def allowed_url(self):
#FIXME: Should use the geturl address as it may have been redirected
scheme, netloc, path, query, fragment = urlsplit(self.url)
robot_url = urlunsplit([scheme, netloc, "robots.txt", "", ""])
#FIXME: Should cache robots.txt in a better persistent data structure
if robot_url in ROBOT_CACHE:
rp = ROBOT_CACHE[robot_url]
else:
rp = RobotExclusionRulesParser()
try:
rp.fetch(robot_url)
# Currently if there's a problem we assume there is no robots.txt
except IOError:
# Should be catching the urllib2.URLError exception
logging.debug("Couldn't retrieve robots.txt for %s" %
robot_url)
rp = None
except UnicodeDecodeError:
logging.debug("Unicode decode error for robots.txt at %s" %
robot_url)
rp = None
except httplib.HTTPException:
logging.debug("Generic HTTPException for robots.txt at %s" %
robot_url)
rp = None
ROBOT_CACHE[robot_url] = rp
if rp is None or rp.is_allowed("*", self.url):
base_url = urlunsplit([scheme, netloc, "", "", ""])
# If there's a current delay on the site respect robots.txt and stall
if self.db.exists(netloc):
logging.debug("Obeying robot overlord for %s..." % netloc)
URLHandler.add_to_busy(self.db, self.url)
return False
# Set a delay for any other requests to this site to respect robots.txt
delay = rp.get_crawl_delay("*") if rp else None
if delay:
delay = int(math.ceil(float(rp.get_crawl_delay("*"))))
else:
delay = SETTINGS["DEFAULT_ROBOTS_DELAY"]
self.db.setex(netloc, "1", delay)
return True
else:
return False
class RerpWrapper(python_common.web.robots_txt.parser_base.RobotsTxtParser):
def __init__(self, content=None, expires=None):
super(RerpWrapper, self).__init__(content, expires)
if content:
self.parser = RobotExclusionRulesParser()
self.parser.use_local_time = False
self.parser.expiration_date = self.expires
self.parser.parse(content)
else:
self.parser = None
self.my_super = super(RerpWrapper, self)
def allowed(self, user_agent, url):
return self.parser.is_allowed(
user_agent, url) if self.parser else self.my_super.allowed(
user_agent, url)
def delay(self, user_agent):
return self.parser.get_crawl_delay(
user_agent) if self.parser else self.my_super.delay(user_agent)
@property
def expired(self):
return self.parser.is_expired if self.parser else self.my_super.expired
@property
def sitemaps(self):
return self.parser.sitemaps if self.parser else self.my_super.sitemaps
def allowed_url(self):
#FIXME: Should use the geturl address as it may have been redirected
scheme, netloc, path, query, fragment = urlsplit(self.url)
robot_url = urlunsplit([scheme, netloc, "robots.txt", "", ""])
#FIXME: Should cache robots.txt in a better persistent data structure
if robot_url in ROBOT_CACHE:
rp = ROBOT_CACHE[robot_url]
else:
rp = RobotExclusionRulesParser()
try:
rp.fetch(robot_url)
# Currently if there's a problem we assume there is no robots.txt
except IOError:
# Should be catching the urllib2.URLError exception
logging.debug("Couldn't retrieve robots.txt for %s" % robot_url)
rp = None
except UnicodeDecodeError:
logging.debug("Unicode decode error for robots.txt at %s" % robot_url)
rp = None
except httplib.HTTPException:
logging.debug("Generic HTTPException for robots.txt at %s" % robot_url)
rp = None
ROBOT_CACHE[robot_url] = rp
if rp is None or rp.is_allowed("*", self.url):
base_url = urlunsplit([scheme, netloc, "", "", ""])
# If there's a current delay on the site respect robots.txt and stall
if self.db.exists(netloc):
logging.debug("Obeying robot overlord for %s..." % netloc)
URLHandler.add_to_busy(self.db, self.url)
return False
# Set a delay for any other requests to this site to respect robots.txt
delay = rp.get_crawl_delay("*") if rp else None
if delay:
delay = int(math.ceil(float(rp.get_crawl_delay("*"))))
else:
delay = SETTINGS["DEFAULT_ROBOTS_DELAY"]
self.db.setex(netloc, "1", delay)
return True
else:
return False
class Robot:
def __init__(self, url):
self.url = Url(urljoin(url, '/robots.txt'))
self.rerp = RobotExclusionRulesParser()
self.rerp.user_agent = 'Mozilla/5.0'
self.rerp.fetch(self.url.url())
def throttle_time(self):
return self.rerp.get_crawl_delay('Mozilla/5.0')
def should_block(self, url):
return not self.rerp.is_allowed('Mozilla/5.0', url.url())
