You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
 
 
 
 
 

133 lines
5.5 KiB

using log4net;
using Lucene.Net.Analysis;
using Lucene.Net.Index;
using Lucene.Net.Store;
using System.Collections.Generic;
using System.IO;
using System.Threading;
using System.Web.Hosting;
namespace WebApp
{
public class IndexManager
{
public static readonly IndexManager Instance = new IndexManager();
private static readonly string IndexPath = HostingEnvironment.MapPath("~/Index");
private static ILog log = LogManager.GetLogger(typeof(IndexManager));
private IndexManager()
{ }
static IndexManager()
{ }
public void Start()
{
var thread = new Thread(WatchIndexTask)
{
IsBackground = true
};
thread.Start();
log.Debug("IndexManager has been lunched successfully!");
}
private readonly Queue<IndexTask> indexQueue = new Queue<IndexTask>();
private void WatchIndexTask()
{
while (true)
{
if (indexQueue.Count > 0)
{
// 索引文档保存位置
var directory = FSDirectory.Open(new DirectoryInfo(IndexPath), new NativeFSLockFactory());
var isUpdate = IndexReader.IndexExists(directory); //判断索引库是否存在
log.Debug($"The status of index : {isUpdate}");
if (isUpdate)
{
// 如果索引目录被锁定(比如索引过程中程序异常退出),则首先解锁 Lucene.Net在写索引库之前会自动加锁,在close的时候会自动解锁 不能多线程执行,只能处理意外被永远锁定的情况
if (IndexWriter.IsLocked(directory))
{
log.Debug("The index is existed, need to unlock.");
IndexWriter.Unlock(directory); //unlock:强制解锁,待优化
}
}
// 创建向索引库写操作对象 IndexWriter(索引目录,指定使用盘古分词进行切词,最大写入长度限制) 补充:使用IndexWriter打开directory时会自动对索引库文件上锁
var writer = new IndexWriter(directory, new PanGuAnalyzer(), !isUpdate,
IndexWriter.MaxFieldLength.UNLIMITED);
log.Debug($"Total number of task : {indexQueue.Count}");
while (indexQueue.Count > 0)
{
var task = indexQueue.Dequeue();
var id = task.TaskId;
//ArticleService articleService = new ArticleService();
//Article article = articleService.GetById(id);
//if (article == null)
//{
// continue;
//}
//// 一条Document相当于一条记录
//Document document = new Document();
//// 每个Document可以有自己的属性(字段),所有字段名都是自定义的,值都是string类型
//// Field.Store.YES不仅要对文章进行分词记录,也要保存原文,就不用去数据库里查一次了
//document.Add(new Field("id", id.ToString(), Field.Store.YES, Field.Index.NOT_ANALYZED));
//// 需要进行全文检索的字段加 Field.Index. ANALYZED
//// Field.Index.ANALYZED:指定文章内容按照分词后结果保存,否则无法实现后续的模糊查询
//// WITH_POSITIONS_OFFSETS:指示不仅保存分割后的词,还保存词之间的距离
//document.Add(new Field("title", article.Title, Field.Store.YES, Field.Index.ANALYZED,
// Field.TermVector.WITH_POSITIONS_OFFSETS));
//document.Add(new Field("msg", article.Msg, Field.Store.YES, Field.Index.ANALYZED,
// Field.TermVector.WITH_POSITIONS_OFFSETS));
if (task.TaskType != TaskTypeEnum.Add)
{
// 防止重复索引,如果不存在则删除0条
writer.DeleteDocuments(new Term(nameof(id), id.ToString()));// 防止已存在的数据 => delete from t where id=i
}
// 把文档写入索引库
//writer.AddDocument(document);
log.Debug($"Index {id.ToString()} has been writen to index library!");
}
writer.Dispose(); // Close后自动对索引库文件解锁
directory.Dispose(); // 不要忘了Close,否则索引结果搜不到
log.Debug("The index library has been closed!");
}
else
{
Thread.Sleep(2000);
}
}
}
public void AddArticle(IndexTask task)
{
task.TaskType = TaskTypeEnum.Add;
indexQueue.Enqueue(task);
}
public void UpdateArticle(IndexTask task)
{
task.TaskType = TaskTypeEnum.Update;
indexQueue.Enqueue(task);
}
}
public class IndexTask
{
public long TaskId { get; set; }
public TaskTypeEnum TaskType { get; set; }
}
public enum TaskTypeEnum
{
Add,
Update
}
}