Tracking size in http connection process state
was unnecessary as it wasn't queried for logic
or obvious reporting and ets table should have
right (within 1) value already.

Got rid of duplicated information in order to
reduce complexity and change points for upcoming
algorithm changes.

Via ets man page:
Note that for tables of the ordered_set type, safe_fixtable/2 is not
necessary as calls to first/1 and next/2 will always succeed

Preparing to rework the pipelining algorithm

Verified that all code paths using saved table
identifier from state are guarded either by a
check for undefined or a call to maybe_create_ets,
so the table should always be there when it
is needed without creating before real use.

Prior version was "2". New functional tests suite
was not being picked up by local rebar, but was
execute with my local which is version "2.5.2"
which targets the same version of Erlang/OTP
as iBrowse (R16).

For consistency, added macro aliases for remaining
ets table names to be found all in one place.

Reuse of ibrowse_lb for named table holding
load balancers and for name of unnamed table
holding connections was confusing.

In preparation for work to ensure that request
load is balanced across connections, created failing
functional test to demonstrate current behavior.

New endpoint will never send back a response and can
be used for verifying timeouts and long running processes.

New eunit based test harness for functional
tests that require a running test server in
order to run.

Moved receive of stop message to listening process
instead of connection and modified listener to come
up for air every once in a while to process.

Fix failing requests when ibrowse_lb process does not stop gracefully.

Enhanced SOCKS5 Support

Supports HTTP and HTTPS destinations and performs all hostname
resolution via the SOCKS proxy, preventing privacy leakage if used
with Tor.

If hostname is a valid IP address then it is used directly, the
hostname is not resolved again via the SOCKS5 proxy (Patch suggested
by Klaus Trainer but implementation here slightly different).

This reverts commit d8150feeaf.

Conflicts:
	.gitignore
	src/ibrowse_http_client.erl
	src/ibrowse_socks5.erl

ibrowse_socks5: fix 'can't find include lib "kernel/src/inet_dns.hrl"'

logic is  wrong in parse_response with the return_raw_request option

if the option return_raw_request` was unset or false, ibrowse was
still returning the returning thhe raw request. This change fix it.
(Spotted while running couchdb tests)

fix send_body pattern matching. Thanks Benoit

fix function_clause error:

        {error,
        {error,
            {'EXIT',
                {{function_clause,
                     [{ibrowse_http_client,do_send_body_1,
                          [{{ok,<<>>},

The case {Function, State} wasn't took in consideration while handling
the body.

Fix silent connection termination - thanks Paul

in that place leads to silent termination, since client will never be
informed. Instead of crashing wait for tcp_closed and cleanup nicely.

Allow missing Content-Length for 303 response.

Run dialyzer by default when compiling. Update travis builds.

Below are the warnings that were fixed by this commit.

[git: master]Iris:ibrowse andreas$ make dialyzer
Building local plt at .dialyzer_plt

dialyzer --output_plt .dialyzer_plt --build_plt \
	   --apps erts kernel stdlib ssl crypto public_key asn1 compiler hipe edoc gs syntax_tools inets xmerl runtime_tools mnesia
  Compiling some key modules to native code... done in 0m19.73s
  Creating PLT .dialyzer_plt ... done in 1m40.98s
done (passed successfully)
dialyzer --fullpath --plt .dialyzer_plt -Wrace_conditions -Wunmatched_returns -Werror_handling -r ./ebin
  Checking whether the PLT .dialyzer_plt is up-to-date... yes
  Proceeding with analysis...
src/ibrowse_http_client.erl:197: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:298: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:322: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:335: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:352: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:582: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:587: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:590: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:706: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:789: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:797: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:1402: Expression produces a value of type 'ok' | {'error',_}, but this value is unmatched
src/ibrowse_http_client.erl:1859: Expression produces a value of type 'ok' | integer(), but this value is unmatched
src/ibrowse_socks5.erl:25: Expression produces a value of type {'ok',<<_:16>>}, but this value is unmatched
 done in 0m8.07s
done (warnings were emitted)
make: *** [dialyzer] Error 2

Add dialyzer target to Makefile

Retry requests when the worker has the connection close

When a tcp_closed or ssl_closed message is handled while ibrowse is
trying to send a request to a worker, ibrowse would incorrectly return a
req_timedout message, when, in fact, the request had not even been
handled.

This patch attempts to avoid that race condition by signalling tcp or
ssl closed messages back to ibrowse differently than a normal exit.

Allow comma-separated list in Transfer-Encoding response header.