| Commit message (Collapse) | Author | Age | Files | Lines |
| |
|
|
| |
set() has been a built-in since 2.4, and we don't support older versions.
|
| |
|
|
| |
removing all references to the LGPL as a result
|
| | |
|
| | |
|
| | |
|
| |
|
|
| |
test it in cross-test-server
|
| |
|
|
|
|
|
|
|
|
|
|
| |
byte_arrays parameters kwargs when exporting a
method. These change the calling convention in the same way as
Message.get_args_list().
- dbus.proxies.ProxyMethod: allow the same kwargs to be passed to any proxy
method; this time, they change the representation of the remote method's
return value(s).
- Test that the above work
- Improve correctness of setting the NAME_FLAG_* flags
- Whitespace tweaks (remove hard tabs)
|
| | |
|
| |
|
|
|
|
|
| |
discussion with J5.
Switching Byte to be a single-character string is arguably more Pythonic, but
needlessly breaks API.
|
| | |
|
| |
|
|
|
|
|
|
|
|
| |
interop with dbus-java:
* Don't assume we have introspection data - only call methods with correctly
typed parameters
* Make InvertMapping check order-independent
* Make signal test not fail if the Triggered signal is meant to come from /Test
* Use logging rather than sys.stderr
|
| |
|
|
| |
format dbus-java does (the <> in the specification were not meant to be literal, apparently)
|
| |
|
|
|
|
|
|
|
| |
The lists of copyright holders are derived from the D-Bus CVS changelog: I've
erred on the side of adding people to the list rather than not.
For now I've assumed that J5's contributions are owned by Red Hat, Collabora
people's contributions (Rob McQueen, Rob Taylor, Ole Andre Ravnaas,
myself) are owned by Collabora and everything else is owned by the author.
|
| |
|
|
|
|
| |
Add document API_CHANGES.txt listing visible API changes.
Add more test cases, for low-level Python <-> D-Bus type mappings.
Amend existing test cases to cope with the API changes.
|
|
|
Currently only tested with Python <-> Python, and some of the Byte tests fail
until I get an opinion from the list on whether the API should be in terms
of strings-of-length-1, ints, or both.
|