| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| ... | |
| | | |
|
| | |
| |
| |
| |
| |
| |
| | |
Some imports are fixtures, that need to be in the current namespace when
pytest runs. However, pylint does not know this.
Disabling pylint checks on those imports
|
| |/
|
|
|
|
|
|
|
|
| |
This is due to pytest fixtures having to be named the same as the test
arguments.
This is a pre-requisite to enable pylint on this directory
We need to do this per file as we can't blanket disable for directories.
See upstream issue: https://github.com/PyCQA/pylint/issues/618
|
| |\
| |
| |
| |
| | |
Remove tests/testutils/site.py and move everything to buildstream/testing/_utils/site.py
See merge request BuildStream/buildstream!1375
|
| | |
| |
| |
| |
| |
| | |
What this checks does is check if the ostree plugin is installed and
fully configured. Renaming to HAVE_OSTREE_PLUGIN makes it slightly more
explicit
|
| | |
| |
| |
| |
| |
| | |
The only place needing to know about OSTREE is the flatpak-autotools for
the documentation. We can therefore move the OSTREE business here and
remove completely the site file, ending up with a single one left.
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| | |
We currently have two sites file containing redundant information.
- tests/testutils/site.py: Remove IS_{LINUX,WSL,WINDOWS}
- rest: Change imports to import those from
buildstream/tests/_utils/site.py
|
| | |
| |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both
define some variables in BuildStream environment.
Remove HAVE_SANDBOX from tests/testutils/site.py and change all imports
to point to the other one
|
| | |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both
define some variables in BuildStream environment.
Moving the MACHINE_ARCH to a single place.
|
| | |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both
define some variables in BuildStream environment.
Moving all the arpy related ones in a single place.
|
| | |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both
define some variables in BuildStream environment.
Moving all the lzip related ones in a single place.
|
| | |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both define
some variables in BuildStream environment.
Moving all the bubblewrap related ones in a single place.
|
| | |
| |
| |
| |
| |
| |
| | |
We have two different 'site' files that are redundant and both define
some variables in BuildStream environment.
Moving all the git related ones in a single place.
|
| |/
|
|
|
|
|
| |
We have two different 'site' files that are redundant and both define
some variables in BuildStream environment.
This is a first step in consolidating them in a single file
|
| |\
| |
| |
| |
| | |
Rename (spawn, fork) -> 'start process'
See merge request BuildStream/buildstream!1374
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
| |
Avoid confusion by not referring to starting another process as
'spawning'. Note that 'spawn' is a process creation method, which is an
alternative to forking.
Say 'create child process' instead of 'fork' where it doesn't harm
understanding. Although we currently only use the 'fork' method for
creating subprocesses, there are reasons for us to support 'spawn' in
the future.
More information on forking and spawning:
https://docs.python.org/3/library/multiprocessing.html#contexts-and-start-methods
|
| |\
| |
| |
| |
| | |
Don't assert gRPC messages which may change
See merge request BuildStream/buildstream!1381
|
| |/
|
|
|
|
|
|
| |
When we are testing push failures, we are being too strict
when we try to verify that the remote has failed to be initialized.
The assertion was failing because of slightly different gRPC error
messages. Now the assertion is less strict but still appropriate.
|
| |\
| |
| |
| |
| | |
_scheduler/jobs: refactor, defensive send_message
See merge request BuildStream/buildstream!1373
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Simplify the custom 'handle_message' / 'send_message' protocol by not
requiring a message_type. These message types share a namespace with the
base Job implementation, which could cause trouble.
Introduce a new private '_send_message' to implement the old
functionality.
Subclasses are free to pack a message type into their own messages, this
isn't necessary at present and simplifies existing subclass code.
|
| |/ |
|
| |\
| |
| |
| |
| | |
Ensure pylint runs in tests/integration
See merge request BuildStream/buildstream!1376
|
| | |
| |
| |
| |
| |
| |
| |
| | |
tests/integration was lacking a __init__.py file, meaning it was not
recognized as a python package, and thus, pylint would not check
anything inside the directory.
Adding __init__.py ensures we have correct checks here.
|
| | |
| |
| |
| |
| |
| | |
Python method default arguments should never be mutable.
Using 'None' as a canary and setting them afterwards instead.
|
| | |
| |
| |
| | |
Pylint was complaining about it.
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| | |
Some imports are fixtures, that need to be in the current namespace when
pytest runs. However, pylint does not know this.
Disabling pylint checks on those imports
|
| |/
|
|
|
|
|
|
|
|
| |
This is due to pytest fixtures having to be named the same as the test
arguments.
This is a pre-requisite to enable pylint on this directory
We need to do this per file as we can't blanket disable for directories.
See upstream issue: https://github.com/PyCQA/pylint/issues/618
|
| |\
| |
| |
| |
| | |
Ensure pylint runs in some tests paths
See merge request BuildStream/buildstream!1378
|
| | |
| |
| |
| |
| | |
tests/plugins was missing an __init__.py, which meant pylint
was never run there.
|
| | |
| |
| |
| |
| | |
tests/remoteexecution missing an __init__.py, which meant pylint
was never run there.
|
| |/
|
|
|
| |
tests/sandboxes was missing an __init__.py, which meant pylint
was never run there.
|
| |\
| |
| |
| |
| | |
_loader/loader: cythonize valid_chars_name
See merge request BuildStream/buildstream!1371
|
| |/
|
|
|
|
|
|
|
| |
- Create a new _loader/utils.pyx cython module for functions cythonized
in the loader module.
- Move valid_chars_name from loader to utils and cythonize.
This function is called extensively, and easy to extract
|
| |\
| |
| |
| |
| |
| |
| | |
Improved handling of legacy remotes
Closes #1025
See merge request BuildStream/buildstream!1366
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
This creates a new ArtifactRemote class, derived from CASRemote that
extends initialisation to check for an artifact service. This drops the
remote early rather than raising an error on method not found each time
it tries to use it.
Fixes #1025
|
| | |
| |
| |
| |
| |
| | |
And update relavant grpc files.
Part of #1025
|
| |/ |
|
| |\
| |
| |
| |
| | |
Split ChildJob out from Job class
See merge request BuildStream/buildstream!1334
|
| | | |
|
| | | |
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Make it clearer what happens in which process by splitting out a
'ChildJob', which encapsulates the work that happens in the child
process.
This also makes it possible to control what is transferred to the
child process. This is very useful for adding support for the 'spawn'
method of creating child processes as opposed to the 'fork' method.
|
| | |
| |
| |
| | |
This just does the default behaviour, clearer to remove it.
|
| |/ |
|
| |\
| |
| |
| |
| | |
Optimize _yaml.node_get
See merge request BuildStream/buildstream!1370
|
| | |
| |
| |
| |
| | |
This reduces considerably the number of nodes created and thus speeds
up the calls to node_get.
|
| |/
|
|
|
|
|
|
|
|
|
| |
Calls to `isinstance` can be particularily costly. Using type() is
much faster. The only known case where the `isinstance` was useful
is for dictionnaries where we would ask for a 'Mapping' instead.
Disallowing 'Mapping' for expected_type considerably speeds up the
calls to this functions.
Also add into NEWS
|
