = { ',' };
t1 = t1.Split(separator)[0];
items.Add(t1);
t = stream.Next();
}
return(items);
}
private void RetriveTokenFromText(string sourceText, int fileIndex)
{
List <string> dataTableItem = new List <string>();
string curNewToken = "";
Analyzer analyzer = new Lucene.China.ChineseAnalyzer();
StringReader sr = new StringReader(sourceText);
TokenStream stream = analyzer.TokenStream(null, sr);
Token nextToken = stream.Next();
while (nextToken != null)
{
curNewToken = nextToken.ToString(); //显示格式： (关键词,0,2) ，需要处理
curNewToken = curNewToken.Replace("(", "");
char[] separator = { ',' };
curNewToken = curNewToken.Split(separator)[0];
//StoreTokenIntoList(curNewToken, fileIndex);
dataTableItem.Add(curNewToken);
nextToken = stream.Next();
}
this.dataTable.Add(dataTableItem);
}
private void RetriveTokenFromText(string sourceText, int fileIndex)
{
List<string> dataTableItem = new List<string>();
string curNewToken = "";
Analyzer analyzer = new Lucene.China.ChineseAnalyzer();
StringReader sr = new StringReader(sourceText);
TokenStream stream = analyzer.TokenStream(null, sr);
Token nextToken = stream.Next();
while (nextToken != null)
{
curNewToken = nextToken.ToString(); //显示格式： (关键词,0,2) ，需要处理
curNewToken = curNewToken.Replace("(", "");
char[] separator = { ',' };
curNewToken = curNewToken.Split(separator)[0];
//StoreTokenIntoList(curNewToken, fileIndex);
dataTableItem.Add(curNewToken);
nextToken = stream.Next();
}
this.dataTable.Add(dataTableItem);
}
