66f1a85221
Add connection.blocked and connection.unblocked extension methods
2013-01-09 18:10:05 +00:00
08de6606b4
I'm slightly paranoid that people will use these. Putting them at the end and making their IDs not possible to express with the codec might help.
2013-01-09 12:23:52 +00:00
b13db0c6a3
Resurrect.
2013-01-07 17:41:06 +00:00
5fdcb37152
"When opening a file, it’s preferable to use open() instead of invoking this constructor directly." Also fixes it under Python 3 after 2to3 conversion, according to Alan Antonuk.
2012-10-04 15:23:24 +01:00
9d6d1c6c78
Update copyright 2012
2012-02-02 12:51:57 +00:00
bcb2cf081b
cosmetic
2011-10-10 12:53:36 +01:00
f443a12b95
add extra license info
...
This has always been part of the Debian copyright notice.
2011-10-10 11:40:56 +01:00
2312640129
Merge in default.
2011-07-22 15:52:57 +01:00
742f5b6aeb
Changed "name": "server properties" to "name": "server-properties" in both json files.
2011-03-09 11:55:23 +00:00
6c96330489
Add count as a base for credit. This means messages in flight can be accounted for -- the sender of basic.credit may not have received all the messages the queue has delivered.
2011-02-09 16:26:29 +00:00
70bac1aa88
Remove redundant fields from credit-ok.
2011-02-03 15:49:05 +00:00
f5440b22b2
Change mind again on how this should work - have a synchronous reply method, and an async method for spontaneous notification.
2011-02-03 13:38:05 +00:00
97c44175d0
In fact this should *not* be synchronous.
2011-02-03 13:03:59 +00:00
d75743e953
Mark basic.credit as synchronous, so that clients can interpret the credit-state as the reply. I *think* this is non-evil.
2011-02-03 11:44:18 +00:00
841cfb164e
Credit needs to be per ctag.
2011-02-02 12:03:58 +00:00
f7679c8e72
First pass at channel.credit.
2011-02-01 14:49:58 +00:00
2462b206bd
Opps, lost a starting year in the previous commit
2011-01-19 14:01:42 +00:00
ace7ee4c37
Correct all copyright notices
2011-01-19 14:00:43 +00:00
aa844322a9
Merge with default
2011-01-14 13:14:22 +00:00
e2a84111e1
m-x untabify
2011-01-14 13:12:22 +00:00
d099e83f75
merge bug23645 into default (C client is lacking some methods in its API)
2011-01-14 12:04:42 +00:00
60f1a32145
Don't delete the output file before closing it
...
Windows won't let you, and produces an error, disguising whatever
error led us to try to delete the output file in the first place.
2011-01-13 11:23:01 +00:00
dc7e06aa39
Removed tabs from basic.nack section
2011-01-12 09:33:33 +00:00
960c9af7f2
Added requeue flag to basic.nack
2011-01-12 09:19:50 +00:00
1b1dd38e6e
Merge default
2011-01-11 17:07:06 +00:00
6ec9ba9e32
Merge bug23656 into default
2011-01-11 15:59:00 +00:00
7de63185ac
Added basic.nack to the JSON protocol definition
2011-01-11 15:31:14 +00:00
4f13afcb88
default queue name to "" everywhere
...
since the MRDQ feature operates everywhere
2011-01-07 17:56:42 +00:00
2c8aa0a93e
Added .gitignore.
2011-01-07 13:39:16 +00:00
6372794445
Make AmqpClass.spec hold the parsed spec, not the raw spec
...
AmqpClass previously held the raw JSON spec data, rather than the
parsed AmqpSpec. But the latter is clearly better - it's more useful
in general, and contains the former.
No supported clients were using this anyway.
2011-01-07 01:19:12 +00:00
a2e0b9b68a
remove confirm.select{multiple}
2010-12-30 00:01:35 +00:00
fc48c15f9a
merge default into bug20284
2010-10-18 18:13:49 +01:00
49d8194700
merge bug21377 into default
2010-10-14 10:23:31 +01:00
a191947944
cosmetic
2010-10-10 14:38:26 +01:00
302b1e1eb7
cosmetic
2010-09-27 19:34:03 +01:00
f48768127c
merge default into bug20284
2010-09-22 16:27:43 +01:00
046ea717ca
exchange.unbind has nowait
2010-09-12 17:35:12 +01:00
a4fa7fd893
Implement exchange-to-exchange bindings
2010-09-11 23:46:30 +01:00
48303cb310
moved confirm extension into 0-9-1 json
2010-09-03 10:45:53 +01:00
b8ec5273b8
rename basic.consume's 'filter' arg to 'arguments', as per the 0-9-1 spec
2010-09-01 06:42:53 +01:00
4ea0240d72
rename pubAck -> confirm
...
