diff options
Diffstat (limited to 'python/sandcrawler/misc.py')
| -rw-r--r-- | python/sandcrawler/misc.py | 7 |
1 files changed, 7 insertions, 0 deletions
diff --git a/python/sandcrawler/misc.py b/python/sandcrawler/misc.py index 88669e6..d9c9d55 100644 --- a/python/sandcrawler/misc.py +++ b/python/sandcrawler/misc.py @@ -6,8 +6,15 @@ import datetime import requests from requests.adapters import HTTPAdapter from requests.packages.urllib3.util.retry import Retry # pylint: disable=import-error +import urlcanon +def clean_url(s): + parsed = urlcanon.parse_url(s) + if not parsed.port and parsed.colon_before_port: + parsed.colon_before_port = b'' + return str(urlcanon.whatwg(parsed)) + def gen_file_metadata(blob): """ Takes a file blob (bytestream) and returns hashes and other metadata. |