1868 Commits

Author SHA1 Message Date
HoneyryderChuck
3b882f59d6 Merge branch 'issue-171' into 'master'
fix: remove mime-types support for mime type file detection

Closes #171

See merge request honeyryderchuck/httpx!180
2021-12-24 11:42:39 +00:00
HoneyryderChuck
11f6bb401f Merge branch 'issue-172' into 'master'
faraday: added support for #on_data

Closes #172

See merge request honeyryderchuck/httpx!182
2021-12-24 11:42:26 +00:00
HoneyryderChuck
dea70478ea faraday: added support for #on_data 2021-12-22 16:01:48 +02:00
HoneyryderChuck
be5300ea4a fix: remove mime-types support for mime type file detection
mime-types uses filenames, which is a terrible and inaccurate strategy
to infer mime types (example: "a.mp4" can be "application/mp4" or
"audio/mp4" before it's "video/mp4").

Added support for `marcel` and `filemagic`, which both support magic
bytes detection strategy.

Fixes #171
2021-12-18 00:47:19 +00:00
HoneyryderChuck
17b3b6c1d1 bumped version to 0.18.3 v0.18.3 2021-12-14 20:19:20 +00:00
HoneyryderChuck
9adf8fbef9 Merge branch 'fix-webmock-multipart-body' into 'master'
fix: request bodies eager-loaded from enumerables yield duped partial strings

See merge request honeyryderchuck/httpx!178
2021-12-14 20:02:36 +00:00
HoneyryderChuck
c8d32b7363 fix: request bodies eager-loaded from enumerables yield duped partial
strings

An error was observed while looking at webmock integration, where
requests formed via the multipart plugin where returning an empty string
as body. The issue was caused by an optimization on multipart encoder,
which reuses the same buffer when reading chunks. Unfortunately, these
cannot be yielded the same way via IO.copy_stream, as the same (cleared)
buffer will be used to generate the eager-loaded body chunks.
2021-12-14 19:12:58 +00:00
HoneyryderChuck
5187bc95b5 added missing images to blog post 2021-12-12 12:29:08 +00:00
HoneyryderChuck
facffe7644 bumped version to 0.18.2 v0.18.2 2021-12-09 17:53:41 +00:00
HoneyryderChuck
938d69aed7 Merge branch 'gh-issue-3' into 'master'
do not loop forever if there are no connections to listen nor any timer

See merge request honeyryderchuck/httpx!177
2021-12-09 17:44:08 +00:00
HoneyryderChuck
edc2c4e6c4 do not loop forever if there are no connections to listen nor any timer
to wait for.

Timeout calculation may trigger errors which lead to connection
unregistering, such as tital timeout errors. In such a case, we can end
up in a state where #select gets called with no timeout and no
selectable connections.

https://github.com/HoneyryderChuck/httpx/issues/3
2021-12-09 15:02:32 +00:00
HoneyryderChuck
8e77f0464a bumping version to 0.18.1 v0.18.1 2021-11-23 10:32:12 +00:00
HoneyryderChuck
ab70f6a586 total timeout call should bubble up exception, not timeout value 2021-11-19 23:03:08 +00:00
HoneyryderChuck
589e47e98b Merge branch 'issue-160' into 'master'
do not ignore goaway with no errors sent by the server

Closes #160

See merge request honeyryderchuck/httpx!176
2021-11-19 22:45:01 +00:00
HoneyryderChuck
f3cff86775 cleaning up state within an http2 connection receiving a goaway, to properly mark it to be close; this will prevent unfinished pending requests to fall in the same connection, resulting in an unrecoverable internal state error 2021-11-19 16:24:50 +00:00
HoneyryderChuck
0f80baf351 retries plugin: make errorless http2 connnection goaway a retryable error 2021-11-18 18:12:32 +00:00
HoneyryderChuck
5a61586fd5 do not ignore goaway with no errors sent by the server, let the connection go down
raise a specific error for this, which will make it easier to rescue
from.
2021-11-18 18:07:12 +00:00
HoneyryderChuck
c7d73d09b8 coalescing: do not try to use a self-signed certificate for coalescing verification purposes 2021-11-17 21:35:31 +00:00
HoneyryderChuck
f31b06c150 improving servlet layer to be able to quickly build an HTTP2 test server 2021-11-17 21:34:16 +00:00
HoneyryderChuck
d25b61fa00 set mfa required in gemspec 2021-11-16 15:07:08 +00:00
HoneyryderChuck
7e73c3a066 Merge branch 'issue-159' into 'master'
Fix for HTTP/1 logging on pipelining