> pubAck -> confirm
> confirm.select{multiple, nowait}
> drop pubAck.deselect
2010-08-11 16:05:34 +01:00
2c6c2887b0
added pubAck.deselect-ok
2010-08-10 15:54:50 +01:00
81e792505a
puback -> pubAck
2010-08-10 15:31:52 +01:00
d715381af2
added Publisher Acks class and methods
...
This introduces one new class and two new methods within said class.
> puback
> class-id: 85
>
> puback.select{many : bit}
> put the channel into one of the two publisher ack modes, namely
> "ack publisher messages as soon as they arrive" or "let them
> pile up a bit and ack many messages at once"; this cannot be
> called on a transacted channel
>
> puback.deselect{}
> handle any outstanding acks, put the channel back in the normal
> mode, then return
2010-08-10 15:12:09 +01:00
0659342d94
add unexpected_frame error code to 0-8 spec
...
we agreed we'd use the same error codes across version, but somehow
missed this one.
2010-07-30 19:46:01 +01:00
966c8fdb36
merge bug22889 into default
2010-07-30 15:18:24 +01:00
a4aa0d556b
revert gratuitous change
2010-07-09 17:08:00 +01:00
d8123867d6
Make acc win in domains_merger.
2010-07-07 16:07:45 +01:00
23e5d9ff50
Revert these two merges to new into old, I can't remember why I swapped them round and it makes things rather clearer this way.
2010-07-07 14:42:46 +01:00
0e14045eaf
Merging bug22901 into 22889.
2010-07-07 13:43:58 +01:00
5fa3fbf1a7
More renaming :)
2010-07-07 13:43:36 +01:00
487f64894f
Merged bug22901 into bug22889.
2010-07-07 13:06:50 +01:00
b6e6b01baf
Stylistic changes: Rename old to acc, allow_overwrite to allow_accumulate, and o to merged. Add comment. Tweak usage message.
2010-07-07 13:03:24 +01:00
861db4aeba
Merge bug22901
2010-07-05 17:47:30 +01:00
af25b5695c
Remove accidental print statement
2010-07-05 17:30:21 +01:00
4a5a3c0c13
...and back out all those checks after discussion with Matthias.
2010-07-05 17:18:33 +01:00
e7635893f7
Merged bug22901 into bug22889
2010-07-05 16:41:29 +01:00
93096e13ae
Refactor merge_dict_lists_by a bit.
2010-07-05 16:40:31 +01:00
07d8aa56d1
Actually, default-value conflicts all over the place.
2010-07-05 16:34:01 +01:00
fd5af7eb9e
Always break when there's
...
- a method argument defined with two different types or default values
2010-07-05 16:30:28 +01:00
3ee744c477
Always break when there's
...
- a method with two different "synchronous" settings
2010-07-05 16:15:29 +01:00
6929c42ecc
Merge bug 22901 into bug 22889.
2010-07-05 16:09:50 +01:00
0427e055ba
Always break when there's
...
- a domain defined with two different types
- a constant defined with two different values
- differences in properties signatures
2010-07-05 16:07:07 +01:00
fba4fd20d0
Restart bug 22889
2010-07-05 16:06:13 +01:00
60df6cd41a
Merge bug 22901
2010-07-05 15:19:08 +01:00
bd9fef9680
Add --allow-overwrite command line switch.
2010-07-05 15:18:46 +01:00
96a7dbea0d
Major and minor are integers, revision should be too.
2010-07-02 17:42:58 +01:00
42ad999d1d
Rename JSON files, and merge the 0-8 extensions back into our main 0-8 file.
2010-07-02 14:25:08 +01:00
8afa5424a2
Remove close08 hack, and remove deprecated- prefix everywhere so it will merge with 0-8.
