如何获取京东商品价格和商品评分

2012年12月15日 22:50

由于课程作业的需求，我们要抓取京东商品的信息。在抓取商品的价格、评论人数以及评分的时候遇到一些麻烦。下面提供我的解决方案。

1. 商品价格：

京东上面的商品价格都是用图片显示的，不过庆幸的时，所有数字的字体、大小、颜色都是一样的。简单起见，直接把二维图片变成灰度图，取个反，影射成一维结构后进行切割和匹配。用了一个简单的匹配评估函数，只是对于数字不变的情况总是能得到正确的结果，如果字体、颜色、大小等变了，估计就得不到正确结果了。具体代码如下

?
price2number.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
#!/usr/bin/python
# coding=utf-8
import Image, sys
 
class PriceReco:
    img_data = []
    size_x, size_y = 0, 0
    def __init__(self, filename): #加载变换图片
        try:
            img = Image.open(filename)
        except:
            print filename, "load error"
            return
        self.size_x, self.size_y = img.size
        self.img_data = list(img.convert('L').getdata())
        for i in range(0, len(self.img_data)):
            self.img_data[i] = 255 - self.img_data[i]
        #print filename, "load success, image size is", self.size_x, self.size_y
        #print self.img_data
         
    def getone(self, single): #识别单个数字
        table_value = [
                [189, 378, 945, 1512, 2079, 1701, 1701, 1134, 945, 378, 189], #￥
                [567, 567], # .
                [1323, 1701, 756, 378, 378, 378, 756, 1701, 1323], # 0
                [378, 378, 2079, 2079, 189, 189], # 1
                [567, 945, 756, 756, 756, 756, 945, 945, 567], # 2
                [756, 1134, 378, 567, 567, 567, 1323, 1512, 756], # 3
                [378, 378, 378, 378, 378, 378, 2079, 2079, 189, 189], # 4
                [378, 1512, 1134, 567, 567, 567, 945, 1134, 756], # 5
                [1134, 1512, 945, 756, 567, 567, 945, 1134, 567], # 6
                [189, 189, 378, 756, 945, 945, 945, 756, 378], # 7
                [756, 1512, 1323, 567, 567, 567, 1323, 1512, 756], # 8
                [567, 1134, 945, 567, 567, 756, 945, 1512, 1134], # 9
                ]
        table_key = ['￥', '.', '0', '1', '2', '3', '4', '5', '6', '7', '8', '9']
        key_id, min_value = 0, 100000
        #print single
        for k in range(0, len(table_key)):
            #print len(table_value[k]), len(single)
            value = 10 * (len(table_value[k]) - len(single)) ** 2
            #print value
            for i in range(0, min(len(table_value[k]), len(single))):
                value += (table_value[k][i] - single[i]) ** 2
            #print value
            if value < min_value:
                key_id, min_value = k, value
                #print "updata: ", key_id, min_value
        #print min_value
        if min_value > 100:
            return 'N'
        else:
            return table_key[key_id]
             
    def recognita(self): #切分和识别图片
        cnt = [0] * self.size_x
        for x in range(0, self.size_x):
            for y in range(0, self.size_y):
                index = y * self.size_x + x
                cnt[x] += self.img_data[index]
        #print cnt
        x = 0
        number = ""
        while x < self.size_x:
            if cnt[x]:
                single = []
                while x < self.size_x and cnt[x]:
                    single.append(cnt[x])
                    x += 1
                number += self.getone(single)
            x += 1
        return number
         
if __name__ == '__main__':
    if len(sys.argv) != 2:
        print "Usage: price_reco image"
    price = PriceReco(sys.argv[1])
    print price.recognita()

2. 评论人数以及商品评分

这个问题就简单多了，虽然京东304(重定向)了n次，稍加分析就可以知道，直接访问如下url就可以得到评论人数和商品评分了

http://club.360buy.com/ProductPageService.aspx?method=GetCommentSummaryBySkuId&referenceId=$id&callback=GetCommentSummaryBySkuId

其中 $id是商品的id，这个得到很简单。

 

于是就成功攻破京东，o(∩∩)o...哈哈！

 

Posted by xhSong Filed in python Tagged by 京东 360buy 商品价格 图片 商品评分 Comments[6]