`passthrough:""` or `passthrough:"all"` (the default) will pass through
all further arguments including unrecognised flags.
`passthrough:"partial"` will validate flags up until the `--` or the
first positional argument, then pass through all subsequent flags and
arguments.
* fix: Check if negatable duplicates another flag
Add a check for flags with the `negatable` option if the negative flag
conflicts with another tag, such as:
Flag bool `negatable:""`
NoFlag bool
The flag `--no-flag` is ambiguous in this scenario.
* feat: Make negatable flag name customisable
Allow a value on the `negatable` tag to specify a flag name to use for
negation instead of using `--no-<flag-name>` as the flag.
e.g.
Approve bool `default:"true",negatable:"deny"`
This example will allow `--deny` to set the `Approve` field to false.
* Feat: Add xand group and check for missing
* Fix: Split and combine err in TestMultiand for consistency
* Feat: Check missing required flags in xand groups
* Feat: Handle combined xor and xand
* Docs: Add info about combined xand and required use
* Docs: Fix language error in xand description
Co-authored-by: Stautis <thkrst@gmail.com>
* Feat: Rename xand to and
* Refactor: Switch from fmt.Sprintf to err.Error
* Refactor: Get requiredAndGroup map in separate function
---------
Co-authored-by: Stautis <thkrst@gmail.com>
The golangci-lint being used was quite dated.
This change upgrades to the latest version.
Adds and updates exclusions based on new failures.
Note on revive:
I've included an opt-out for unused parameters for revive
because turning `newThing(required bool)` to `newThing(_ bool)`
is a loss of useful information.
The changes to the Go files are to fix the following issues:
```
camelcase.go:16: File is not `gofmt`-ed with `-s` (gofmt)
config_test.go:50:18: directive `//nolint: gosec` is unused for linter "gosec" (nolintlint)
defaults_test.go:28:25: G601: Implicit memory aliasing in for loop. (gosec)
doc.go:5: File is not `gofmt`-ed with `-s` (gofmt)
kong.go:446:18: directive `//nolint: gosec` is unused for linter "gosec" (nolintlint)
kong_test.go:503:20: G601: Implicit memory aliasing in for loop. (gosec)
model.go:493:10: composites: reflect.ValueError struct literal uses unkeyed fields (govet)
scanner.go:112: File is not `gofmt`-ed with `-s` (gofmt)
```
And to address broken nolint directives reported as follows by
golangci-lint.
```
[.. skipped .. ]
tag.go:65:1: directive `// nolint:gocyclo` should be written without leading space as `//nolint:gocyclo` (nolintlint)
tag.go:206:51: directive `// nolint: gocyclo` should be written without leading space as `//nolint: gocyclo` (nolintlint)
util_test.go:23:43: directive `// nolint: errcheck` should be written without leading space as `//nolint: errcheck` (nolintlint)
util_test.go:51:22: directive `// nolint: errcheck` should be written without leading space as `//nolint: errcheck` (nolintlint)
```
The former allows arbitrary structs to be embedded in the root of the
CLI, with optional tags.
The latter allows an arbitrary function to be called using Kong's
binding functionality.
This is a breaking change, but the previous behaviour was broken so I'm
not concerned.
Also made most programmer errors more useful by giving type.field
context information.
Fixes#179.
This new tag tells the parser to stop processing flags after the
positional argument is encountered.
This is particularly useful for subcommands that forward their arguments
to an external program, and makes it possible to implement commands
like `kubectl exec` or `docker run`.
Fixes#80.
Factor out some repetition from `parseTag()` into a `GetSep()` method so
as to simplify `parseTag()` and formalise the semantics of a separator.
The existing behaviour in corner/error cases has been preserved, but an
appropriate error is returned (currently ignored) so a future breaking
change could be made to handle these conditions.
Tests have not been added as the test files are in a separate package to
the unit under test, making direct `Tag` construction difficult. Testing
through the exported APIs is already performed in the
`FormatPlaceHolder()` tests.
According to the README.md, a value of `"none"` for the `sep` tag (for
slices) and `mapsep` (for maps) is meant to disable the separator that
would allow multiple entries to be added from a single argument.
However, using `"none"` just set the separator to the rune `'n'`.
Fix the parsing of the tag, and also update `SplitEscaped` to not split
with a separator of `-1`.
Add a new test for the new cases.
This provides much more convenient composition when reusing structs in
different parts of the command grammar.
eg.
type Embedded struct {
Key string `help:"A key from ${where}."`
}
var cli struct {
Embedded `set:"where=somewhere"`
}
Kong supports limited variable interpolation into help strings, enum lists and
default values.
Variables are in the form:
${<name>}
Variables are set with the `Vars(map[string]string)` option. Undefined
variable references in the grammar will result in an error at construction
time.
Previously, there was a confusing mix of functionality shared between
the two wherein you would need to use the Kong type for printing errors,
etc. but it did not have access to the context in order to print
context-sensitive usage information. This has been fixed.
Additionally, there are now fuzzy correction suggestions for flags and
commands
Also added a server example which shows how Kong can be used for parsing
in interactive shells. Run with:
$ go run ./_examples/server/*.go
Then interact with:
$ ssh -p 6740 127.0.0.1
Map key/value separator is now hardcoded to `=`. This allows the
`sep` tag to continue to be used for slice separation, cascading to
support maps with slice values eg. `--set a=1,2,3`.
* Propagate errors.
* Use junit test output.
* Expand role of DecodeContext to include Scanner.
* Inject resolved flags as Path elements in the Context.
This allows all existing logic to apply seamlessly: hooks, required
flags, etc.
* Clarify that hooks can be called multiple times.
This allows help to be called even when the parse trace is invalid.
Without this, the command-line would have to be valid in order to use
help at all, which defeats the purpose.
* Removed bubbling errors in favour of panic
* Added a test for Parse with a bad build and an arg.
* Removed is() function in favour of a switch with strings
* Collect key and value in parseCSV, reducing complexity.
* Fix in global to not use parser instance on error.
Alec Thomas