When running tests in verbose mode (or other options), tests involving
Cobra may fail if the test does not explicitly set Command.args to an
empty slice; in this case, Cobra defaults to using `os.Args`, which
will contain arguments passed to the test (such as `-v` (verbose)).
Commits e5762054c9 and 1ef0913976
implemented a workaround for this when running (unit) tests for Cobra
itself, but this check is specifig to Cobra (checking for `cobra.test`),
and don't work on Windows (which will have a `.exe` extension),
This patch implements a more universal check, so that users of Cobra
as a module also benefit from this workaround.
go1.21 and up provides a `testing.Testing()` utility ([1]); as the Cobra
module still supports Go1.16 and up, an alternative implementation was
added for older versions, based on golang.org/x/mod/lazyregexp [2].
[1]: https://pkg.go.dev/testing#Testing
[2]: https://cs.opensource.google/go/x/mod/+/refs/tags/v0.19.0:internal/lazyregexp/lazyre.go;l=66-78
Signed-off-by: Sebastiaan van Stijn <github@gone.nl>
* Address golangci-lint linter deprecation warnings
1.59.0 outputs:
WARN [lintersdb] The name "gas" is deprecated. The linter has been renamed to: gosec.
WARN [lintersdb] The linter named "megacheck" is deprecated. It has been split into: gosimple, staticcheck, unused.
* Enable some more linters, address finding
* Remove fully inactivated linters
Currently golangci-lint fails with these errors:
ERRO [linters_context] golint: This linter is fully inactivated: it will not produce any reports.
ERRO [linters_context] interfacer: This linter is fully inactivated: it will not produce any reports.
ERRO [linters_context] maligned: This linter is fully inactivated: it will not produce any reports.
I could not find any docs explaining what "fully inactivated" mean, but
based this PR[1] it seems that these linters do nothing now. Removing
the linters fixes this issue without changing linting, as they did not
produce any report.
Looking in the linters docs[2] I did not find a replacement for
"interfacer" and "malinged" linters. "stylecheck" seems to be a
replacement for "golint", but we need to fix the code to enable it.
[1] https://github.com/golangci/golangci-lint/pull/4436
[2] https://golangci-lint.run/usage/linters/
* Add stylecheck linter, replacement for golint
This revealed 2 capitalized error messages.
https://golangci-lint.run/usage/linters/#stylecheck
When creating a plugin without sub commands, the help text included the
command name (kubectl-plugin) instead of the display name (kubectl plugin):
Usage:
kubectl-plugin [flags]
The issue is that the usage line for this case does not use the
command path but the raw `Use` string, and this case was not tested.
Add a test for this case and fix UsageLine() to replace the command name
with the display name.
Tested using https://github.com/nirs/kubernetes/tree/sample-cli-plugin-help
When using `CommandDisplayNameAnnotation` we want to use it instead of
the command name in `--help` message or in the default help command.
With current code we get the wrong text in the --help usage text:
Flags:
-h, --help help for kubectl-plugin
And in the long description of the default help command:
$ kubectl cobraplugin help -h
Help provides help for any command in the application.
Simply type kubectl-plugin help [path to command] for full details.
The issue was hidden since the test checked only the Usage line.
Fixed by extracting a displayName() function and use it when creating
FlagSet and when formatting the default help flag usage and the help
command long description.
Enhance the TestPlugin to check all the lines including the command
name.
* Avoid redundant string splits
There likely isn't actually more than once to split in the source
strings in these cases, but avoid doing so anyway as we're only
interested in the first.
* Avoid redundant completion output target evaluations
The target is not to be changed while outputting completions, so resolve
it only once.
* Avoid redundant active help enablement evaluations
The enablement state is not to be changed during completion output, so
evaluate it only once.
* Preallocate some slices and maps with known size
* Avoid some unnecessary looping
* Use strings.Builder to construct suggestions
In this case the executable is `kubectl-plugin`, but we run it as:
kubectl plugin
And the help text should reflect the actual usage of the command.
To create a plugin, add the cobra.CommandDisplayNameAnnotation:
rootCmd := &cobra.Command{
Use: "plugin",
Annotations: map[string]string{
cobra.CommandDisplayNameAnnotation: "kubectl plugin",
}
}
Internally this change modifies CommandPath() for the root command to
return the command display name instead of the command name. This is
used for error messages, help text generation, and completions.
CommandPath() is expected to have spaces and code using it already
handle spaces (e.g replacing with _), so hopefully this does not break
anything.
Fixes: #2017
Signed-off-by: Nir Soffer <nsoffer@redhat.com>
Currently, only one of the persistent pre-runs and post-runs is executed.
It is always the first one found in the parents chain, starting at this command.
Expected behavior is to execute all parents' persistent pre-runs and post-runs.
Dependent projects implemented various workarounds for this:
- manually building persistent hook chains (in every hook).
- applying some kind of monkey-patching on top of Cobra.
This change eliminates the necessity for such workarounds
by allowing to set a global variable EnableTraverseRunHooks.
Tickets:
- https://github.com/spf13/cobra/issues/216
- https://github.com/spf13/cobra/issues/252
Signed-off-by: Volodymyr Khoroz <volodymyr.khoroz@foundries.io>
Fixes#1816
Previously, arguments with a dash as the second character (e.g., 1-ff00:0:1)
were detected as a flag by mistake. This resulted in auto completion misbehaving
if such an argument was last in the argument list during invocation.
