This module provides a decorator to allow to write Scrapy's spider callbacks which performs multiple requests without the need to write multiple callbacks for each request.
The code is experimental, might not work in all cases and even might be hard to debug.
from inline_requests import inline_requests class MySpider(CrawlSpider): ... @inline_requests def parse_item(self, response): item = self.build_item(response) # scrape more information response = yield Request(response.url + '?info') item['info'] = self.extract_info(response) # scrape pictures response = yield Request(response.url + '?pictures') item['pictures'] = self.extract_pictures(response) # a request that might fail (dns error, network timeout, error 404/500, etc) try: response = yield Request(response.url + '?protected') except Exception as e: log.err(e, spider=self) else: item['protected'] = self.extract_protected_info(response) # finally yield the item yield item
The example directory includes a example spider for StackOverflow.com:
cd example scrapy crawl stackoverflow
- Python 2.6+
- Scrapy 0.14+