def __init__(self, *a, **kw):
super(SuperSpider, self).__init__(*a, **kw)
self.rd = redis.Redis(settings.get("REDIS_HOST"), settings.get("REDIS_PORT"),
db=settings.get("MAIN_REDIS_DB"))
domain = settings.get("DOMAIN")
self.domain = Domain(self.rd, domain)
self.rule = Rule(self.rd, domain)
settings.overrides['SCHEDULER'] = "scrapy.core.scheduler.Scheduler"
settings.overrides['DOWNLOAD_DELAY'] = float(self.domain["download_delay"])
settings.overrides['CONCURRENT_REQUESTS'] = int(self.domain["concurrent_requests"])
def __init__(self, *a, **kw):
super(SuperSpider, self).__init__(*a, **kw)
self.rd = redis.Redis(settings.get("REDIS_HOST"),
settings.get("REDIS_PORT"),
db=settings.get("MAIN_REDIS_DB"))
domain = settings.get("DOMAIN")
self.domain = Domain(domain)
self.rule = Rule(domain)
settings.set('DOWNLOAD_DELAY', float(self.domain["download_delay"]))
settings.set('CONCURRENT_REQUESTS', int(self.domain["concurrent_requests"]))
class SuperSpider(CrawlSpider):
name = "main"
def __init__(self, *a, **kw):
super(SuperSpider, self).__init__(*a, **kw)
self.rd = redis.Redis(settings.get("REDIS_HOST"),
settings.get("REDIS_PORT"),
db=settings.get("MAIN_REDIS_DB"))
domain = settings.get("DOMAIN")
self.domain = Domain(domain)
self.rule = Rule(domain)
settings.set('DOWNLOAD_DELAY', float(self.domain["download_delay"]))
settings.set('CONCURRENT_REQUESTS', int(self.domain["concurrent_requests"]))
def _get_realurl(self, response, url):
if url.startswith("http://"):
u = url
else:
u = urljoin_rfc(get_base_url(response), url)
return u
def parse(self, response):
flow = Flow(response.meta["flow"])
page_no, = response.xpath(flow['pageno_xpath']).re(flow['pageno_regex'])
if int(page_no) != int(flow["page_limit"]):
next_pages = response.xpath(flow['list_page_xpath']).re(flow['list_page_regex'])
for u in next_pages:
yield Request(url=self._get_realurl(response, u),
meta=response.meta, callback=self.parse)
detail_pages = response.xpath(flow['detail_page_xpath']).re(flow['detail_page_regex'])
for u in detail_pages:
yield Request(url=self._get_realurl(response, u),
meta=response.meta, callback=self.parse_item)
def _load_raw(self, rule, response, jps, sep=''):
raw = ""
if rule:
p_type, rule_case = rule.split("##")
if p_type == 'xpath':
xpath, regex = rule_case.split("#")
rs = response.xpath(xpath).re(regex)
elif p_type == 'jpath':
rs = jps.jpath(rule_case)
elif p_type == 'value':
rs = [rule_case]
elif p_type == 'regex':
rs = re.findall(rule_case, response.body)
raw = sep.join(s.strip() for s in rs)
return raw
def parse_item(self, response):
jps = JsonPathSelector(response)
item = CrawlerItem()
item["url"] = response.url
item["id"] = md5(response.url).hexdigest()
item["domain"] = self.domain.name
item["site"] = self.domain["domain_name"]
item["flow"] = response.meta["flow"]
item["version"] = "1.6"
frs = defaultdict(dict)
for key, rule in self.rule.items():
try:
rule = rule.decode("utf-8")
except:
pass
if key == "domain":
item[key] = rule
continue
sep = ""
if key == "car_images":
sep = "###"
if rule.startswith("fr"):
_, func, frule = rule.split("###")
frs[func][key] = frule
else:
item[key] = self._load_raw(rule, response, jps, sep=sep)
seq = 1
max_seq = len(frs) + 1
if frs:
for fr, cates in frs.items():
fr_func = further_request.__getattribute__(fr)
frequest = fr_func(response)
if frequest:
frequest.meta["fr"] = cates
frequest.meta["url"] = response.url
frequest.meta["id"] = item["id"]
frequest.meta["max_seq"] = max_seq
frequest.meta["seq"] = seq
frequest.callback = self.further_parse
yield frequest
seq += 1
if seq == 1:
seq = 0
item["_max_seq"] = max_seq
item["_seq"] = seq
yield item
def further_parse(self, response):
jps = JsonPathSelector(response)
item = CrawlerItem()
cates = response.meta["fr"]
for key, rule in cates.items():
item[key] = self._load_raw(rule, response, jps)
item["url"] = response.meta["url"]
item["id"] = response.meta["id"]
item["_max_seq"] = response.meta["max_seq"]
item["_seq"] = response.meta["seq"]
return item
class SuperSpider(CrawlSpider):
name = "test"
def __init__(self, *a, **kw):
super(SuperSpider, self).__init__(*a, **kw)
self.rd = redis.Redis(settings.get("REDIS_HOST"), settings.get("REDIS_PORT"),
db=settings.get("MAIN_REDIS_DB"))
domain = settings.get("DOMAIN")
self.domain = Domain(self.rd, domain)
self.rule = Rule(self.rd, domain)
settings.overrides['SCHEDULER'] = "scrapy.core.scheduler.Scheduler"
settings.overrides['DOWNLOAD_DELAY'] = float(self.domain["download_delay"])
settings.overrides['CONCURRENT_REQUESTS'] = int(self.domain["concurrent_requests"])
def start_requests(self):
start_urls = settings.getlist("URL")
for url in start_urls:
request = Request(url=url, callback=self.parse_item, dont_filter=True)
request.meta["flow"] = "test"
yield request
def _load_raw(self, rule, hxs, jps, sep=''):
raw = ""
if rule:
p_type, rule_case = rule.split("##")
if p_type == 'xpath':
xpath, regex = rule_case.split("#")
rs = hxs.select(xpath).re(regex)
elif p_type == 'jpath':
rs = jps.jpath(rule_case)
raw = sep.join(s.strip() for s in rs)
return raw
def parse_item(self, response):
hxs = HtmlXPathSelector(response)
jps = JsonPathSelector(response)
item = CrawlerItem()
item["url"] = response.url
item["id"] = md5(response.url).hexdigest()
frs = defaultdict(dict)
for key, rule in self.rule.items():
try:
rule = rule.decode("utf-8")
except:
pass
if key == "domain":
item[key] = rule
continue
sep = ""
if key == "car_images":
sep = "###"
if rule.startswith("fr"):
_, func, frule = rule.split("###")
frs[func][key] = frule
else:
item[key] = self._load_raw(rule, hxs, jps, sep=sep)
seq = 1
max_seq = len(frs) + 1
if frs:
for fr, cates in frs.items():
fr_func = further_request.__getattribute__(fr)
frequest = fr_func(response)
frequest.meta["fr"] = cates
frequest.meta["url"] = response.url
frequest.meta["id"] = item["id"]
frequest.meta["max_seq"] = max_seq
frequest.meta["seq"] = seq
frequest.callback = self.further_parse
yield frequest
seq += 1
if seq == 1: seq = 0
item["_max_seq"] = max_seq
item["_seq"] = seq
yield item
def further_parse(self, response):
hxs = HtmlXPathSelector(response)
jps = JsonPathSelector(response)
item = CrawlerItem()
cates = response.meta["fr"]
for key, rule in cates.items():
item[key] = self._load_raw(rule, hxs, jps)
item["url"] = response.meta["url"]
item["id"] = response.meta["id"]
item["_max_seq"] = response.meta["max_seq"]
item["_seq"] = response.meta["seq"]
return item
