sharpetronics/httpx
1
0
Fork 0
mirror of https://github.com/HoneyryderChuck/httpx.git synced 2025-10-05 00:02:38 -04:00
Code Issues Projects Releases Wiki Activity
1,537 Commits 52 Branches 126 Tags
Commit Graph

477 Commits

Author SHA1 Message Date
HoneyryderChuck
a29961962b allowing multipart requests to be retried (although, to be fair, user has to push a lot of knobs to do so) 2021-01-13 13:40:05 +00:00
HoneyryderChuck
7fa283097d added signatures to new modules 2021-01-13 12:27:13 +00:00
HoneyryderChuck
80f0ddfbc8 adding file tool to the CI 2021-01-13 12:27:13 +00:00
HoneyryderChuck
42297cd38d simplifying apis, testing pathnames 2021-01-13 12:27:13 +00:00
HoneyryderChuck
25d5e93248 added additional tests for the future accepted variationns of multipart parameters 2021-01-13 00:00:09 +00:00
HoneyryderChuck
e1ccae25f9 added a global store for domains not supporting 100-continue 
it's a pretty naive implementation, as it grows unbounded, and isn't
thread safe, mostly due to negligent cost of fake fetch.
 2021-01-12 23:53:18 +00:00
HoneyryderChuck
68baf484da reshaping the way 100 responses are handled, by keeping informational status (right now working only for 100, as 101 breaks h2c) 2021-01-12 17:43:14 +00:00
HoneyryderChuck
75089a334a added test for Socks5 connection establishment using an IP. 
The test only tests IPv4 due to lack of support for IPv6 in the current
docker test suite, but IPv6 is expected to work as well.
 2021-01-09 17:14:36 +00:00
HoneyryderChuck
0c646ced85 fixed expect 100 plugin delay feature 
another bug gets fixed after adding a test... the delay of sending the
request body when server doesn't support the Expect-100 continue dance
wasn't using the proper API. Adapted the handling of the state machine
to be more readable and compliant as well.
 2021-01-09 16:20:01 +00:00
HoneyryderChuck
256ec59bdd added resolver alias test 2021-01-09 01:58:03 +00:00
HoneyryderChuck
fab44b5993 verifyig error responses annd deltas using internal methods 2021-01-05 18:57:39 +00:00
HoneyryderChuck
a6176ec01f using session pool for coalescing test, thereby minimizing interference (although it would have been nice to have more entropy) 2021-01-05 10:38:28 +00:00
HoneyryderChuck
cdf6868059 moar logging 2021-01-04 19:12:25 +00:00
HoneyryderChuck
a1b27f5349 make origins comparison less flakey 2021-01-04 19:08:01 +00:00
HoneyryderChuck
2853cde6db mapping origins correctly now 2021-01-04 19:01:03 +00:00
HoneyryderChuck
8a570f7d4c fixing location of origins var 2021-01-04 18:53:54 +00:00
HoneyryderChuck
1537ffb10f improving error message of coalescing test 2021-01-04 18:47:49 +00:00
HoneyryderChuck
d460a9c839 another try at fixing the RUBYOPT issue 2021-01-03 18:22:25 +00:00
HoneyryderChuck
5083186654 do not overwrite RUBYOPT (because of jit values) 2021-01-03 18:12:35 +00:00
HoneyryderChuck
cacd740843 giving the native resolver timeout test a bit more delta, as it fails often with concurrency set 2021-01-03 15:39:13 +00:00
HoneyryderChuck
5b0cbe4ec7 fail when any command in build.sh fails 2021-01-02 03:39:11 +00:00
HoneyryderChuck
4c5b59d2c3 integrating the integration tests in CI, disable parallel mode for them 
as one can't guarantee that the third-party modules we integrate with
are safe to run concurrently, we're left best running them sequentially.
For our first integration, webmock, this holds true.
 2021-01-01 16:56:36 +00:00
HoneyryderChuck
1313478f6c added test for dns timeouts, which uncovered a bug. 
There were errors inn the code handling the retries. This has been dealt
with, along with a new exception (HTTPX::ResolveTimeoutError)
 2020-12-24 12:38:14 +00:00
HoneyryderChuck
623ad7d50a enabling auto-compaction in GC 2020-12-18 18:16:01 +00:00
HoneyryderChuck
65f5e86f3f allowing nested params also when posting multipart requests 2020-12-10 13:46:57 +00:00
HoneyryderChuck
0bb18ad500 added urlformencoder which encodes form data the way rack is used to, when nested 2020-12-10 01:07:18 +00:00
HoneyryderChuck
392dadcb78 typing the connection parsers 2020-12-07 17:46:45 +00:00
HoneyryderChuck
beb0a82baa typing selector and connection 2020-12-07 16:30:39 +00:00
HoneyryderChuck
d3a6edeecb fixed socks4a handshake while adding more socks4 tests 
the packet wasn't being well formed in certain situations. Also, raised
restrictions for IPv6 addresses, as hton is implemented in more recent
rubies for them as well.
 2020-12-02 01:30:38 +00:00
HoneyryderChuck
201591e9fe added test for the socks5 non-happy path 2020-12-02 01:30:38 +00:00
HoneyryderChuck
c67ae0582d added test for host unreachable when connecting to the nameservers 2020-12-02 01:30:38 +00:00
HoneyryderChuck
282dff83cb reworking resolver error tests, making resolver resilient to wrong record types from user 2020-12-02 01:30:38 +00:00
HoneyryderChuck
7a5eaa3795 using the same strategy when exiting early from unexpected dns errors 
there was a :resolve_error already in place for this. this extends the
same strategy for https and native resolvers, thereby fixing potential
bugs.
 2020-12-02 01:30:38 +00:00
HoneyryderChuck
c98d568e07 fixed some bugs from the https resolver while improving coverage 
* it now fails hard  if the dns server is a domain and it's not
  resolvable;
* it properly handles retries for other record types (p.ex. when A query
  fails, and AAAA follows);
 2020-12-02 01:30:38 +00:00
HoneyryderChuck
37999cc39e improved testing of body transcoder 2020-12-02 01:30:38 +00:00
HoneyryderChuck
3bf3223c60 fixing another loop error when proxy needs to be resolved, by resolving 
fails

