[issue10231] SimpleHTTPRequestHandler directory bugs

[Jeremy Thurgood]

Jeremy Thurgood <firxen at gmail.com> added the comment:

Thanks for the comments.

There are two separate things here: the URL and the filesystem path. The only part of the URL we care about is the path section, but the fragment ("#anchor") and query parameters ("?foo") are valid -- SimpleHTTPRequestHandler just ignores them. translate_path() turns the URL into the filesystem path, which may be a file or a directory, by extracting the URL path and mapping it onto the filesystem.

The bug is that the fragment and query parameters are stripped in translate_path(), but are *not* stripped when manipulating the URL for the redirect.

This means that when the URL is "/something?foo" and the cwd is "/tmp", the filesystem path is "/tmp/something" (which is a directory) and therefore the response needs to be a redirect. The redirect needs to modify the path section of the URL (which is "/something") to add a slash. This means the redirect needs to be to "/something/" (or "/something/?foo" if you want to preserve the query parameters) rather than "/something?foo/" which is what the current implementation does.

translate_path() unescapes the URL path before mapping it to the filesystem, which means that "/something%3Ffoo" (and even "/something%3Ffoo?bar") will be turned into the filesystem path "/tmp/something?foo".

I'll add some tests for the "/something%3Ffoo" case and possibly update send_head() to preserve the fragment and query parameters on redirect.

----------
status: pending -> open

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue10231>
_______________________________________