python实现海量分词

其中，pyhlseg安装和使用参考
​​​https://toscode.gitee.com/ryan70/pyhlseg#https://gitee.com/ryan70/segEvalTool​​

settings.py

hlseg_table_name='柘荣地名地址_hlseg'   #数据库名称

user_dict_path="HLseg/data/userdict.txt"   #用户词典路径

data_path="HLseg/data/zrbzdz.txt"  # 数据源路径

dbname="addresss"

create_hlseg_Mysql_1.py

import requests
from bs4 import BeautifulSoup
import re
import pymysql
import settings
class Mysqlhlseg(object):
    dbname=settings.dbname
    # 创建表
    def conn(self):
        return pymysql.connect(host='10.0.0.13',user='lyf',password='lyf',db=self.dbname,port=3486,charset='utf8')#连接数据库 
    
    # 创建表
    def create(self,sql,addtbname):
        db= self.conn()
        cursor = db.cursor()
        try:
            cursor.execute(sql)
            db.close()
            print("'{}'数据库表创建成功！！！".format(addtbname))
        except Exception as e:
            print("'{}'数据库表已存在！！！".format(addtbname))
    
    # 插入表数据
    def insert_hlseg(self,hlsegone,addtbname):
        db= self.conn()
        cursor = db.cursor()
        key_list = []
        value_list=[]
        sql = '''insert into {}(%s) value(%s)'''.format(addtbname)
        
        for k,v in hlsegone.items():
            if k!='id':     #id是关键字，自动顺编，不需要人为写入
                key_list.append(k)
                value_list.append('%%(%s)s' % k)
        sql = sql % (','.join(key_list),','.join(value_list))

        try:
            cursor.execute(sql,hlsegone)
            db.commit()
            print(addtbname,'  插入数据成功')
        except Exception as e:
            db.rollback()
            # print("插入数据失败")
        db.close()
    
    # 更新表数据
    def update_hlseg(self,hlsegone,results,addtbname):
        db= self.conn()
        cursor = db.cursor()
        key_list = []
        value_list=[]
        
        sql = '''UPDATE {} SET %s WHERE id=%s'''.format(addtbname) 
        for k,v in hlsegone.items():
            if k!='id':     #id是关键字，自动顺编，不需要人为写入
                key_list.append(k+'='+'%%(%s)s' % k)
                # value_list.append('%%(%s)s' % k)
        sql = sql % (','.join(key_list),60)
        
        cursor.execute(sql,hlsegone)
        # 提交到数据库执行
        db.commit()
        db.close()
    
    def save_hlseg(self,hlsegone,addtbname):
        """ 保存hlseg信息: 如果hlseg不存在就插入, 存在就更新"""
        try:
            results = self.query("SELECT * FROM {} where 'address_part'='%s'".format(addtbname) % (hlsegone['address_part']))
            # 如果hlseg不存在就插入, 否则, 更新原来的hlseg信息
            if len(results) == 0:
                self.insert_hlseg(hlsegone,addtbname)
            else:
                self.update_hlseg(hlsegone,results[0],addtbname)
        except Exception as e:
            print(e)
    # 查询表数据
    def query(self,sql):
        db= self.conn()
        cursor = db.cursor()
        cursor.execute(sql)
        # 返回所有数据
        class_list = cursor.fetchall()
        
        cursor.close()
        db.close()
        return class_list
    # 删除重复数据
    def delete_repeat(self,addtbname,fieldname):
        sql="delete from `{}` where id not in (select a.min_id from (select min(id) as min_id from `{}` group by `{}`)a)".format(addtbname,addtbname,fieldname)
        db=self.conn()
        cursor = db.cursor()
        cursor.execute(sql)
        cursor.close()
        db.close()
if __name__ == "__main__":
    # m=Mysqlhlseg()
    # m.create()
    
    pass
    # create()  #创建表
    
    # #re匹配需要的数据
    # pertern = re.compile(
    #     r'<img.*?data-original="(.*?)".*?<span class="search_now_price">(.*?)</span>.*?<a.*?单品作者.*?title="(.*?)">.*?</a>',
    #     re.S)
    # #添加请求头   修改user-agent来伪装浏览器
    # headers={'User-Agent':'Mozilla/5.0 (Windows NT 10.0; Win64; x64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/70.0.3538.110 Safari/537.36'}
    # url = 'http://category.dangdang.com/cp01.19.34.00.00.00.html'
    # res = requests.get(url,headers=headers)
    # # print(res.status_code)
    # soup = BeautifulSoup(res.text, 'html.parser')
    # data = soup.find_all('ul', attrs={'class': 'bigimg'})
    # data = str(data)
    # item = re.findall(pertern, data)
    # for i in item:
    #     # print(i)
    #     insert(i)

write_hlseg_mysql_2.py

import sys,os,time
sys.path.append(os.path.abspath(os.path.dirname(__file__) + '/' + '..'))

# import sys,time,os
# sys.path.append(".../")
import sys,os,time
# 引用数据库
from create_hlseg_Mysql_1 import Mysqlhlseg
Mysqlhlseg=Mysqlhlseg()
from settings import *
# 加载thulac分词模型

from pyhlseg import *
def load_hlseg():
    HylandaSegment.start_jvm()
    # HylandaSegment.load_dictionary(user_dict_path=HylandaSegment.BUILD_IN_USER_DICT)
    HylandaSegment.load_dictionary(user_dict_path=user_dict_path)
    # HylandaSegment.load_dictionary()
    HylandaSegment.set_option(grain_size=GrainSize.LARGE)
    # 如果分词选项multi_grain_size设为了True，则可以通过change_result_grain_size方法获取其他粒度的分词结果
    #HylandaSegment.set_option(grain_size=GrainSize.LARGE, multi_grain_size=True)
  
def unload_hlseg():
      HylandaSegment.shutdown_jvm()

if __name__ == "__main__":
    load_hlseg()
    t1 = time.time()
    address_dict={}

    sql="""CREATE TABLE {} (
            id int(255) NOT NULL AUTO_INCREMENT,
            address text COLLATE utf8_bin,
            address_part text COLLATE utf8_bin,
            PRIMARY KEY (id)
        ) ENGINE=InnoDB AUTO_INCREMENT=1 DEFAULT CHARSET=utf8 COLLATE=utf8_bin;""".format(hlseg_table_name)
    Mysqlhlseg.create(sql,hlseg_table_name)
    with open(data_path,"r",encoding='utf-8-sig') as f:
        for line in f:
            line = line.strip()
            outs= HylandaSegment.segment(line)
            seg_text = str(outs.toString())
            # result=[out[0] for out in outs]
            address_dict["address"]=line
            address_dict["address_part"]=seg_text
            Mysqlhlseg.insert_hlseg(address_dict,hlseg_table_name)
        
    tt =  time.time()-t1
    print('time ' + str(tt))


    #卸载分词
    unload_hlseg()