python 全文检索引擎详解

时间:2022-06-22 04:12:54

python 全文检索引擎详解

最近一直在探索着如何用Python实现像百度那样的关键词检索功能。说起关键词检索,我们会不由自主地联想到正则表达式。正则表达式是所有检索的基础,python中有个re类,是专门用于正则匹配。然而,光光是正则表达式是不能很好实现检索功能的。

python有一个whoosh包,是专门用于全文搜索引擎。

whoosh在国内使用的比较少,而它的性能还没有sphinx/coreseek成熟,不过不同于前者,这是一个纯python库,对python的爱好者更为方便使用。具体的代码如下

安装

输入命令行 pip install whoosh

需要导入的包有:

?
1
2
3
4
5
6
7
fromwhoosh.index import create_in
 
fromwhoosh.fields import *
 
fromwhoosh.analysis import RegexAnalyzer
 
fromwhoosh.analysis import Tokenizer,Token

中文分词解析器

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
class ChineseTokenizer(Tokenizer):
  """
  中文分词解析器
  """
  def __call__(self, value, positions=False, chars=False,
         keeporiginal=True, removestops=True, start_pos=0, start_char=0,
         mode='', **kwargs):
    assert isinstance(value, text_type), "%r is not unicode "% value
    t = Token(positions, chars, removestops=removestops, mode=mode, **kwargs)
    list_seg = jieba.cut_for_search(value)
    for w in list_seg:
      t.original = t.text = w
      t.boost = 0.5
      if positions:
        t.pos = start_pos + value.find(w)
      if chars:
        t.startchar = start_char + value.find(w)
        t.endchar = start_char + value.find(w) + len(w)
      yield t
 
 
def chinese_analyzer():
  return ChineseTokenizer()

构建索引的函数

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
@staticmethod
  def create_index(document_dir):
    analyzer = chinese_analyzer()
    schema = Schema(titel=TEXT(stored=True, analyzer=analyzer), path=ID(stored=True),
            content=TEXT(stored=True, analyzer=analyzer))
    ix = create_in("./", schema)
    writer = ix.writer()
    for parents, dirnames, filenames in os.walk(document_dir):
      for filename in filenames:
        title = filename.replace(".txt", "").decode('utf8')
        print title
        content = open(document_dir + '/' + filename, 'r').read().decode('utf-8')
        path = u"/b"
        writer.add_document(titel=title, path=path, content=content)
    writer.commit()

检索函数

?
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
@staticmethod
 def search(search_str):
   title_list = []
   print 'here'
   ix = open_dir("./")
   searcher = ix.searcher()
   print search_str,type(search_str)
   results = searcher.find("content", search_str)
   for hit in results:
     print hit['titel']
     print hit.score
     print hit.highlights("content", top=10)
     title_list.append(hit['titel'])
   print 'tt',title_list
   return title_list

感谢阅读,希望能帮助到大家,谢谢大家对本站的支持!

原文链接:http://blog.csdn.net/qq_30843221/article/details/48184311