我希望使用循环连续爬行和更新数组值,因为我需要单击某个按钮才能获得数组上的下一个值。然而,在循环内部,它似乎像并行线程一样工作,并且这个项目产生了很多次。我想要的是只进行一次循环,更新数据和生成项目。示例:当前输出:
{'field1': 'data1',
'filed2' : 'data2',
'field3' : ['data31']}
{'field1': 'data1',
'filed2' : 'data2',
'field3' : ['data32']}
{'field1': 'data1',
'filed2' : 'data2',
'field3' : ['data33']}预期:
{'field1': 'data1',
'filed2' : 'data2',
'field3' : ['data31', 'data32', 'data3']}这是我的密码
def parse_individual_listings(self, response):
...
data = {}
data[field1] = 'data1'
data[field1] = 'data2'
...
for i in range(3):
yield scrapy.Request(
urlparse.urljoin(response.url, link['href']), #different link
callback=self.parse_individual_tabs,
meta={'data': data, 'n':i};
)
def parse_individual_tabs(self, response):
data = response.meta['data']
i = response.meta['i']
...
# keep populating `data`
data[field3][i] = "data3[i]" #this value change when I click a button to update
yield data发布于 2019-04-19 08:28:58
尝试使用inline_requests库(https://pypi.org/project/scrapy-inline-requests/)。让我们在同一个函数中提出请求。将数据收集到一个对象而不是产生不同的数据是有用的。用一些伪代码检查这个示例:
from inline_requests import inline_requests
from scrapy import Selector
@inline_requests
def parse_individual_listings(self, response):
...
data = {}
data[field1] = 'data1'
data[field1] = 'data2'
...
for i in range(3):
extra_req = yield scrapy.Request(
response.urljoin(link['href']), #different link
)
# apply your logics here, say extract some data
sel = Selector(text=extra_req.text)
data['field3'].append(sel.css('some css selector').get())
yield data发布于 2019-04-19 09:48:25
遵循多线程和不同步的思想,我最终在parse_individual_tabs中使用互斥锁,它看起来工作得很好。
from threading import Lock
...
mutex = Lock()
count = 0
...
def parse_individual_tabs(self, response):
self.mutex.acquire(1)
try:
count += 1
data = response.meta['data']
i = response.meta['i']
...
# keep populating `data`
data[field3][i] = "data3[i]" #this value change when I click a button to update
finally:
self.mutex.release()
if (count == 3): #check if this is callback of last yield Request
yield data
else:
returnhttps://stackoverflow.com/questions/55758718
复制相似问题