python推荐淘宝物美价廉商品

完成的目标:

  输入搜索的商品 以及 淘宝的已评价数目、店铺的商品描述(包括如实描述、服务态度、快递的5.0打分);

  按要求,晒选出要求数量的结果,并按“物美价廉算法”排序后输出    

思路:

1,利用淘宝搜索'https://s.taobao.com/search?'的价格filter 先进行价格筛选,得到结果的网站

2,用urllib打开结果网站,构造正则表达式匹配出各个商品结果的 价格、已评价数量、店铺的如实描述等信息;

  并把结果保存至二维数组里。

3,利用商品及店铺信息,用“物美价廉算法”给各个商品打分

4,按打分排序, 各个信息总结果按排序输出到新建txt文件里;

  并将各个商品图片下载到文件及建立相同排序开头的txt(其名字包好简要的商品信息),这样图片和商品信息同时能在一个文件夹里用大图排列看到。

5.,可以把输入的参数(价格范围等要求)以函数输入,,用pyinstaller 把整个py程序打包为EXE 就可以发布了。

 “物美价廉算法”说明:

  利用商铺顾客的评分作为主要筛选保证质量,价格偏离平均值的幅度筛选便宜的商品,销量热门度以已有评价数度量。具体算法见源代码,参数与函数都是自己调试的,因为抓取的评分5分制分布比较集中,从而算法 提高了评分的重要性,削弱价格、销量的重要性(销量越高,增长一个级别体现的评价统计显著性越不重要,因为二八法则,以及很多消费者是都跟着销量买的,有叠加效应)

  优点就是,结果里商品很少有差评的,都是高分评价

  这样推荐的缺点是,可能会高估销量较少的商铺、新开商铺的商品质量。因为总的评价的人少,打分的体现的价值就不稳定,高分的偶然性也越大,不过是按照淘宝综合排序搜索前多少个满足的,销量还算是比较靠前的,只要不是那些市场份额很少商家也少的商品。另一个问题是 淘宝和天猫没有区分,淘宝的审查没有天猫的严格。

如要求条件为:

reserch_goods='ssd120g' #淘宝搜索词
price_min=22 #价格区间
price_max=400
descripHrequ=0 # % 默认高于average, 输出结果大于此值
servHrequ=0 # % 默认高于average, 输出结果大于此值
descripNrequ=6
counts=10 #要求选出多少个商品

结果显示在results文件里

 打包发布的EXE体验版分享在链接:http://pan.baidu.com/s/1boXwjTH 密码:m32i

源代码如下:

