#!/usr/bin/env python 
#-*- coding:utf-8 -*- 
 
''''' 
用逆向最大匹配法分词，不去除停用词 
''' 
import codecs 
import xlrd 
 
#读取待分词文本,readlines（）返回句子list 
def readfile(raw_file_path): 
  with codecs.open(raw_file_path,"r",encoding="ANSI") as f: 
    raw_file=f.readlines() 
    return raw_file 
#读取分词词典,返回分词词典list 
def read_dic(dic_path): 
  excel = xlrd.open_workbook(dic_path) 
  sheet = excel.sheets()[0] 
  # 读取第二列的数据 
  data_list = list(sheet.col_values(1))[1:] 
  return data_list 
#逆向最大匹配法分词 
def cut_words(raw_sentences,word_dic): 
  word_cut=[] 
  #最大词长，分词词典中的最大词长,为初始分词的最大词长 
  max_length=max(len(word) for word in word_dic) 
  for sentence in raw_sentences: 
    #strip()函数返回一个没有首尾空白字符(‘\n'、‘\r'、‘\t'、‘')的sentence，避免分词错误 
    sentence=sentence.strip() 
    #单句中的字数 
    words_length = len(sentence) 
    #存储切分出的词语 
    cut_word_list=[] 
    #判断句子是否切分完毕 
    while words_length > 0: 
      max_cut_length = min(words_length, max_length) 
      for i in range(max_cut_length, 0, -1): 
        #根据切片性质，截取words_length-i到words_length-1索引的字，不包括words_length,所以不会溢出 
        new_word = sentence[words_length - i: words_length] 
        if new_word in word_dic: 
          cut_word_list.append(new_word) 
          words_length = words_length - i 
          break 
        elif i == 1: 
          cut_word_list.append(new_word) 
          words_length = words_length - 1 
    #因为是逆向最大匹配，所以最终需要把结果逆向输出，转换为原始顺序 
    cut_word_list.reverse() 
    words="/".join(cut_word_list) 
    #最终把句子首端的分词符号删除，是避免以后将分词结果转化为列表时会出现空字符串元素 
    word_cut.append(words.lstrip("/")) 
  return word_cut 
#输出分词文本 
def outfile(out_path,sentences): 
  #输出模式是“a”即在原始文本上继续追加文本 
  with codecs.open(out_path,"a","utf8") as f: 
    for sentence in sentences: 
      f.write(sentence) 
  print("well done!") 
def main(): 
  #读取待分词文本 
  rawfile_path = r"逆向分词文本.txt" 
  raw_file=readfile(rawfile_path) 
  #读取分词词典 
  wordfile_path = r"words.xlsx" 
  words_dic = read_dic(wordfile_path) 
  #逆向最大匹配法分词 
  content_cut = cut_words(raw_file,words_dic) 
  #输出文本 
  outfile_path = r"分词结果.txt" 
  outfile(outfile_path,content_cut) 
if __name__=="__main__": 
  main()