def parse(self):
tags = [('a', 'href'), ('img', 'src'), ('link', 'href'),
('iframe', 'src')]
# TODO: check rel="canonical"
urls = set([])
for tag_name, attr_name in tags:
for tag in self.doc.findall('.//%s' % tag_name):
attr = tag.get(attr_name)
if attr is None:
continue
url = normalize_url(urljoin(self.url, attr))
if url is not None:
urls.add(url)
on_parse.send(self, urls=urls)
for url in urls:
self.state.crawl(url, path=self.next_path)
def url(self):
url = self.raw_url
if self._has_response():
url = self._response.url
url = normalize_url(url)
return url
def seeds(self):
if not hasattr(self, '_seeds'):
seeds = [normalize_url(s) for s in self.get_list('seed')]
self._seeds = [s for s in seeds if s is not None]
return self._seeds