  1 # -*- coding: utf-8 -*-
  2 '''
  3 author : willowj
  4 http://www.cnblogs.com/willowj
  5 注意robots 协议,别给网站带来较多负载
  6 '''
  7 
  8 import urllib
  9 import urllib2
 10 import re
 11 import time 
 12 import random
 13 import os
 14 from math import log
 15 from math import log10
 16 from math import sqrt
 17 import sys
 18 
 19 '''在Python自己IDE上要注释掉一下两行'''
 20 reload(sys)  
 21 sys.setdefaultencoding('utf8')  # python2.x的的defaultencoding是ascii
 22 
 23 class counter(object):
 24     #计数器
 25     def __init__(self):
 26         self.count  = 0
 27         self.try_time = 0
 28         self.fail_time = 0
 29         self.url_list = []
 30         self.new_flag = True
 31         self.results=[]
 32         self.p=0
 33         self.d=0
 34 
 35     def print_counter(self):
 36         print 'try_time:', self.try_time,   "  get_count:" , self.count,   "  fail_time:",self.fail_time
 37 
 38 counter1 = counter()
 39 
 40 
 41 def post_request(url):
 42     '''
 43     #使用代理
 44     proxy = {'http':'27.24.158.155:84'}
 45     proxy_support = urllib2.ProxyHandler(proxy)
 46     # opener = urllib2.build_opener(proxy_support,urllib2.HTTPHandler(debuglevel=1))
 47     opener = urllib2.build_opener(proxy_support)
 48     urllib2.install_opener(opener)
 49     '''
 50 
 51     #构造随机头部文件访问请求
 52     User_Agents=["Mozilla/5.0 (Windows NT 10.0; Win64; x64; rv:47.0) Gecko/20100101 Firefox/47.0",
 53     "Mozilla/5.0 (Windows NT 10.0; WOW64) AppleWebKit/537.36 (KHTML, like Gecko) Chrome/50.0.2661.102 Safari/537.36",
 54     "Mozilla/5.0 (Macintosh; U; Intel Mac OS X 10_6_8; en-us) AppleWebKit/534.50 (KHTML, like Gecko) Version/5.1 Safari/534.50", #
 55     "Opera/9.80 (Windows NT 6.1; U; en) Presto/2.8.131 Version/11.11",
 56     "Opera/9.80 (Macintosh; Intel Mac OS X 10.6.8; U; en) Presto/2.8.131 Version/11.11O"
 57     ]
 58     random_User_Agent = random.choice(User_Agents)
 59     #print random_User_Agent
 60 
 61     req =urllib2.Request(url) #!!
 62 
 63     req.add_header("User-Agent",random_User_Agent)
 64     req.add_header("GET",url)
 65     req.add_header("Referer",url)
 66     return req
 67 
 68 
 69 def recommend_rate(price,description,delivery,service,comments):
 70     #描述为绝对值
 71     av_p=counter1.p/counter1.count
 72     av_d=counter1.d/counter1.count
 73     rate=(description/av_d)**20*(description+delivery+service)*(av_p/(price))**0.1+log((comments+5),1000)
 74     #print 'all count=',counter1.count
 75     #print "avrage price=",av_p,';',av_p/(price),';price',price,';comments=',comments,';descrip=',description
 76     #print 'rate=',rate,'(price)yinzi',(av_p/(price))**0.1,'descrip_yinzi',(description/av_d)**20,'comments_factor=',log((comments+50),100)
 77     return rate
 78 
 79 
 80 def product_rank(list):
 81     for x in list:
 82         #0开始为  x0商品名 、x1图片链接、x2商品链接、x3价格、x4评论数、 x5店铺名、 x6快递情况、x7描述相符情况3个、x8服务情况 
 83         rate=recommend_rate(x[3],x[7],x[6],x[8],x[4])
 84         x.append(rate)
 85 
 86 
 87 def get_user_rate(item_url):
 88     #暂时未使用该功能
 89     '''获取卖家信用情况;未登录情况不能访问,或者需要在头部文件中加入cookie。。。;'''
 90     html=urllib2.urlopen(item_url)
 91     #"//rate.taobao.com/user-rate-282f910f3b70f2128abd0ee9170e6428.htm"
 92     regrex_rate='"(//.*?user-rate.*?)"'
 93     codes= re.findall(regrex_rate,html.read())
 94     html.close()
 95 
 96     user_rate_url= 'http:'+codes[0]
 97     print 'uu', user_rate_url
 98 
 99     user_rate_html = urllib2.urlopen(user_rate_url)
100     print user_rate_html.read()
101     #title="4.78589分"
102     desc_regex=u'title="(4.[0-9]{5}).*?'
103     de_pat=re.compile(desc_regex)
104     
105     descs = re.findall(de_pat,user_rate_html.read())
106     print len(descs)
107     item_url='https://item.taobao.com/item.htm?id=530635294653&ns=1&abbucket=0#detail'
108 #get_user_rate(item_url)
109 '''获取卖家信用情况;未登录情况不能访问。。。暂时 无用'''
110 
111 def makeNewdir(savePath):
112     while os.path.exists(savePath):
113         savePath = savePath+'%s'%random.randrange(1,10)
114         #print "the path exist,we'll make a new one"
115     try:
116         os.makedirs(savePath)
117         print 'ok,file_path we reserve results:  %s'%savePath
118         print '保存的路径为:'.decode('utf-8')
119     except:
120         print "failed to make file path
