urllib.robotparser.RobotFileParser.can_fetch crashes on a malformed URL

Bug description:

urllib.robotparser.RobotFileParser.can_fetch is documented to return a bool,
but it passes url straight to urllib.parse._urlsplit, so a malformed URL
(an unterminated IPv6 authority) raises a bare ValueError instead of a
result:

import urllib.robotparser
p = urllib.robotparser.RobotFileParser()
p.parse(["User-agent: *", "Disallow: /"])
p.can_fetch("*", "http://[::1")   # ValueError: Invalid IPv6 URL

The url argument is crawl-frontier-controlled (it comes from parsed pages).

Proposed behavior: a URL that cannot be parsed matches no rule, so can_fetch
returns True (RFC 9309 default that access is not restricted when no rule
applies), rather than raising or being conservatively False.

This is a sibling of #153404 (robotparser raising ValueError on a non-decimal
digit in Crawl-delay / Request-rate); the sink there is the robots.txt
content, here it is the url argument.

CPython versions tested on:

3.13, 3.14, 3.15

Operating systems tested on:

macOS

Linked PRs

• gh-153796

Start at urllib.robotparser.RobotFileParser.can_fetch and reproduce the malformed URL case with an unterminated IPv6 authority. Check how urllib.parse._urlsplit handles the input. Done means can_fetch returns True instead of raising ValueError, with coverage for the malformed URL behavior.