| Commit message (Collapse) | Author | Age | Files | Lines |
|---|
| |
|
|
|
|
|
|
|
|
|
|
| |
Now that we are also looking for WORKSPACE_PROJECT_FILE, we should add
it to the error message. While we're there, also mention the directory
we are looking in - this could help scripters be clear about what went
wrong.
The new error message looks like this:
Error loading project: None of ['project.conf', '.bstproject.yaml']
found in '/src/temp/blah' or any of its parent directories
|
| |\
| |
| |
| |
| |
| |
| | |
Expose base class for Git source plugins
Closes #739
See merge request BuildStream/buildstream!1019
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Add a `_GitSourceBase` class to act as the base class for deriving
source plugins that work with Git. The first user of this base class is
the `git` source plugin that's in core at the moment.
The contents of this base class is almost identical to the existing `GitSource`
class. The two notable differences are:
- the private methods now have leading underscores, as some were missing
it originally
- the name of the class
Note that we are exposing a private member here as we expect it to move to a
separate package soon. See the following discussion for more details:
https://gitlab.com/BuildStream/buildstream/issues/739#note_124819869
Fixes #739.
|
| |\
| |
| |
| |
| | |
Do not call fetch() for cached sources
See merge request BuildStream/buildstream!992
|
| | | |
|
| | | |
|
| |/
|
|
|
| |
The queue shouldn't need to know about individual sources. This is in
line with _track() and _get_consistency().
|
| |\
| |
| |
| |
| | |
README.rst: Add license badge
See merge request BuildStream/buildstream!1014
|
| |/ |
|
| |\
| |
| |
| |
| |
| |
| | |
Shell --build has optional buildtrees
Closes #740
See merge request BuildStream/buildstream!986
|
| |/
|
|
| |
Fixes issue #740
|
| |\
| |
| |
| |
| |
| |
| | |
Remote-execution instance configuration support
Closes #627
See merge request BuildStream/buildstream!952
|
| | | |
|
| | |
| |
| |
| |
| |
| |
| |
| |
| | |
Adds the 'instance_name' parameter, which may be None, to most GRPC
calls in the CASCache object. ByteStream requests already have
instance_name supplied in the resource name, so do not need the
parameter.
Closes #627.
|
| | |
| |
| |
| | |
Adds instance name support for the remote execution storage service.
|
| |/
|
|
|
|
|
| |
This is used when sending execution requests, to specify which instance of the
execution server to use.
Partial fix for #627.
|
| |\
| |
| |
| |
| |
| |
| | |
Introduce new "source" command group
Closes #814
See merge request BuildStream/buildstream!1003
|
| | |
| |
| |
| |
| |
| | |
Also, change the existing entry about `source-checkout` to be
`source checkout` instead. And, while we are there, move it near the
other announcements about the source command group.
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Following the message thread
https://mail.gnome.org/archives/buildstream-list/2018-November/msg00106.html,
implement a new command group called `source`. Move existing `track`,
`fetch`, and the recently added `source-checkout` commands under this
group.
For `track` and `fetch`, this is a BREAKING change, as the old commands
have been marked as obsolete. Using them will result in an error message
that refers people to use the new versions, like `bst source fetch`
instead of old `bst fetch`. `source-checkout` will now become
`source checkout` (the dash has turned into a space), and is not a
breaking change as it was added in the current development cycle.
Note that the functionality to hide commands from help output was added
only recently in Click, so the minimum version of Click that we now
require is 7.0.
Summary of changes:
* _frontend/cli.py: Add `source` command group, mark previous versions
as obsolete and hide them from the help output.
* _frontend/complete.py: Fix completion for hidden commands.
* setup.py: Bump Click minimum version to 7.0.
* tests: Update to cope with the new command names.
Fixes #814.
|
| |\
| |
| |
| |
| | |
contributing: more clarity on testing
See merge request BuildStream/buildstream!973
|
| |/
|
|
|
|
|
|
|
|
|
|
|
|
|
| |
Update the section on adding tests, to fulfill these broad points:
- Mandate end-to-end testing, with rationale.
- Mention internal APIs as endpoints for testing.
- Warn against pitfalls of unit-testing.
It's more text than I would have liked, perhaps in later work we'll be
able to say it with less.
This change integrates feedback from the mailing list
'Guidance on Unit Tests' thread:
https://mail.gnome.org/archives/buildstream-list/2018-November/msg00045.html
|
| |\
| |
| |
| |
| |
| |
| | |
Dont batch prepare assemble by default
Closes #800
See merge request BuildStream/buildstream!1009
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Some external plugins depend derive from BuildElement and are broken
by BuildElement enabling this batching by default.
Instead, enable it in all of the individual build element plugin
implementations.
This fixes issue #800
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| |/ |
|
| |\
| |
| |
| |
| | |
element.py: fix unbounded variable in nested python exception
See merge request BuildStream/buildstream!1006
|
| |/
|
|
|
|
|
|
|
|
|
| |
As per https://docs.python.org/3.7/reference/compound_stmts.html#except
variables used in the "except" clause are erased at the end of the
block.
We had a conflict with a nested exception using the same variable as
before.
This renames the nested variable to avoid the clash
|
| |\
| |
| |
| |
| |
| |
| | |
Mock storage space checks for tests.
Closes #530
See merge request BuildStream/buildstream!702
|
| |/
|
|
|
|
|
|
|
|
|
| |
Fix #530
- Extract free space computation in a function for easier
mocking
- Mock space computation during cache quota tests
- Mock cache size during cachque quota tests
- Add two more tests when the configuration would require
to much storage space
|
| |\
| |
| |
| |
| | |
Add artifact log command
See merge request BuildStream/buildstream!920
|
| | | |
|
| | | |
|
| | | |
|
| | | |
|
| |/ |
|
| |\
| |
| |
| |
| |
| |
| | |
Retire bst source bundle command
Closes #672
See merge request BuildStream/buildstream!959
|
| | | |
|
| | |
| |
| |
| | |
This is part of the work towards #672
|
| | | |
|
| | | |
|
| |/
|
|
| |
This commit is part of the work towards #672
|
| |\
| |
| |
| |
| | |
Guess the element when running commands in a workspace
See merge request BuildStream/buildstream!954
|
| | |
| |
| |
| | |
This is a part of #222
|
| | |
| |
| |
| | |
This is a part of #222
|
| | |
| |
| |
| | |
This is a part of #222
|
| | |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| |
| | |
Known issues:
* `bst shell` works, but `bst shell COMMANDS...` doesn't, because click
has no way of separating optional args from variable-length args.
* `bst checkout` and `bst source-checkout`'s usage strings mark LOCATION
as an optional argument. Because click gets confused if there's an
optional argument before a mandatory argument, I had to mark LOCATION
as optional internally.
* `bst workspace open` makes no sense with element being optional, so
I skipped it.
* `bst workspace close` will probably need to be revisited when multiple
projects can own one workspace.
* `bst workspace reset` will happily delete the directory you're
currently in, requiring you to `cd $PWD` to see the contents of your
directory.
I could exclude the top-level directory of the workspace being
deleted, but it is entirely valid to run workspace commands from deeper
in the workspace.
This is a part of #222
|
| | | |
|