Closes #159

See merge request honeyryderchuck/httpx!175
2021-11-16 14:57:06 +00:00
HoneyryderChuck
17d05e987d fixed cond around waiting for localstack to come up again 2021-11-14 02:56:32 +00:00
HoneyryderChuck
72924e420b fixing HTTP/1 logging on pipelining, where the buffered requests would appear in the second request log 2021-11-14 02:56:32 +00:00
HoneyryderChuck
b18b715818 some small rbs improvemments 2021-11-13 11:09:29 +00:00
HoneyryderChuck
8a8e0998df bumping version to 0.18.0 v0.18.0 2021-10-02 23:53:07 +01:00
HoneyryderChuck
34405df522 skipping vary test 2021-10-02 01:21:40 +01:00
HoneyryderChuck
11a02946ad Merge branch 'issue-122' into 'master'
Response Cache plugin

Closes #122

See merge request honeyryderchuck/httpx!171
2021-10-01 23:52:10 +00:00
HoneyryderChuck
b2a5371fe2 fix missing var 2021-10-02 00:27:37 +01:00
HoneyryderChuck
8f00539d6d bigger delta overall for timer-based tests 2021-10-02 00:19:53 +01:00
HoneyryderChuck
a3ee98f410 implementation of the response cache plugin 2021-10-01 23:53:21 +01:00
HoneyryderChuck
911c803d83 added test for vary header handling 2021-10-01 23:38:38 +01:00
HoneyryderChuck
a60fde07d6 added plugin tests, and the plugin body 2021-10-01 23:38:38 +01:00
HoneyryderChuck
5bafed433a Merge branch 'issue-157' into 'master'
Fixing AWS apis

Closes #157

See merge request honeyryderchuck/httpx!173
2021-10-01 22:38:04 +00:00
HoneyryderChuck
bdcd4b31b5 aws_sdk plugin: removing S3 plugin APIs, replacing it with barebones aws-sdk-core components 2021-10-01 23:20:19 +01:00
HoneyryderChuck
8f5c45225b removing aws-sdk-s3 constraint in sigv4 plugin (does not require anything aws-sdk related) 2021-10-01 23:20:19 +01:00
HoneyryderChuck
a076da6549 disable h2 upgrade test for 2.3 2021-10-01 22:29:41 +01:00
HoneyryderChuck
c2131c8e41 disable ssl verify for nghttp2, letsencrypt expired cert + old openssl, thx 2021-10-01 18:00:34 +01:00
HoneyryderChuck
f4dea3cc10 Merge branch 'fix-timers' into 'master'
Fix timers and remaining busy loops on multi-requests

See merge request honeyryderchuck/httpx!174
2021-10-01 14:58:13 +00:00
HoneyryderChuck
01efc4405c usinng better base images 2021-09-30 02:27:10 +01:00
HoneyryderChuck
03d57a918f bump to latest jruby 2021-09-27 16:18:02 +01:00
HoneyryderChuck
1e4177a079 allow connections to be closed before they even started, if some error coming from the parser requires it (can happen if user input is evaluated in httpx time and fails) 2021-09-27 16:18:02 +01:00
HoneyryderChuck
eb5bdc1fe6 telemetry: fixed visibility of #close; more log data 2021-09-27 16:18:02 +01:00
HoneyryderChuck
207ecb2da0 bugfix: making https resolver subscribe to the right pool
(such as when it's overwritten, such as in the case of SessionWithPool Module)
2021-09-27 16:18:02 +01:00
HoneyryderChuck
efddd72caa removing persistent connections from the selector whe inactive
keeping them around was resulting in some busy loops on timer events
(i.e. retry after), making them unreliable, innacurate  and CPU
draining. they're now kept out whenever they're inactive.
2021-09-27 16:18:02 +01:00
HoneyryderChuck
edf7357a5f improved persistent plugin test, in order to test not only the number of connections kept, but the ones actually being used for selecting (connections with no request shouldn't) 2021-09-25 00:44:31 +01:00
HoneyryderChuck
52948e0f83 bugfix: prevent stream close callback from being called 2 times
an issue was observed when stream was closed from our side, that the
the request in-flight count on the connection. This was fixed by not
reacting to :stream_closed events if request has been previously deleted.
2021-09-23 12:12:48 +01:00
HoneyryderChuck
81a41d889c bugfix: remove connections from selector which have been unregistered
during interest calculation

A quirk was found whereby a connection which failed while connecting
(such as the badssl test) was properly unregistered from the pool, was
however kept in the selectables selector pool, because if this operation
happening during the interest calculation pool, and the var substitution
being performed right afterwards, leaving the pool and selector out of
sync and causing all sorts of miscalculations around timers later on.
2021-09-22 12:53:33 +01:00
HoneyryderChuck
6425e5d8ee moved calculation of pool timeout to #next_timeout fully 2021-09-22 12:48:40 +01:00
HoneyryderChuck
c1881dc65b preparing for URI::HTTP#origin in mainline 2021-09-22 12:48:40 +01:00
HoneyryderChuck
bd1946999f improving retry logs 2021-09-22 12:48:40 +01:00