When the command searches args to find the arg matching a
particular subcommand name, it needs to ignore flag values,
as it is possible that the value for a flag might match
the name of the sub command.
This change improves argsMinusFirstX() to ignore flag values
when it searches for the X to exclude from the result.
Fixes#1831
By moving the check for help group existence to "ExecuteC()" we no
longer need groups to be added before AddCommand() is called. This
provides more flexibility to developers and works better with the use
of "init()" for command creation.
Signed-off-by: Marc Khouzam <marc.khouzam@gmail.com>
* Add tests for grouping commands
* Adds Additional Command section in help
Signed-off-by: Marc Khouzam <marc.khouzam@gmail.com>
Co-authored-by: Marc Khouzam <marc.khouzam@gmail.com>
Fixes#1786
The --help, -h, --version and -v flags are normally added when the
`execute()` function is called on a command. When doing completion
we don't call `execute()` so we need to add these flags explicitly to
the command being completed.
Also, we disable all further completions if the 'help' or 'version'
flags are present on the command-line.
Signed-off-by: Marc Khouzam <marc.khouzam@gmail.com>
This method is the OnInitialize counterpart. Like OnInitialize which allows
loading the configuration before each command is executed, OnFinalize allows
saving the configuration after each command has been executed.
Add a global `EnableCaseInsensitive` variable to allow
case-insensitive command names.
The variable supports commands names and aliases globally.
Resolves#1382
This fixes a bug where a child flag that shadows (has the same
name as) a parent persistent flag would not be shown in the
child command's help output and the parent flag would be shown
instead under the global flags section.
This change makes the help output consistent with the
observed behavior during execution, where the child flag is
the one that is actually used.
Since go 1.13 you can wrap errors. This make it no longer possible to
compare with `==`, instead you have to compare with `errors.Is()`.
I noticed this problem because -h was no longer working after I stared
wrapping the errors in my custom FlagErrorFunc function.
Note that this is only a problem when a custom help flag is defined.
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
This change adds two features for dealing with flags:
- requiring flags be provided as a group (or not at all)
- requiring flags be mutually exclusive of each other
By utilizing the flag annotations we can mark which flag groups
a flag is a part of and during the parsing process we track which
ones we have seen or not.
A flag may be a part of multiple groups. The list of flags and the
type of group (required together or exclusive) make it a unique group.
Signed-off-by: John Schnake <jschnake@vmware.com>
* Fix flag completion
The flag completion functions should not be stored in the root cmd.
There is no requirement that the root cmd should be the same when
`RegisterFlagCompletionFunc` was called. Storing the flags there does
not work when you add the the flags to your cmd struct before you add the
cmd to the parent/root cmd. The flags can no longer be found in the rigth
place when the completion command is called and thus the flag completion
does not work.
Also #1423 claims that this would be thread safe but we still have a map
which will fail when accessed concurrently. To truly fix this issue use a
RWMutex.
Fixes#1437Fixes#1320
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
* Fix trailing whitespaces in fish comp scripts
Signed-off-by: Paul Holzinger <pholzing@redhat.com>
* Bash completion v2
This v2 version of bash completion is based on Go completions.
It also supports descriptions like the other shells.
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
* Only consider matching completions for formatting
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
* Use bash compV2 for the default completion command
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
* Update comments that still referred to bash completion
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
Make it easier for programs to provide shell completion by creating
the 'completion' command automatically.
If a 'completion' command is already provided by the program, Cobra
will use that one instead.
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
* Fix stderr printing functions
Follow-up of #822
* Errors go to stderr as per POSIX
* use PrintErrf() instead of extra call to Sprintf()
* Error messages should always be printed to os.Stderr.
* add test case for Print* redirection
Thanks: @bukowa for the patch.
* Don't exclude 'help' from bash completions
Fixes#1000.
* Provide completion for the help command
1- Show 'help' as a possible completion
2- Provide completions for the help command itself
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
Co-authored-by: Zaven Muradyan <voithos@google.com>
When a command request to DisableFlagParsing, it should not fail due
to a missing required flag. In fact, such a check will always fail
since flags weren't parsed!
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
Since the completion directives will be used for all shells, and that
these names will be consumed by users, this is a more appropriate name.
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
This commit allows programs using Cobra to code their custom completions
in Go instead of Bash.
The new ValidArgsFunction field is added for commands, similarly to
ValidArgs. For flags, the new function
Command.RegisterFlagCompletionFunc() is added.
When either of the above functions is used, the bash completion script
will call the new hidden command '__complete', passing it all
command-line arguments. The '__complete' command will call
the function specified by Command.ValidArgsFunction or by
Command.RegisterFlagCompletionFunc to obtain completions from the
program itself.
Signed-off-by: Marc Khouzam <marc.khouzam@montreal.ca>
Issue Reference: https://github.com/spf13/cobra/issues/1056https://github.com/spf13/cobra/pull/922 introduced a new error
type that emitted when a command was not runnable. This caused
all commands w/o a run function set to error w/ that message and a status code of 1.
This change reverts the addition of that new error. Similar
functionality can be accomplished by leveraging RunE.
