Reid Draper
4642f2cf45
Only conditionally produce HTTP 100 Continue
...
Only produce HTTP 100 Continue if the application asks for the request
body. This allows the application to respond with a 4xx or 5xx status,
having only looked at the request headers, and never receiving the body.
An example use-case for this would be responding with HTTP 403 or 413,
without having to waste resources sending a large request body.
2015-04-09 13:16:47 +09:00
Kazu Yamamoto
9d567170e0
updating doc.
2015-04-03 13:39:58 +09:00
Michael Snoyman
c7acb86d5c
version bumps
2015-04-03 06:53:06 +03:00
Kazu Yamamoto
ed5afaec49
renaming isGet to isHead. ( #357 )
2015-04-02 22:17:37 +09:00
Kazu Yamamoto
699b9f823b
ensuring that GET and HEAD returns the same headers. ( #354 )
2015-04-02 10:34:49 +09:00
Michael Snoyman
27f1df0dcc
Version bump
2015-03-28 21:10:46 +03:00
RyanGlScott
0c4887edc8
Fix Windows build with time-1.5
2015-03-27 15:55:56 -07:00
Michael Snoyman
a8eb726ba5
Add missing test dep
2015-03-20 11:29:37 +02:00
Michael Snoyman
1db5d2eb75
pauseTimeout function
2015-03-20 11:27:30 +02:00
Michael Snoyman
8655714bbb
Disable timeouts as soon as request body is fully consumed.
...
This addresses the common case of a non-chunked request body.
Previously, we would wait until a zero-length `ByteString` is returned,
but that is suboptimal for some cases. For more information, see issue #351 .
2015-03-20 11:10:23 +02:00
Michael Snoyman
3dd01eb724
Fix a minor bug in timeout handling
2015-03-19 19:28:52 +02:00
Michael Snoyman
f006383c77
Don't serve content-length for 416 responses #346
2015-03-08 08:57:31 +02:00
Michael Snoyman
f1f658ba6a
Don't serve a 416 status code for 0-length files
...
keter issue #75 (https://github.com/snoyberg/keter/issues/75 )
2015-03-08 08:43:44 +02:00
Maximilian Tagher
9e9c7c6f6d
Revert "Turn off doctests by default to get Travis to succeed again"
...
This reverts commit 3004b5865ea37f749b28
2015-02-26 21:18:31 -08:00
Michael Snoyman
3004b5865e
Turn off doctests by default to get Travis to succeed again
2015-02-24 19:18:38 +02:00
Michael Snoyman
8cfcdf93d5
Version bump
2015-02-23 14:39:48 +02:00
Emanuel Borsboom
67adcd7635
warp: fix build when bytestring < 0.10.2.0 ( #340 )
2015-02-23 04:18:49 -08:00
Michael Snoyman
5eba374f10
Version bump
2015-02-22 22:29:39 +02:00
Michael Snoyman
6656a87bdc
Attempt to fix doctest on Travis
...
Pinging @manny-fp
2015-02-22 22:27:04 +02:00
Emanuel Borsboom
a35895aaff
warp: support blaze-builder >= 0.4
2015-02-22 07:20:13 -08:00
Michael Snoyman
0ffb30e569
Version bump
2015-02-08 21:31:54 +02:00
patrick brisbin
ac1effe251
Raise informative error on failure to read $PORT
2015-02-08 13:39:33 -05:00
patrick brisbin
ef8ceaa890
Add runEnv: like run but uses $PORT
2015-02-06 13:12:03 -05:00
Michael Snoyman
c30e0a91ab
Version bump
2015-02-04 19:30:03 +02:00
Kazu Yamamoto
02e21202cb
changing the type of runSettingsConnectionMakerSecure #302
2015-02-04 21:23:52 +09:00
Kazu Yamamoto
49928b8681
renaming recvSendLoop to http1 to get ready for http2.
2015-02-04 21:00:51 +09:00
Kazu Yamamoto
920ba5fcb8
better naming for readability.
2015-02-04 20:54:14 +09:00
Kazu Yamamoto
cb4556d220
fixing the doc for sendResponse.
2015-02-04 20:38:35 +09:00
Dylan Simon
d9cb4b798e
Improve range handling to return 416
...
Prevent parsing negative range end-points, truncate ranges to file size, and
correctly return 416 when requested range is outside file bounds.
2015-02-02 00:34:54 -05:00
Kazu Yamamoto
7701f2624c
warp: ver bumps up.
2015-01-23 15:50:25 +09:00
Kazu Yamamoto
cee1871155
Makeing recvSendLoop tail recursive ( #326 ).
2015-01-23 15:45:04 +09:00
Kazu Yamamoto
e6028bfce9
warp: ver bumps up.
2015-01-22 11:21:58 +09:00
Kazu Yamamoto
7500bc7b8e
if FilePart is specified, obtaining the file size from it ( #328 ).
...
stat(2) is not called if FilePart is specified.
This is a workaround for sophisticated applications.
2015-01-21 11:43:59 +09:00
Michael Snoyman
55c446cd54
Version bump
2015-01-14 10:37:30 +02:00
Roman S. Borschel
4b97050f96
Pass the Request to settingsOnException handlers when available.
2015-01-13 15:58:46 +01:00
Michael Snoyman
93ed763ff2
Version bump
2015-01-05 17:17:27 +02:00
Luke Taylor
3805a01a50
Fix some warp comment typos
2015-01-05 15:10:14 +00:00
Michael Snoyman
1977a1a7ce
Minor doc improvements
2014-12-30 05:56:01 +02:00
Emanuel Borsboom
4cfcd14f74
Fix test build failure
2014-12-23 05:16:53 -08:00
Emanuel Borsboom
8f3bc4aa33
warp supports the PROXY protocol
2014-12-22 13:56:40 -08:00
Michael Snoyman
50d7a20ca3
Doc link updates
2014-12-21 11:27:53 +02:00
Kazu Yamamoto
17541c9635
layout only.
2014-12-02 09:57:06 +09:00
Michael Snoyman
c09ef1d4ef
Doc fixes, add setFork, add missing pragma
2014-12-01 23:00:33 +02:00
David Barbour
6edfbccd39
warp with bound threads #308
2014-12-01 13:56:50 -06:00
Michael Snoyman
0be7202622
setMaximumBodyFlush settings
...
Pinging @kazu-yamamoto @meteficha. This should be the last commit in
this request body flushing saga.
2014-11-27 07:13:05 +02:00
Michael Snoyman
38934e7ecc
Modify request body flushing behavior.
...
See explanation in ChangeLog.md. Pinging @kazu-yamamoto : do you see any
problems with this change?
2014-11-26 13:53:44 +02:00
Michael Snoyman
5043338823
README.md and ChangeLog.md
2014-11-26 13:50:51 +02:00
Gabor Greif
16b3ed78f1
Make this compilable on GHC HEAD (v7.9)
2014-10-30 16:28:09 +01:00
Michael Snoyman
9edc2aab34
Revert "Don't have double content-length headers snoyberg/keter#52"
...
This reverts commit 76d474873e
2014-10-20 07:33:53 +03:00
Michael Snoyman
76d474873e
Don't have double content-length headers snoyberg/keter#52
...
Pinging @kazu-yamamoto . Can you review this patch? The issue is that,
currently, any responseFile respone which sets content-length ends up
with two content-length headers. I'm also going to patch wai-app-static
to not provide a content-length header, since it's incorrect in the case
of range requests.
2014-10-19 23:56:09 +03:00
