Update references to RFC2616.

Update references to RFC2616 in HttpStreamParser.

R=bnc@chromium.org,cbentzel@chromium.org
BUG=382120
TEST=none

Committed: https://crrev.com/1fc22bb1eaa8ddf7effa0e3313609291b8fb5f2e
Cr-Commit-Position: refs/heads/master@{#333096}

[wtc, 2015-06-05 01:51:05]

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
File net/http/http_stream_parser.cc (right):

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:843: if (response_body_length_ == 0) {

This is the code that depends on CalculateResponseBodySize() setting
response_body_length_ to 0 when necessary.

I don't understand why it is necessary to move the extra data to the
beginning of read_buf_.

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:976: // [RFC7231]) switch to tunnel mode instead
of having a message body.

I studied the relevant code. It seems useful to set response_body_length_
to 0 for a 2xx (successful) response to a CONNECT request method if we
ever support a protocol where the server speaks first after the proxy
tunnel is established. Right now we only do TLS/SSL and the client speaks
first in a TLS/SSL handshake.

I would hate to add a TODO comment during the Fixit. But I won't have time
to handle the consequences of this change. So I suggest we not change this.
Alternatively, can you suggest a person who can drive this to completion?

The code we would need to add is simple:

  if (request_->method == "CONNECT" && response_->headers->response_code() ==
200)
    response_body_length_ = 0;

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:985: // other words, a server MUST do one of the
following for a 205

Chris: do you think we need to quote the entire paragraph? I think quoting the
first sentence is enough.

[Bence, 2015-06-05 11:28:33]

LGTM.  Please consider updating title to "Update references to RFC2616 in
HttpStreamParser." or similar.

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
File net/http/http_stream_parser.cc (right):

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:976: // [RFC7231]) switch to tunnel mode instead
of having a message body.
On 2015/06/05 01:51:05, wtc wrote:
> 
> I studied the relevant code. It seems useful to set response_body_length_
> to 0 for a 2xx (successful) response to a CONNECT request method if we
> ever support a protocol where the server speaks first after the proxy
> tunnel is established. Right now we only do TLS/SSL and the client speaks
> first in a TLS/SSL handshake.
> 
> I would hate to add a TODO comment during the Fixit. But I won't have time
> to handle the consequences of this change. So I suggest we not change this.
> Alternatively, can you suggest a person who can drive this to completion?
> 
> The code we would need to add is simple:
> 
>   if (request_->method == "CONNECT" && response_->headers->response_code() ==
> 200)
>     response_body_length_ = 0;

I think this will naturally come up if someone wants to implement such a
protocol, so I wouldn't worry about future-proofing at this point.

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:985: // other words, a server MUST do one of the
following for a 205
On 2015/06/05 01:51:04, wtc wrote:
> 
> Chris: do you think we need to quote the entire paragraph? I think quoting the
> first sentence is enough.

I think the first sentence is enough.  Even better, for the purposes of this
method, you could possibly combine it with the previous paragraph: "All 1xx
(informational), 204 (No Content), 205 (Reset Content), and 304 (Not Modified)
responses...".

[cbentzel, 2015-06-05 14:07:31]

LGTM

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
File net/http/http_stream_parser.cc (right):

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:985: // other words, a server MUST do one of the
following for a 205
On 2015/06/05 11:28:33, Bence wrote:
> On 2015/06/05 01:51:04, wtc wrote:
> > 
> > Chris: do you think we need to quote the entire paragraph? I think quoting
the
> > first sentence is enough.
> 
> I think the first sentence is enough.  Even better, for the purposes of this
> method, you could possibly combine it with the previous paragraph: "All 1xx
> (informational), 204 (No Content), 205 (Reset Content), and 304 (Not Modified)
> responses...".

Agree that first sentence is enough. Thanks for finding this quote!

[wtc, 2015-06-05 17:28:51]

On 2015/06/05 11:28:33, Bence wrote:
> LGTM.  Please consider updating title to "Update references to RFC2616 in
> HttpStreamParser." or similar.

Done.

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
File net/http/http_stream_parser.cc (right):

https://codereview.chromium.org/1169603002/diff/20001/net/http/http_stream_pa...
net/http/http_stream_parser.cc:985: // other words, a server MUST do one of the
following for a 205

On 2015/06/05 11:28:33, Bence wrote:
> On 2015/06/05 01:51:04, wtc wrote:
> > 
> > Chris: do you think we need to quote the entire paragraph? I think quoting
the
> > first sentence is enough.
> 
> I think the first sentence is enough.

Done.

[wtc, 2015-06-05 17:29:03]

The CQ bit was checked by wtc@chromium.org

[wtc, 2015-06-05 17:29:03]

The patchset sent to the CQ was uploaded after l-g-t-m from
cbentzel@chromium.org, bnc@chromium.org
Link to the patchset: https://codereview.chromium.org/1169603002/#ps40001
(title: "Shorten the quoted RFC text for the 205 status code.")

[commit-bot: I haz the power, 2015-06-05 17:29:55]

CQ is trying da patch. Follow status at
 https://chromium-cq-status.appspot.com/patch-status/1169603002/40001

[commit-bot: I haz the power, 2015-06-05 18:29:05]

Committed patchset #3 (id:40001)

[commit-bot: I haz the power, 2015-06-05 18:35:20]

Patchset 3 (id:??) landed as
https://crrev.com/1fc22bb1eaa8ddf7effa0e3313609291b8fb5f2e
Cr-Commit-Position: refs/heads/master@{#333096}