please restart program"
121         print '创建文件夹失败,请重新启动程序'.decode('utf-8')
122 
123 
124 def get_praised_good(url, file_open, keyword, counts, descripHrequ, servHrequ, descripNrequ):
125     #从给定的淘宝链接中 获取符合条件的商品list
126     html = urllib2.urlopen(post_request(url))
127     code = html.read()
128     html.close()
129 
130     regrex2=ur'raw_title":"(.*?)","pic_url":"(.*?)","detail_url":"(.*?)","view_price":"(.*?)".*?"comment_count":"(.*?)".*?"nick":"(.*?)".*?"delivery":[(.*?),(.*?),(.*?)],"description":[(.*?),(.*?),(.*?)],"service":[(.*?),(.*?),(.*?)]' 
131     #每一个匹配项 返回  15个 字符串 
132     #x[0]开始为  x0商品名 、x1图片链接、x2商品链接、x3价格、x4评论数、 x5店铺名、 x6快递情况3个、x9描述相符情况3个、x12服务情况3个
133     pat = re.compile(regrex2)
134     meet_code = re.findall(regrex2,code)#
135 
136     for x in meet_code:
137         if counter1.count>=counts :
138             print "have get enough pruducts"
139             break
140         description_higher=int(x[10])*float(x[11])/100
141         service_higher=int(x[13])*float(x[14])/100
142         try:
143             x4=int(x[4]) #description_count
144         except:
145             x4=0
146         if  (description_higher>=descripHrequ) and (service_higher>=servHrequ) and x4>=descripNrequ:
147             if re.findall(keyword,x[0]) : # 中文keyword在结果中匹配问题暂时没有解决,,直接加在搜索词里吧 
148                 x0=x[0].replace(' ','').replace('/','')
149                 detail_url='http:'+x[2].decode('unicode-escape').encode('utf-8')
150                 x1='http:'+x[1].decode('unicode-escape').encode('utf-8')
151                 #print type(x)
152                 if detail_url  in counter1.url_list:
153                     counter1.new_flag=False
154                     print 'no more new met products'
155                     print counter1.url_list
156                     print detail_url
157                     break
158                 counter1.url_list.append(detail_url)
159                 counter1.try_time+=1
160                 counter1.count+=1
161 
162                 x11=float(x[11])/100
163                 x9=float(x[9])/100
164                 x12=float(x[12])/100
165                 x6=float(x[6])/100
166                 x3=float(x[3])
167                 counter1.p+=x3
168                 counter1.d+=x9
169                 x5=unicode(x[5],'utf-8')
170                                 
171                 result_list=[]
172                 result_list.append(x0)
173                 result_list.append(x1)
174                 result_list.append(detail_url)
175                 result_list.append(x3)
176                 result_list.append(x4)
177                 result_list.append(x5)
178                 result_list.append(x6)
179                 result_list.append(x9)
180                 result_list.append(x12)
181                 #0开始为  x0商品名 、x1图片链接、x2商品链接、x3价格、x4评论数、 x5店铺名、 x6快递情况、x7描述相符情况、x8服务情况 
182                 counter1.results.append(result_list)
183             
184 
185 def save_downpic(lis,file_open,savePath):
186     '''从商品list下载图片到reserve_file_path,并写入信息至fileopen'''
187     #0开始为  x0商品名 、x1图片链接、x2商品链接、x3价格、x4评论数、 x5店铺名、 x6快递情况、x7描述相符情况、x8服务情况、x9:rate
188     len_list=len(lis)
189     print len_list
190     cc=0        
191     for x in lis:
192         try :
193             urllib.urlretrieve(x[1],savePath+'\%s___'%cc +unicode(x[0],'utf-8')+'.jpg')
194 
195             txt_name = savePath+'\'+ '%s__'%cc+ 'custome_description_%s __'%x[7] +'__comments_%s_'%x[4]+ '___price_%srmb___'%x[3] +x[5] +'.txt'
196                     
197             file_o = open(txt_name,'a')
198             file_o.write(x[2])
199             file_o.close()
200             
201             print '
get_one_possible_fine_goods:
','good_name:',x[0].decode('utf-8')
202             print 'rate=',x[9]
203             print 'price:',x[3],x[5].decode('utf-8')
204             print 'custome_description:',x[7],'--','described_number:',x[4],'  service:',x[8]
205             print x[2].decode('utf-8'),'
good_pic_url:',x[1].decode('utf-8')
206 
207             print txt_name
208             print cc+1,"th"
209 
210             file_open.write(u'%s__'%cc +u'%s'%x[0]+'
price:'+str(x[3])+'¥,'+'
'+str(x[2])+'  
'+str(x[5])+'
customer_description:'+str(x[7])+'described_number:'+str(x[4])+'


