def crawl_archives(self):
registry = getUtility(IRegistry)
base_url = registry.get('castle.aws_s3_base_url', None)
storage = archival.Storage(self.site)
urls = []
for key, archive_data in storage.archives.items():
# archives do not need to be re-indexed ever.
# see if the key is in ES, if it is move on
url = archive_data.get('view_url', None) or archive_data['url']
urls.append(aws.swap_url(url, base_url=base_url))
query = {"bool": {"filter": {"term": {"sitemap": "archives"}}}}
existing_urls = self.get_all_from_es(query)
for _id in set(urls) - set(existing_urls):
# pages that have not yet been crawled
try:
self.crawl_archive_url(_id)
except Exception:
logger.error('Error indexing archive url: ' + _id,
exc_info=True)
for _id in set(existing_urls) - set(urls):
# pages that have been removed from the archive
self.delete_from_index(_id)
def test_swap_url(self):
awsurl = 'https://s3-us-gov-west-1.amazonaws.com/bucketname/archives/path/to/resource'
baseurl = 'http://foo.com/'
swappedurl = 'http://foo.com/archives/path/to/resource'
resulturl = aws.swap_url(awsurl, base_url=baseurl)
self.assertEqual(swappedurl, resulturl)
def __call__(self):
shield.protect(self.request, recheck=True)
self.notfound = self.context
self.context = api.portal.get()
if '++' in self.request.URL:
self.request.response.setStatus(404)
try:
return self.index()
except Exception:
logger.warn(
"Failed to render 404 template, had to return simple response"
)
return "not found"
archive_storage = archival.Storage(self.context)
site_url = self.context.absolute_url()
path = self.request.ACTUAL_URL[len(site_url):].rstrip('/')
wants_view = False
if path.endswith('/view'):
wants_view = True
path = path.rsplit('/view', 1)[0]
new_url = None
if path.startswith('/resolveuid'):
uid = path.replace('/resolveuid/', '')
try:
new_url = archive_storage.get_archive_url_by_uid(uid)
except Exception:
pass
else:
try:
new_url = archive_storage.get_archive_url_by_path(
path, wants_view)
except Exception:
pass
if new_url:
# XXX need to force redirect this way since normal redirect
# gets overridden with 404
if self.request.environ.get('QUERY_STRING'):
new_url += '?' + self.request.environ['QUERY_STRING']
raise Redirect(aws.swap_url(new_url))
self.attempt_redirect()
self.request.response.setStatus(404)
return self.index()
def fix_urls(storage, dom):
parsed_endpoint = urlparse(storage.s3_conn.meta.client.meta.endpoint_url)
for Mover in storage.Movers:
mover = Mover(dom)
for el in mover.get_elements():
url = mover.get_url(el)
if url is None:
continue
# check that the url is an s3 url
if not is_s3_url(url, parsed_endpoint):
continue
original = None
if 'original-url' in el.attrib:
# need to maintain the original original
original = el.attrib['original-url']
mover.modify(el, aws.swap_url(url))
if original:
el.attrib['original-url'] = original
def fix_urls(storage, dom):
for Mover in storage.Movers:
mover = Mover(dom)
for el in mover.get_elements():
url = mover.get_url(el)
if url is None:
continue
# check that the url is an s3 url
parsed = urlparse(url)
if parsed.netloc != storage.s3_conn.server_name():
continue
original = None
if 'original-url' in el.attrib:
# need to maintain the original original
original = el.attrib['original-url']
mover.modify(el, aws.swap_url(url))
if original:
el.attrib['original-url'] = original
def __call__(self):
shield.protect(self.request)
self.notfound = self.context
self.context = api.portal.get()
archive_storage = archival.Storage(self.context)
site_url = self.context.absolute_url()
path = self.request.ACTUAL_URL[len(site_url):].rstrip('/')
wants_view = False
if path.endswith('/view'):
wants_view = True
path = path.rsplit('/view', 1)[0]
new_url = None
if path.startswith('/resolveuid'):
uid = path.replace('/resolveuid/', '')
try:
new_url = archive_storage.get_archive_url_by_uid(uid)
except:
pass
else:
try:
new_url = archive_storage.get_archive_url_by_path(path, wants_view)
except:
pass
if new_url:
# XXX need to force redirect this way since normal redirect
# gets overridden with 404
if self.request.environ.get('QUERY_STRING'):
new_url += '?' + self.request.environ['QUERY_STRING']
raise Redirect(aws.swap_url(new_url))
# seems this overrides plone.app.redirector handler
redirector = queryMultiAdapter((self.context, self.request),
name=u'plone_redirector_view')
if redirector:
redirector.attempt_redirect()
return self.index()
def transform_content(self, content, from_url):
parsed_url = urlparse(from_url)
domain = parsed_url.netloc
dom = fromstring(content)
for Mover in self.Movers:
mover = Mover(dom)
for el in mover.get_elements():
url = mover.get_url(el)
if url is None:
continue
if url[0] == '/':
url = '{}://{}{}'.format(parsed_url.scheme, domain, url)
elif 'https://' not in url and 'http://' not in url:
url = urljoin(from_url, url)
# check that the url is on the site...
rdomain = urlparse(url).netloc
if rdomain and domain != rdomain:
continue
if url not in self.resources:
# need to move resource
resource_url = url
if not url.startswith('http'):
resource_url = urljoin(from_url, url)
moved_url = self.move_resource(resource_url,
mover.keep_ext)
if moved_url:
self.resources[url] = moved_url
if url in self.resources:
mover.modify(el, aws.swap_url(self.resources[url]))
content = tostring(dom)
for Util in getAllUtilitiesRegisteredFor(IArchiveContentTransformer):
try:
util = Util(self)
content = util(content)
except Exception:
logger.info('Error with archive utility', exc_info=True)
return content
