def create_importer(page):
importer = Importer(page=page, style='')
resp = urlfetch.fetch(page.url, deadline=10)
if resp.status_code == 200:
soup = BeautifulSoup(resp.content)
parser = CSSParser()
for tag in soup.findAll(re.compile(r'^(link|style)$')):
if tag.name == 'link':
if tag.get('href', None) and tag.get('rel', 'stylesheet').lower() == 'stylesheet':
url = urljoin(page.url, tag['href'])
if urlparse(url).netloc != urlparse(request.url).netloc:
importer.urls.append(url)
elif tag.name == 'style':
media = tag.get('media', None)
sheet = parser.parseString(''.join(tag.contents).strip('\n'), href=url)
style = sheet.cssText
if media:
style = '@media %s {\n%s\n}' % (media, style)
style = '/* Imported directly from %s */\n%s\n' % (page.url, style)
importer.style += style
# Patch around AppEngine's frame inspection
del parser
importer.put()
queue_import(page)
def do_import():
page = Page.get(request.form.get('page_key', ''))
if not page or page.import_state != IMPORTING:
return 'NO_IMPORTER' # We're done
importer = Importer.gql('WHERE page=:1', page.key()).get()
if not importer:
# This requires a request to fetch the page and parse the URLs.
# It also enqueues the next run.
create_importer(page)
return 'CREATED'
if importer.urls:
url = importer.urls.pop(0)
parser = None
try:
resp = urlfetch.fetch(url, deadline=10)
if resp.status_code == 200:
parser = CSSParser()
sheet = parser.parseString(resp.content, href=url)
style = sheet.cssText
importer.style += '\n\n/* Imported from %s */\n%s' % (url, style)
else:
raise Exception('Error fetching %s' % url)
except Exception, e:
import traceback
importer.errors.append('Error importing %s' % url)
logging.error('Error importing for Page %s from %s:\n%s\n%s', page.key().id(), url, e, traceback.format_exc())
finally:
def run(self, initialized):
importer_model = Importer.get(id=self.importer_name, ignore=404)
if importer_model is None:
importer_model = Importer(_id=self.importer_name, last_updated=datetime(1900, 1, 1))
importer_model.save()
try:
for doc_type, doc in self.fetch(importer_model.last_updated):
if doc_type == 'case':
model = Case.get(id=doc['case_id'], ignore=404)
if model is None:
model = Case(_id=doc['case_id'], **doc)
model.save()
elif model.last_updated < dateutil.parser.parse(doc['last_updated']):
model.update(**doc)
elif doc_type == 'notice':
if dateutil.parser.parse(doc['date_published']) > importer_model.last_updated or (initialized == False):
case_id = doc['case_num']+doc['case_code']
existing_case = Case.get(id=case_id, ignore=404)
if existing_case is None:
case_doc = {
'case_id': case_id,
'last_updated': doc['last_updated'],
'address': doc['address'],
'name': 'NEWCASE'
}
new_case = Case(_id=case_id, **case_doc)
new_case.save()
model = Notice(**doc)
model.meta.parent = case_id
model.save()
else:
model = Notice(**doc)
model.meta.parent = case_id
model.save()
else:
continue
importer_model.last_updated = doc['last_updated']
finally:
importer_model.save()
def __init__(self, *args, **kwargs):
method_path = kwargs['method_path']
self.raw_methods = self.load_raw_data(method_path)
self.importer = Importer(db_name=kwargs['db_name'])