質問編集履歴
1
コードを追記しました。
title
CHANGED
|
File without changes
|
body
CHANGED
|
@@ -1,7 +1,90 @@
|
|
|
1
1
|
### 前提・実現したいこと
|
|
2
2
|
|
|
3
3
|
scrapyのcrawlSpiderを用いてデータを収集した後、itemが一つなら scrapy spidername -o file.csv でcsvにできるのですがItemが2つ以上になったときにcsvをItemごとに生成したいです。
|
|
4
|
+
```python
|
|
5
|
+
import logging
|
|
6
|
+
import scrapy
|
|
7
|
+
from scrapy.spiders import CrawlSpider, Rule
|
|
8
|
+
from scrapy.linkextractors import LinkExtractor
|
|
9
|
+
import re
|
|
4
10
|
|
|
11
|
+
from .. import items
|
|
12
|
+
|
|
13
|
+
|
|
14
|
+
class JtnewsSpider(CrawlSpider):
|
|
15
|
+
name = 'jtnews'
|
|
16
|
+
allowed_domains = ['www.jtnews.jp']
|
|
17
|
+
start_urls = ['http://www.jtnews.jp/cgi-bin/revlist.cgi?PAGE_NO=1']
|
|
18
|
+
|
|
19
|
+
rules = (
|
|
20
|
+
Rule(
|
|
21
|
+
LinkExtractor(allow=r"revlist.cgi?&?PAGE_NO=\d+$"), # レビュワー名簿
|
|
22
|
+
callback="parse_user",
|
|
23
|
+
follow=False
|
|
24
|
+
),
|
|
25
|
+
Rule(
|
|
26
|
+
LinkExtractor(allow=r"revper.cgi?&?REVPER_NO=\d+$"), # 個人ページ
|
|
27
|
+
follow=True
|
|
28
|
+
),
|
|
29
|
+
Rule(
|
|
30
|
+
LinkExtractor(allow=r"revper.cgi?&?PAGE_NO=\d+&REVPER_NO=\d+&TYPE=2$"), # レビューの変更日付順化
|
|
31
|
+
callback="parse_review"
|
|
32
|
+
)
|
|
33
|
+
)
|
|
34
|
+
|
|
35
|
+
user_pattern = re.compile(r"REVPER_NO=(?P<user_id>\d+)")
|
|
36
|
+
movie_pattern = re.compile(r"TITLE_NO=(?P<movie_id>\d+)")
|
|
37
|
+
|
|
38
|
+
def parse_user(self, response):
|
|
39
|
+
try:
|
|
40
|
+
user_table = response.css("table.hover-table")
|
|
41
|
+
for link in user_table.css("a"):
|
|
42
|
+
user = items.UserItem()
|
|
43
|
+
user_url = link.css("a::attr(href)").get()
|
|
44
|
+
user["user_id"] = int(self.user_pattern.findall(user_url)[0])
|
|
45
|
+
user["name"] = link.css("a::text").get()
|
|
46
|
+
yield user
|
|
47
|
+
except:
|
|
48
|
+
self.log(f"parse failed: {response.url}", level=logging.ERROR)
|
|
49
|
+
yield scrapy.Request(
|
|
50
|
+
response.url, callback=self.parse_user, dont_filter=True
|
|
51
|
+
)
|
|
52
|
+
|
|
53
|
+
def parse_review(self, response):
|
|
54
|
+
try:
|
|
55
|
+
user_table = response.css("table.normal-table")[2]
|
|
56
|
+
for link in user_table.css("tr")[1:]:
|
|
57
|
+
review = items.ReviewItem()
|
|
58
|
+
review_url = link.css("a::attr(href)")[0].get()
|
|
59
|
+
review["movie_id"] = int(self.movie_pattern.findall(review_url)[0])
|
|
60
|
+
review["title"] = link.css("a::text")[0].get()
|
|
61
|
+
review["point"] = link.css("td::text").get()
|
|
62
|
+
yield review
|
|
63
|
+
except:
|
|
64
|
+
self.log(f"parse failed: {response.url}", level=logging.ERROR)
|
|
65
|
+
yield scrapy.Request(
|
|
66
|
+
response.url, callback=self.parse_review, dont_filter=True
|
|
67
|
+
)
|
|
68
|
+
|
|
69
|
+
```
|
|
70
|
+
|
|
71
|
+
```python
|
|
72
|
+
import scrapy
|
|
73
|
+
|
|
74
|
+
|
|
75
|
+
class ReviewItem(scrapy.Item):
|
|
76
|
+
point = scrapy.Field(serializer=str)
|
|
77
|
+
movie_id = scrapy.Field(serializer=str)
|
|
78
|
+
title = scrapy.Field(serializer=str)
|
|
79
|
+
|
|
80
|
+
|
|
81
|
+
class UserItem(scrapy.Item):
|
|
82
|
+
user_id = scrapy.Field(serializer=str)
|
|
83
|
+
name = scrapy.Field(serializer=str)
|
|
84
|
+
|
|
85
|
+
|
|
86
|
+
```
|
|
87
|
+
|
|
5
88
|
### 試したこと
|
|
6
89
|
|
|
7
90
|
https://qiita.com/bakeratta/items/6fe9030ad838a2a71aa5
|