remove '&' from URL patterns

'/?&#' -> '/?#' and '?&#' -> '?#'

According to https://www.ietf.org/rfc/rfc3986.txt, URLs are
"organized hierarchically" by using "the slash ("/"), question
mark ("?"), and number sign ("#") characters to delimit components"
This commit is contained in:
Mike Fährmann
2020-10-22 23:12:59 +02:00
parent 1686dc1757
commit 968d3e8465
74 changed files with 158 additions and 158 deletions

View File

@@ -184,7 +184,7 @@ def generate_extractors():
Extr.category = category
Extr.instance = instance
Extr.pattern = (r"(?:https?://)?" + pattern +
r"/@([^/?&#]+)(?:/media)?/?$")
r"/@([^/?#]+)(?:/media)?/?$")
Extr.test = info.get("test-user")
Extr.root = root
Extr.access_token = token
@@ -197,7 +197,7 @@ def generate_extractors():
Extr.__doc__ = "Extractor for images from a status on " + instance
Extr.category = category
Extr.instance = instance
Extr.pattern = r"(?:https?://)?" + pattern + r"/@[^/?&#]+/(\d+)"
Extr.pattern = r"(?:https?://)?" + pattern + r"/@[^/?#]+/(\d+)"
Extr.test = info.get("test-status")
Extr.root = root
Extr.access_token = token