Age | Commit message (Collapse) | Author |
|
There are better counterparts in the rfc7231_SUITE.
|
|
|
|
Match inside the clause instead. This should make clearer
the occasional failure.
|
|
|
|
|
|
|
|
To attempt to fix a timetrap_timeout issue occuring sometimes.
I have little confidence about this fixing anything but it will
not hurt either.
|
|
If we set Max-Age to 1 we may end up GCing the cookie immediately.
Instead we set it to 2 and increase all time values by 1 second.
|
|
Instead of timeouts, favor receiving events from the event
handler gun_test_event_h. Remove most other timeouts in particular
receive timeouts in favor of a Common Test timetrap timeout global
to this test suite of 30 seconds.
|
|
|
|
Turns out we can't rely on IPv4 being available if we set
the socket in IPv6 mode. Instead the origin is modified to
either setup IPv4 or IPv6 depending on the test's needs.
|
|
|
|
|
|
|
|
The call to Transport:sockname/1 can return an error if the
socket was closed as we were gathering info. In that case we
simply do not fill in the address and port information.
|
|
It seems that the retry behavior changed recently. It would
be better to retrieve the exact values and calculate how much
time we want to wait but that will do for now.
|
|
|
|
|
|
There's no need to have idna listed in ALL_DEPS_DIR we only
use it for "make gen" which does not require it.
|
|
Also contains a few small changes and Dialyzer fixes.
|
|
|
|
|
|
|
|
|
|
|
|
As reported by Dialyzer.
|
|
|
|
|
|
And make sure all HTTP/2 connection_error(s) result in a
gun_down message containing the error. In the preface case
we do not send a gun_error message (because there's no stream
open yet) and gun_down was always saying normal.
Also make sure the human readable reason is included in the
gun_error message, if any.
|
|
|
|
|
|
This should be much faster than using lists:keyfind and friends.
This matters for connections that have a lot of concurrent
streams.
|
|
|
|
|
|
Also rework the different handle_head scenarios in separate
functions.
|
|
|
|
|
|
The details are not given for now. The best would be to
document them in Cowlib and refer to that documentation
instead of duplicating the information in both the Gun
and Cowboy manuals.
|
|
|
|
|
|
|
|
|
|
The reply_to option is also propagated when we switch protocols.
|
|
While at it the gun:info/1 function has been fixed to work
even when we are in the not_connected state, and the owner
is now also returned.
|
|
The auto-ping will at regular interval send a ping frame.
The silence_pings option defaults to true. It can be set
to false when the user needs to receive ping/pong frames.
|
|
|
|
Gun can now be used to send or receive arbitrary data in the
following scenarios:
* Directly after connecting to a server (this is not terribly
useful but it works nevertheless due to the Gun architecture)
* After connecting through one or more Socks and/or HTTP proxies.
This allows using Gun's proxy capabilities to access servers
located beyond firewalls.
* After performing an HTTP/1.1 Upgrade. This allows using Gun
to implement custom protocols that require upgrading from
an HTTP/1.1 connection.
As there is still no support for HTTP/2 CONNECT for the time
being, there are no relevant streams attached to those use
cases and therefore the raw protocol currently expects users
to use 'undefined' as the StreamRef value. This is not a
final decision and will most likely produce a Dialyzer
warning at this time.
|
|
|
|
This essentially reverts to the old behavior for initial TLS
handshakes, and improves TLS support for the Socks/HTTP proxy
cases.
|
|
|