')
211             
212             print 'get one -^-'
213         except :
214             print "failed to down picture or creat txt"
215             counter1.fail_time += 1
216         cc+=1
217         time.sleep(0.5)
218 
219 
220 def get_all_praised_goods(serchProd,counts,savePath ,keyword, price_min=0,price_max=0,descripHrequ =0,servHrequ=0 ,descripNrequ=0):
221     #边里搜索结果每一页
222     #initial url and page number
223     initial_url='https://s.taobao.com/search?q='+serchProd
224     if  price_min :
225         if price_min < price_max :
226             initial_url = initial_url+'&filter=reserve_price%5B'+'%s'%price_min+'%2C' +'%s'%price_max
227 
228     initial_url = initial_url +'%5D&s='
229 
230     #tian_mall = 'https://list.tmall.com/search_product.htm?q='
231 
232     print "initial_url",initial_url
233     page_n=0
234     reserve_file=savePath+r'found_goods.txt'
235     file_open=open(reserve_file,'a')
236 
237     file_open.write('****************************
')
238     file_open.write(time.ctime())
239     file_open.write('
****************************
')
240 
241     while counter1.new_flag and counter1.count<counts :
242         
243         url_1=initial_url+'%s'%(44*page_n)
244         #print initial_url
245         print 'url_1:', url_1
246         #print 'ss',initial_url+'%s'%(44*page_n)
247         page_n += 1
248 
249         get_praised_good(url_1,file_open,keyword,counts,descripHrequ,servHrequ ,descripNrequ)
250         print "let web network rest for 2s lest  make traffic jams "
251         time.sleep(2)
252         # except:
253         print "%s"%page_n,"pages have been searched"            
254         if page_n >=11 :
255             print "check keyword,maybe too restrict"
256             break
257     print url_1        
258     product_rank(counter1.results)
259 
260     counter1.results.sort(key=lambda x :x[9],reverse=True)        
261 
262     save_downpic(counter1.results,file_open,savePath)
263     
264     #
265     for a in  counter1.results:
266         for b in a :
267             file_open.write(unicode(str(b),'utf-8'))
268             file_open.write('	')
269         file_open.write('