While adding yet another test, an edge case was found where a proxy had
to be resolved, but when the resolving failed, the process wasn't
recovering from it.
 2020-11-30 12:16:36 +00:00
HoneyryderChuck
035eda1f95 fixed another loop caused by certain connection goaway frames from 
server not being processed

While introducing yet another test to catch frame processing errors, in
this case with the SETTINGS_TIMEOUT error, another loop was found. It
was caused by two reasons:

* connection was signaling it was "closing" on such an error, which is
  not really true (server already closed the stream, so no need to
  close it again); it should be marked as closed instead.
* write buffer was still full (with the handshake in this case), so the
  connection was still trying to write;
 2020-11-27 13:47:36 +00:00
HoneyryderChuck
e6c9bb4714 fix: http/1.1 recover from connection exhausted 
While adding the test, the code to recover from an exhausted HTTP/1.1
connection proved not to be reliable, as it wasn't taking inflight
requests nor update-once keep-alive max requests counters into
account.

This has been fixed by implementing our own test dummy using
webrick.
 2020-11-26 20:11:30 +00:00
HoneyryderChuck
34dc7df495 ensurinng that the push promise test sends requests sequentially 2020-11-26 17:20:13 +00:00
HoneyryderChuck
7f9e68ba3b fixed: session hangs when HTTP/2 receives a 421 
While adding a test, a bug was found, where an HTTP/2 connection
receiving a 421 response wasn't redirecting the request well to an
HTTP/1.1 one. This has been fixed by changing the way a new connection
is instantiated.
 2020-11-26 17:13:02 +00:00
HoneyryderChuck
322aa06f22 added a test for when an http/2 connection exhausts and a second connection has to be established 2020-11-26 17:12:09 +00:00
HoneyryderChuck
350a030473 sleep skews a bit under load, increasing it for reliability 2020-11-24 17:21:29 +00:00
HoneyryderChuck
cb3e57e10d fixing log_exception signature 2020-11-24 16:38:46 +00:00
HoneyryderChuck
6cc24b63f5 forcing request fo fail with exception when there's an error (since we're supporting only 1 stream request at a time, this should be straightforward) 2020-11-24 00:42:54 +00:00
HoneyryderChuck
742d877c76 added #each_line method to stream response 2020-11-24 00:42:52 +00:00
HoneyryderChuck
1b29e062a0 added streaming plugin tests 2020-11-22 22:32:38 +00:00
HoneyryderChuck
7e26c86dc1 really closing a response (won't buffer stuff anymore after being closed) 2020-11-22 15:55:44 +00:00
HoneyryderChuck
fa2fd818fe added explicit response closing in tests 2020-11-22 15:55:43 +00:00
HoneyryderChuck
86f1db5bf0 disabling IDN support for ruby < 2.3 2020-11-07 15:36:21 +00:00
HoneyryderChuck
1b26977d16 updated rubocop to 1.0.0 on possible rubies, making the necessary changes 2020-11-07 15:36:21 +00:00