jieba分词，识别【带空格的词】

自定义添加【带空格的词】，并分词识别

方法：找到源码的变量进行修改

示例：使【BladeMaster】这类中间有空格的词被识别

jieba

import jieba, re
sentence = 'Blade Master疾风刺杀Archmage'
jieba.add_word('Blade Master')  # 添词
print([word for word in jieba.cut(sentence)])
jieba.re_han_default = re.compile('([\u4E00-\u9FD5a-zA-Z0-9+#&\._% -]+)', re.U)  # 修改格式
print([word for word in jieba.cut(sentence)])

打印结果

[‘Blade’, ’ ', ‘Master’, ‘疾风’, ‘刺杀’, ‘Archmage’]
[‘Blade Master’, ‘疾风’, ‘刺杀’, ‘Archmage’]

jieba.posseg

import jieba, jieba.posseg as jp, re
sentence = 'Demon Hunter斩杀大法师'
jieba.add_word('Demon Hunter', 9, 'hero')  # 添词
jp.re_han_internal = re.compile('(.+)', re.U)  # 修改格式
print(jp.lcut(sentence))

打印结果

[pair(‘Demon Hunter’, ‘hero’), pair(‘斩杀’, ‘v’), pair(‘大法师’, ‘n’)]

import jieba, jieba.posseg as jp, re
sentence = 'D H的D H的DH'
# 修改格式
jp.re_han_internal = re.compile('(.+)', re.U)
# 添词
jieba.add_word('D H')
jieba.add_word('的', tag='DE')
# 打印
print(jp.lcut(sentence))
print(jp.lcut(sentence, HMM=False))

打印结果

[pair(‘D H’, ‘x’), pair(‘的’, ‘DE’), pair(‘D H’, ‘x’), pair(‘的’, ‘uj’), pair(‘DH’, ‘eng’)]
[pair(‘D H’, ‘x’), pair(‘的’, ‘DE’), pair(‘D H’, ‘x’), pair(‘的’, ‘DE’), pair(‘DH’, ‘eng’)]

补充

jieba.posseg的玩法

from jieba import dt
from jieba.posseg import POSTokenizer

text = '正义的洪基伟斩杀邪恶的巨法师'

pos_dt = POSTokenizer(dt)
print(pos_dt.lcut(text))

dt.add_word('巨法师', 1, 'DDD')
print(pos_dt.lcut(text))

dt.add_word('的', tag='DE')
print(pos_dt.lcut(text))

打印结果

[pair(‘正义’, ‘n’), pair(‘的’, ‘uj’), pair(‘洪基伟’, ‘nr’), pair(‘斩杀’, ‘v’), pair(‘邪恶’, ‘a’), pair(‘的’, ‘uj’), pair(‘巨’, ‘a’), pair(‘法师’, ‘n’)]
[pair(‘正义’, ‘n’), pair(‘的’, ‘uj’), pair(‘洪基伟’, ‘nr’), pair(‘斩杀’, ‘v’), pair(‘邪恶’, ‘a’), pair(‘的’, ‘uj’), pair(‘巨法师’, ‘ DDD’)]
[pair(‘正义’, ‘n’), pair(‘的’, ‘ uj’), pair(‘洪基伟’, ‘nr’), pair(‘斩杀’, ‘v’), pair(‘邪恶’, ‘a’), pair(‘的’, ‘ DE’), pair(‘巨法师’, ‘DDD’)]

数字和英文的分词问题

import jieba
s = 'apple均价93600'
print(' '.join(jieba.cut(s)))
jieba.add_word('360')
jieba.add_word('app')
print(' '.join(jieba.cut(s)))

打印结果

apple 均价 93600
app le 均价 9 360 0

解决方法：自写分词器

from os import path
import re
import jieba
from math import log
jieba_dict = path.join(path.dirname(jieba.__file__), 'dict.txt')


class Token:
    re_eng = re.compile('[a-zA-Z][a-zA-Z0-9_-]*')
    re_m = re.compile('[0-9][0-9.+%/~-]*')  # jieba数词标注为m

    def __init__(self, dt, total, max_len):
        self.dt = dt
        self.total = total
        self.max_len = max_len

    @classmethod
    def initialization(cls):
        with open(jieba_dict, encoding='utf-8') as f:
            dt = {line.split()[0]: int(line.split()[1]) for line in f.read().strip().split('\n')}
        # 总频数
        total = sum(dt.values())
        # 词最大长度，默认等于词典最长词（超长英文符会识别不出来）
        max_len = max(len(i) for i in dt.keys())
        return cls(dt, total, max_len)

    def _get_DAG(self, sentence):
        length = len(sentence)
        dt = dict()
        for head in range(length):
            tail = min(head + self.max_len, length)
            dt.update({head: [head]})
            for middle in range(head + 2, tail + 1):
                word = sentence[head: middle]
                # ------------- 词典 + 正则 ------------- #
                if word in self.dt:
                    dt[head].append(middle - 1)
                elif self.re_eng.fullmatch(word):
                    dt[head].append(middle - 1)
                elif self.re_m.fullmatch(word):
                    dt[head].append(middle - 1)
        return dt

    def _calculate(self, sentence):
        DAG = self._get_DAG(sentence)
        route = dict()
        N = len(sentence)
        route[N] = (0, 0)
        logtotal = log(self.total)
        for idx in range(N - 1, -1, -1):
            route[idx] = max(
                (log(self.dt.get(sentence[idx:x + 1], 1)) - logtotal + route[x + 1][0], x)
                for x in DAG[idx])
        return route

    def cut(self, sentence):
        route = self._calculate(sentence)
        x = 0
        N = len(sentence)
        while x < N:
            y = route[x][1] + 1
            l_word = sentence[x:y]
            yield l_word
            x = y

    def lcut(self, sentence):
        return list(self.cut(sentence))

    def add_word(self, word, freq=1):
        original_freq = self.dt.get(word, 0)
        self.dt[word] = freq
        self.total = self.total - original_freq + freq

    def del_word(self, word):
        original_freq = self.dt.get(word)
        if original_freq is not None:
            del self.dt[word]
            self.total -= original_freq


tokenizer = Token.initialization()
cut = tokenizer.cut
lcut = tokenizer.lcut
add_word = tokenizer.add_word
del_word = tokenizer.del_word


if __name__ == '__main__':
    s = '小米60r价值3660rmb'
    print(' '.join(jieba.cut(s)))
    jieba.add_word('60r')
    print(' '.join(jieba.cut(s)))
    print('——')
    print(' '.join(cut(s)))
    add_word('60r')
    print(' '.join(cut(s)))

print

小米 60r 价值 3660rmb
小米 60r 价值 36 60r mb
——
小米 60 r价值 3660 rmb
小米 60r 价值 3660 rmb

add_word和load_userdict

from jieba import dt
print(dt.FREQ)
with open('a.txt', 'w', encoding='utf-8') as f:
    f.write('柳梦璃 99 nr')
dt.load_userdict('a.txt')
print('柳梦璃', dt.FREQ['柳梦璃'])
print('不', dt.FREQ['不'])
dt.add_word('不')
print('不', dt.FREQ['不'])
dt.add_word('不', 9)
print('不', dt.FREQ['不'])

dt.FREQ初始化前是空字典{}

初始化的时候会打印一大串红色文字，初始化完成后，dt.FREQ才不为空