Run Scrapy crawler in a thread
When you run the Scrapy crawler from a program, the code blocks until the Scrapy crawer is finished. This is due to how Twisted (the underlying asynchronous network library) works. This prevents using the Scrapy crawler from scripts or other code.
To circumvent this issue you can run the Scrapy crawler in a thread with the following code:
CrawlerThread
""" Code to run Scrapy crawler in a thread - works on Scrapy 0.8 """ import threading, Queue from twisted.internet import reactor from scrapy.xlib.pydispatch import dispatcher from scrapy.core.manager import scrapymanager from scrapy.core.engine import scrapyengine from scrapy.core import signals class CrawlerThread(threading.Thread): def __init__(self): threading.Thread.__init__(self) self.running = False def run(self): self.running = True scrapymanager.configure(control_reactor=False) scrapymanager.start() reactor.run(installSignalHandlers=False) def crawl(self, *args): if not self.running: raise RuntimeError("CrawlerThread not running") self._call_and_block_until_signal(signals.spider_closed, \ scrapymanager.crawl, *args) def stop(self): reactor.callFromThread(scrapyengine.stop) def _call_and_block_until_signal(self, signal, f, *a, **kw): q = Queue.Queue() def unblock(): q.put(None) dispatcher.connect(unblock, signal=signal) reactor.callFromThread(f, *a, **kw) q.get()
Usage example
Here's a simple script which uses CrawlerThread to crawl a couple of spiders and print the names of the items found.
import os os.environ.setdefault('SCRAPY_SETTINGS_MODULE', 'myproject.settings') from scrapy.xlib.pydispatch import dispatcher from scrapy.core import signals from scrapy.conf import settings from scrapy.crawler import CrawlerThread settings.overrides['LOG_ENABLED'] = False # avoid log noise def item_passed(item): print "Just scraped item:", item dispatcher.connect(item_passed, signal=signals.item_passed) crawler = CrawlerThread() print "Starting crawler thread..." crawler.start() print "Crawling somedomain.com...." crawler.crawl('somedomain.com) # blocking call print "Crawling anotherdomain.com..." crawler.crawl('anotherdomain.com') # blocking call print "Stopping crawler thread..." crawler.stop()
The output of running of that script would be something like this:
Starting crawler thread... Crawling somedomain.com... Just scraped item: ...some item here... Just scraped item: ...some item here... ... Crawling anotherdomain.com Just scraped item: ...some item here... Just scraped item: ...some item here... ... Stopping crawler thread...