2010-06-24 16:11:58 +01:00
cc35c0fdf4
Merged bug 22901
2010-06-24 15:26:46 +01:00
a10c0c6247
Merge everything, don't die on conflicts.
2010-06-24 15:26:11 +01:00
f9ecfbf546
Revert the lie that this file is 0-8
2010-06-24 13:36:14 +01:00
34cc3d2287
Re-add 0-8 JSON for the clients.
2010-06-24 12:54:50 +01:00
e6b7a73ff9
Claim to be 0-8 for the time being. This needs to be stripped out though.
2010-06-23 16:21:13 +01:00
44e7a7c972
Remove 0-8 specific JSON and use the unified one.
2010-06-23 15:41:02 +01:00
6232b61fe5
Merged bug22864 into default
2010-06-22 17:28:48 +01:00
68a62e19d8
Update class records in-place
2010-06-21 22:49:50 +12:00
5c72a777f2
xrange instead of range
2010-06-21 10:32:12 +01:00
6ddef67e2e
Preserve ordering of constants, methods and classes
2010-06-21 14:44:33 +12:00
56f6c4407f
Put all our extensions in one file.
2010-06-17 13:05:30 +01:00
98986c5421
Create an extension for synchronous basic.recover in 0-8. Also move queue.unbind in 0-8 to an extension since that's what it is.
2010-06-17 12:57:53 +01:00
7b5be26842
updated docs
2010-06-11 07:17:14 +01:00
2cc2af31ee
amqp_codegen.py now supports an arbitrary number of functions
2010-06-03 16:11:31 +01:00
a815c3dde0
rabbit_framing_spec.hrl is now generated
2010-06-03 14:51:10 +01:00
2c4850e41d
bug22555 merged into default (synchronous is now True not 'true')
2010-05-28 18:03:12 +01:00
3e1daa6637
More documentation.
2010-05-23 23:11:06 +12:00
e2b3041082
Documentation.
2010-05-23 22:56:49 +12:00
36035d82a7
Check for duplicate properties.
2010-05-23 22:55:52 +12:00
5b45e992b0
"true" => true. This requires no changes elsewhere because the python json library which we're using in codegen seems to be able to cope with "true" and "false" as boolean values.
2010-04-01 14:12:24 +01:00
e458b5bdb3
Move demo extension from rabbitmq-server to rabbitmq-codegen; start documentation
2009-10-19 13:36:42 +01:00
b496184de1
Extensions mergers.
2009-10-07 16:11:54 +01:00
7f1137b69a
Infrastructure for supporting piecewise spec extension.
2009-10-07 14:55:02 +01:00
7ca539bfff
Merged bug21493 into default
2009-08-26 14:19:07 +01:00
db97179e9b
Added synchronous flag to json for methods that have an associated response
2009-08-26 12:11:47 +01:00
6b5b8ffb82
add a few more default method arg values
...
and changed *.ack.multiple's default from true to false, which is the
more sensible (and less dangerous) default
2009-08-12 23:08:22 +01:00
77bd658abb
Avoid try: ... except: ... finally: to syppoet python 2.4
...
Python 2.4 (as still used by RHEL 5) doesn't have this syntax. You
have to nest a try: ... except: inside a try: ... finally: ...
Tested on Centos 5.
2009-05-08 16:17:07 +01:00
5a72c5ebcd
Write autogenerated code to file instead of to stdout.
...
To avoid generating incorrect code, remove the file on exception.
2009-01-09 19:05:59 +01:00
23ed51ce65
More strict matching
2008-12-12 15:01:04 +00:00
552f37bcc8
Use default regex type in find
2008-12-12 14:44:14 +00:00
1052e77bae
Added distclean target
2008-12-12 14:25:19 +00:00
Michael Klishin
Simon MacMullen
Emile Joubert
Alexandru Scvortov
Steve Powell
Michael Bridgen
Matthew Sackman
David Wragg
Rob Harrop
Matthias Radestock
=?UTF-8?q?Jakub=20=C5=A0=C5=A5astn=C3=BD=20aka=20Botanicus?=
Tony Garnock-Jones
Tony Garnock-Jones
Alexandru Scvortov
Marek Majkowski
Matthew Sackman
Ben Hood
Paul Jones
Matthias Radestock
David Wragg
Hubert Plociniczak