benoitc
|
1716877232
|
do it quietly
|
2010-03-07 00:01:33 +01:00 |
|
benoitc
|
6ac5faf763
|
fix chunked response
|
2010-03-06 23:40:59 +01:00 |
|
benoitc
|
19bf5e5d1b
|
just reinit
|
2010-03-06 23:06:18 +01:00 |
|
benoitc
|
9651a809e3
|
no it's now part from gunicorn release too.
|
2010-03-06 20:58:09 +01:00 |
|
benoitc
|
c785be0780
|
new parser using StringIO, faster than concatenate strings. Lot of fixes
in TeeInput.
|
2010-03-06 20:47:39 +01:00 |
|
benoitc
|
54d1a8a5dc
|
fix iter
|
2010-03-05 18:20:58 +01:00 |
|
benoitc
|
83804e2421
|
follow spec
|
2010-03-05 01:11:00 +01:00 |
|
benoitc
|
c67800e1c7
|
I don't see any good reason to keep Status here. Also since we are
HTTP/1.1 make sure to use correct status line
|
2010-03-04 19:46:38 +01:00 |
|
benoitc
|
d26073cc78
|
worker may have already be removed here.
|
2010-03-04 19:43:17 +01:00 |
|
benoitc
|
4a56af4a2b
|
typo.
|
2010-03-04 19:29:22 +01:00 |
|
benoitc
|
8622604f44
|
remove race condition
|
2010-03-04 19:23:37 +01:00 |
|
benoitc
|
6ca87c5cf4
|
right length
|
2010-03-04 18:58:27 +01:00 |
|
benoitc
|
77ea9ea78a
|
recv_info seems slower that just concatening string. let's create a
branch to test it.
|
2010-03-04 18:48:39 +01:00 |
|
benoitc
|
a0d8a16b86
|
improve read_partial
|
2010-03-04 18:07:45 +01:00 |
|
benoitc
|
176ad162b1
|
fix header parsing. spotted by MattCampbell on irc. thx
|
2010-03-03 16:23:32 +01:00 |
|
benoitc
|
7f0b689851
|
fix 100-Continue. spotted by MattCampbell on irc.
|
2010-03-03 16:13:41 +01:00 |
|
benoitc
|
68d1542c20
|
timeout and backlog can now be configured in config file. we should put
more options. Also the default of backlog could be 1024 instead of 2048.
|
2010-03-03 16:03:23 +01:00 |
|
jbergstroem
|
e5e6b70895
|
Fix failed tests due to lack of passing config to a Request
|
2010-03-03 15:44:54 +08:00 |
|
benoitc
|
0774607fab
|
flush on first write
|
2010-03-02 23:49:41 +01:00 |
|
benoitc
|
6fc3772c8c
|
flacky my friend. spotted by jbergstroem
|
2010-03-01 18:22:43 +01:00 |
|
benoitc
|
81ea3c078e
|
don't normalize headers. fix issue #27
|
2010-03-01 18:04:34 +01:00 |
|
benoitc
|
d33c5b6c86
|
fix issue #25.
|
2010-03-01 16:28:37 +01:00 |
|
benoitc
|
368257a733
|
fix issue #24.
|
2010-03-01 16:12:09 +01:00 |
|
benoitc
|
6170d5a503
|
prepare 0.6.2. release
|
2010-03-01 10:12:26 +01:00 |
|
benoitc
|
5f11fd024b
|
remove the worker id. just use pid as ref.
|
2010-03-01 09:11:39 +01:00 |
|
benoitc
|
6deeeecc80
|
clean the code
|
2010-03-01 08:57:46 +01:00 |
|
benoitc
|
49c02c8c4c
|
fix chunked encoding. It's ok now but we need unitests
|
2010-03-01 01:03:24 +01:00 |
|
benoitc
|
19a4a081a2
|
we want to go to last position.
|
2010-03-01 00:25:25 +01:00 |
|
benoitc
|
0629f011e8
|
r should be here anyway...
|
2010-02-28 22:55:37 +01:00 |
|
benoitc
|
ea02c5e073
|
this seems to fix problem with upload. Tested with the django app
|
2010-02-28 22:46:35 +01:00 |
|
benoitc
|
d50420848e
|
add simple upload form test by adding its management in django ex
|
2010-02-28 22:46:35 +01:00 |
|
benoitc
|
c2f12a4977
|
should fix _ensure_length in tee. while I'm here fix chunked response.
|
2010-02-28 22:46:35 +01:00 |
|
benoitc
|
29fd1cf48f
|
empty line
|
2010-02-28 22:46:35 +01:00 |
|
benoitc
|
d92364af71
|
new read_partial function. We also now use buffer in parser.
|
2010-02-28 22:46:35 +01:00 |
|
benoitc
|
05d4673972
|
just make sure we write at the end everywhere
|
2010-02-28 14:59:54 +01:00 |
|
benoitc
|
d1011ff301
|
fix indent in parser.body_eof
|
2010-02-28 00:59:03 +01:00 |
|
benoitc
|
86736e448d
|
help report, bump version in trunk.
|
2010-02-27 23:37:54 +01:00 |
|
benoitc
|
7cc365305d
|
revert this patch since it cause more problem than it solves.
|
2010-02-27 23:02:15 +01:00 |
|
benoitc
|
52978cb16f
|
we want to flush here too.
|
2010-02-27 18:23:53 +01:00 |
|
benoitc
|
2348fc238d
|
seek after we sync
|
2010-02-27 18:04:04 +01:00 |
|
benoitc
|
87d3c4b3e7
|
make sure we write at the end. fix 002-test-request.test_011 . spotted
by jbergstroem on irc. thanks!
|
2010-02-27 18:01:05 +01:00 |
|
benoitc
|
49d340ce2b
|
fix import
|
2010-02-27 17:32:05 +01:00 |
|
benoitc
|
cb706ea818
|
give me some coffee please
|
2010-02-27 15:55:36 +01:00 |
|
benoitc
|
ccedd37235
|
s/sock/util spotted by jbergstroem on irc
|
2010-02-27 15:51:20 +01:00 |
|
benoitc
|
2b556bb830
|
make it more compliant with the spec
|
2010-02-27 15:21:36 +01:00 |
|
Paul J. Davis
|
50e8355c10
|
Swap slashes for hyphen.
Thanks for the suggestion by asenchi.
|
2010-02-27 01:28:51 -05:00 |
|
Paul J. Davis
|
5580b813e9
|
Reformat some docstrings.
|
2010-02-27 01:27:35 -05:00 |
|
Paul J. Davis
|
8b38298114
|
Update the log formatting.
Different date format to remove sub-second timing.
Added process id's to all log messages.
Updated various log messages.
|
2010-02-27 01:16:57 -05:00 |
|
Paul J. Davis
|
1e1207418f
|
Fixes Issue #9 - Recycle oldest workers first.
The worker that's been alive the longest will now be the first
worker to be recycled when the master process receives a TTOU
signal.
|
2010-02-27 00:46:31 -05:00 |
|
Paul J. Davis
|
fd059eaf8f
|
Fixes Issue #22 - Make EPIPE stacks disappear.
Still prints a line like:
2010-02-27 00:25:21,305 - WARNING Ignoring EPIPE
|
2010-02-27 00:25:59 -05:00 |
|