')
270     
271     file_open.close()
272     counter1.print_counter()
273 
274 
275 def main(serchProd):
276     ''' 用户选择是否自定义要求,根据要求进行获取商品,并按推荐排序输出'''
277     if serchProd:
278 
279         print "if customise price_range ,decriptiom require .etc.
input Y/N 
 default by : no price limit avarage than descriptiom,get 30 products 
 默认要求为:无价格限制,商品描述、快递、服务高于均值,获取30个商品。自定义要求请输入 ‘Y’ (区分大小写)".decode('utf-8')
280         if raw_input() == 'Y':
281             print "
please input  _minimal price and _maximal price;   
default by 0,10000
next by 'enter'key input nothing means by default,the same below "
282             print '请输入价格范围 ;默认0-10000 ;两项用半角逗号","分隔 按回车键确认;什么也不输入代表使用默认值 '.decode('utf-8')
283             try:
284                 price_min, price_max=input()
285             except:
286                 print 'not input or wrong number,use default range'
287                 price_min, price_max = 0 ,10000
288             #
289                 # #
290             print "please input _keyword that goods name must include:
(more than one keyword must use Regular Expression); default by no kewords"
291             try:
292                 keyword=raw_input().decode("gbk").encode("utf-8")      #个人限定词,商品名字必须包含,防止淘宝推荐了其他相关词 (正则表达式). 为任意表示不作限制
293             except:
294                 keyword=''
295             #    
296 
297             print "
please input  _description_higher_percent_require and _service_higher__percent_require
 range:(-100,100) ;   
default by 0,0  I.e better than average"
298             print '请输入商品描述、服务高于平均值的百分比-100 ~100'.decode('utf-8')
299                  # %   默认高于average, 输出结果大于此值
300             try:
301                 descripHrequ,servHrequ=input()              
302             except:
303                 print 'not input or wrong number,use default range'
304                 descripHrequ = 0  # %  默认高于average, 输出结果大于此值
305                 servHrequ = 0
306             
307             #    
308             print "
please input description count limit,  default more than 5
" ,'输入最低商品评价数,默认大于5'.decode('utf-8')
309             try:
310                 descripNrequ=input()
311             except :
312                 print 'not input or wrong number,use default range'
313                 descripNrequ=5
314             #
315                 
316                 # print "
IF customise file reserve path, Y or N  
default/sample as:  C:\Users\Administrator\Desktop\find_worthy_goods\results "
317                 # print '是否自定义保存文件目录 Y or N'.decode('utf-8')
318                 # if raw_input()=='Y':
319                 #     print "please input path that you want to reserve;  
 "    
320                 #     savePath = raw_input()
321                 # else:
322                 #     #savePath=r"C:UsersAdministratorDesktopfind_worthy_goods
esults"#结果保存路径        
323             #
324             print "
please input how many results you want,  default by 30
" ,'您要获取的商品数目,默认30'.decode('utf-8')
325             try:
326                 counts=input()
327             except :
328                 counts=30
329         else :
330             counts =30
331             keyword = ''
332             price_min ,price_max ,descripHrequ ,servHrequ ,descripNrequ = 0,0,0,0,0
333 
334         #
335         savePath='results'
336         makeNewdir(savePath)
337 
338         get_all_praised_goods(serchProd, counts, savePath, keyword, price_min ,price_max ,descripHrequ ,servHrequ ,descripNrequ)
339         
340         print '
'
341         counter1.print_counter()
342         print "finished,please look up in %s"%savePath
343         print '下载完成'.decode('utf-8')
344 
345         input()
346     else:
347         print "no search goods,please restart"
348         print '没有输入商品名称,请重新启动程序'.decode('utf-8')
349 
350 
351 def input_para_inner():
352     serchProd='ssd120g'     #淘宝搜索词
353     keyword='非240' #raw_input().decode("gbk").encode("utf-8")        #个人限定词,商品名字必须包含,防止淘宝推荐了其他相关词 (正则表达式). 为任意表示不作限制
354     price_min=22            #价格区间
355     price_max=400
356     descripHrequ=0   # %   默认高于average, 输出结果大于此值
357     servHrequ=0      # %  默认高于average, 输出结果大于此值
358     descripNrequ=6
359     counts=10    
360                         #要求选出多少个商品
361     #savePath=r"C:UsersAdministratorDesktopPython scrapyfind_worthy_goods
esults"#结果保存路径
362     savePath=r"results" #结果保存路径
363     makeNewdir(savePath)
364     
365     get_all_praised_goods(serchProd, counts, savePath, keyword, price_min, price_max ,descripHrequ ,servHrequ ,descripNrequ)
366 
367 
368 if __name__=="__main__" :
369     x = 2
370     if x==1:
371         input_para_inner() #自己在源代码中输入 筛选要求
372     else:
373         print '说明:
本程序用于在淘宝上搜索商品时主动通过 价格范围、商品描述、服务态度、评论数来筛选商品;
筛选出来的商品图片下载保存到磁盘(默认桌面新建find_worty_goods文件夹)并建立同序号开头的txt文件,图片显示商品,其旁的txt文件名显示价格等关键信息,txt里保存商品的淘宝链接'.decode('utf-8')    
374         print "please input reserch _goods_name"
375         serchProd=raw_input().replace(' ','')    #淘宝搜索词 ,并去除中间意外输入的空格
376         main(serchProd)
377     
378     
379     #保存图片,以文件名为商品图片名字,并以序号开头
380     #同时,输出 价格、商家名,商品描述、服务等 到 txt文本
381     #在商品图片看中后,便可按序号查找 
382     #按描述、服务评价高于平均,购物体验应该可以的
View Code

 预计可添加功能:

  交互界面

  MySQL的数据存储,以实现价格变动的比较

原文地址:https://www.cnblogs.com/willowj/p/6238406.html