Use a proper HTTP client to run all tests. This client is currently
undocumented and should not be used.
Includes a few fixes:
* Fix a bug in the max_keepalive test
* Fix a bug with max_keepalive handling
* Fix a bug in stream_body/1 where data was lost under some conditions
The tests now run quite faster than before.
All the tests now run twice: once for TCP, once for SSL.
Introduces 3 low level functions and updates the existing higher
levels functions. The new primitives are has_body/1, body_length/1
and stream_body/1. In addition to that, a helper function
init_stream/4 has been added.
Streaming a body implies to decode the Transfer-Encoding and
Content-Encoding used for the body. By default, Cowboy will try
to figure out what was used and decode them properly. You can
override this if you want to disable this behavior or simply
support more encodings by calling the init_stream/4 function
before you start streaming the body.
This new protocol option is a fun.
It expects a single arg, the Req, and should only return a possibly
modified Req. This can be used for many things like URL rewriting,
access logging or listener-wide authentication.
If a reply is sent inside the hook, then Cowboy will consider the
request handled and will move on to the next one.
If requests go through a proxy, they will have the original uri in the
request, i.e. : GET http://proxy.server.uri/some/query/string HTTP 1.1 ...
That was problematic -- cowboy_http_protocol:request didn't know what to
to with the result of decode_packet applied to this, which would be something
like:
``` erlang
{http_request,'GET',{absoluteURI,http,<<"proxy.server.uri">>,
undefined,<<"/some/query/string">>},{1,1}}
```
So, I just ignore the host, grab the path and pass into
``` erlang
cowboy_http_protocol:request({http_request, Method, {abs_path, Path},
Version}, State)
```
Seems to do the trick without much effort.
Initially recommended by Magnus Klaar, the trick is to add a catch
instruction before the erlang:hibernate/3 call so that Dialyzer
thinks it will return, followed by the expected return value
('ok' for HTTP, 'closed' for websockets).
This should be good enough until a real solution is found.
We try to send a 500 error only if we didn't send the response
headers yet. If they were, then we have no way to be sure the
response was fully sent, nor should we assume anything about
how this will be handled client-side, so we do nothing more
and in both cases close the connection.
This change makes the dependency on quoted optional
by adding a minimal urldecode function to cowboy.
A protocol option for setting the urldecoding function
has been added to the cowboy_http_protocol module.
The default value for this option is set to be
equivalent to the default settings for quoted.
{fun cowboy_http:urldecode/2, crash}
A note has been added in the README to document
how to use quoted instead of this function.
A field to store this option value has been added
to the state record in the cowboy_http_protocol
module and the http_req record in include/http.hrl
Functions that previously used quoted:from_url/1
has been updated to require an equivalent function
in addition to the previously required arguments.
This change removes a C compiler from the build
requirements of cowboy. It also removes the requirement
to cross compile the code if the target arch/OS
is different from the arch/OS used to build it.
It was replying back the correct error, but with a crash message in
the console. This patch prevents it from crashing.
Fixes issue #94 reported by oribrost.
Allows to limit the size of request and header lines, thus preventing
Cowboy from infinitely reading from the socket and never finding an
end of line.
Defaults to 4096 bytes.
This however does not fix the related Dialyzer warnings.
I have no idea what the warnings are about nor how to fix them,
so feel free to work on it and submit a patch!
Now init/3 can return one of the following values to enable loops:
- {loop, Req, State}
- {loop, Req, State, hibernate}
- {loop, Req, State, Timeout}
- {loop, Req, State, Timeout, hibernate}
Returning one of these tuples will activate looping in the HTTP handler.
When looping, handle/2 is never called. Instead, Cowboy will listen
for Erlang messages and forward them to the info/3 function of the
handler. If a timeout is defined, Cowboy will also close the connection
when no message has been received for Timeout milliseconds.
The info/3 function is defined as info(Msg, Req, State). It can return
either of the following tuples:
- {ok, Req, State}
- {loop, Req, State}
- {loop, Req, State, hibernate}
The first one ends the connection, calling terminate/2 before closing.
The others continue the loop.
Loops are useful when writing long-polling handlers that need to wait
and don't expect to receive anything. Therefore it is recommended to
set a timeout to close the connection if nothing arrives after a while
and to enable hibernate everywhere.
Normal HTTP handlers shouldn't need to use this and as such info/3
was made optional.
From the RFC:
The HEAD method is identical to GET except that the server MUST NOT
return a message-body in the response. The metainformation contained
in the HTTP headers in response to a HEAD request SHOULD be identical
to the information sent in response to a GET request.
Replaces the 'Connection' interpretation in cowboy_http_protocol
from raw value to the parsed value, looking for a single token
matching close/keep-alive instead of the whole raw value (which
could contain more than one token, for example with Firefox 6+
using websocket).
Introduce the functions cowboy_http_req:parse_header/2 and /3
to semantically parse the header values and return a proper
Erlang term.
To this end we are formatting the header names just like OTP does
except we do it for names of up to 32 characters, as there are
widely used header names of more than 20 characters, the limit that
OTP follows currently. An example of such header name would be
Sec-Websocket-Version.
The formatting itself is fairly simple: an uppercase character at
the start and after dashes, everything else lowercase.
Also sends a message 'shoot' that can be received by the protocol
to make sure Cowboy has had enough time to fully initialize the
socket. This message should be received before any socket-related
operations are performed.
WebSocket request connections are now moved from the pool 'default'
to the pool 'websocket', meaning we can have a lot of running
WebSockets despite having a low 'max_connections' setting.
Fixes issue #31.
Recursion shouldn't happen in a single catch statement or inside
a try .. catch statement. The only safe construct for catching
exceptions and perform recursion when everything goes well is
to put the recursive call inside a try .. of .. catch construct
in the of .. catch block.
Otherwise the stack gets filled with exception-related information
since they can still be caught if we were to send them and unfold
the whole thing.
Thanks go to lpgauth for reporting the issue and people on IRC
for explaining the hows and whys.
