def __init__(self, storeUri, phantomjsPath):
"""Instantiate for a given storeUri. Creates the WebriverPool.
@type storeUri: string
@param storeUri: the storeUri
@type storeUri: string
@param storeUri: the phantomjsPath
"""
self.store = FSFilesStore(storeUri)
self.webdrivers = WebdriverPool(phantomjsPath)
class RenderJavascript(object):
"""Rendres the page with JavaScript, takes a screenshot and extract Disqus
and Livefyre comments if present"""
def __init__(self, storeUri, phantomjsPath):
"""Instantiate for a given storeUri. Creates the WebriverPool.
@type storeUri: string
@param storeUri: the storeUri
@type storeUri: string
@param storeUri: the phantomjsPath
"""
self.store = FSFilesStore(storeUri)
self.webdrivers = WebdriverPool(phantomjsPath)
@classmethod
def from_settings(cls, settings):
"""Instantiate with storeUri from settings.
@type settings: scrapy.settings.Settings
@param settings: the settings
@rtype: RenderJavascript
@return: the instantiated class
"""
if not settings['FILES_STORE']:
raise CloseSpider("FILES_STORE setting needed to save screenshots.")
if not settings['PHANTOMJS_PATH']:
raise CloseSpider("PHANTOMJS_PATH setting needed to save screenshots.")
return cls(settings['FILES_STORE'], settings['PHANTOMJS_PATH'])
def close_spider(self, _):
"""Closes the WebriverPool."""
self.webdrivers.stop()
def process_item(self, item, _):
"""JavaScript render item's page in a new thread. Populates
item.screenshot and item.comments if appropriate.
@type item: bibcrawl.model.postitem.PostItem
@param item: the item to process
@type _: scrapy.spider.BaseSpider
@param _: the spider that emitted this item
@rtype: bibcrawl.model.postitem.PostItem
@return: the processed item
"""
defered = deferToThread(self.phantomJSProcess, item)
defered.addCallback(lambda _: _)
defered.addErrback(lambda _: item)
return defered
def phantomJSProcess(self, item):
"""Acquires an idle PhantomJS driver, loads the page, saves screenshot,
download Disqus and LiveFyre comments present and release the driver.
@type item: bibcrawl.model.postitem.PostItem
@param item: the item to process
@rtype: bibcrawl.model.postitem.PostItem
@return: the processed item
"""
driver = self.webdrivers.acquire()
driver.get(item.url)
item.comments = disqusComments(driver) # + livefyreComments(driver)
self.saveScreenshot(item, driver)
self.webdrivers.release(driver)
return item
def saveScreenshot(self, item, driver):
"""Save a screeshot of the current page in storeUri://screen/<HASH>.png.
@type item: bibcrawl.model.postitem.PostItem
@param item: the item to process
@type driver: selenium.webdriver.phantomjs.webdriver.WebDriver
@param driver: the driver
"""
uid = sha1(item.url).hexdigest()
png = StringIO(driver.get_screenshot_as_png())
key = 'screen/{0}.png'.format(uid)
self.store.persist_file(key, png, None)
item.screenshot = key
