[New-bugs-announce] [issue22233] http.client splits headers on none-\r\n characters

[Samuel Charron]

New submission from Samuel Charron:

In some cases, the headers from http.client (that uses email.feedparser) splits headers at wrong separators. The bug is from the use of str.splitlines (in email.feedparser) that splits on other characters than \r\n as it should. (See bug http://bugs.python.org/issue22232)

To reproduce the bug : 

import http.client
c = http.client.HTTPSConnection("graph.facebook.com")
c.request("GET", "/%C4%85", None, {"test": "\x85"})
r = c.getresponse()
print(r.headers)
print(r.headers.keys())
print(r.headers.get("WWW-Authenticate"))

As you can see, the WWW-Authenticate is wrongly parsed (it misses its final "), and therefore the rest of the headers are ignored.

----------
components: Library (Lib)
messages: 225562
nosy: scharron
priority: normal
severity: normal
status: open
title: http.client splits headers on none-\r\n characters
type: behavior
versions: Python 3.4, Python 3.5

_______________________________________
Python tracker <report at bugs.python.org>
<http://bugs.python.org/issue22233>
_______________________________________