Struct Runner 

pub struct Runner { /* private fields */ }

Port of clustertool.Runner. The zero value is not usable; use Runner::new.

Implementations

impl Runner

pub fn new() -> Self

Port of clustertool.NewRunner: wired to the process’s own standard streams and terminal.

pub fn new_captured( stdin_content: &[u8], has_terminal: impl Fn() -> bool + Send + Sync + 'static, ) -> Self

Test-only constructor: stdin_content is fed to the child’s stdin, and its stdout/stderr are captured rather than inherited, readable afterwards through Runner::stdout/Runner::stderr. has_terminal stands in for has_controlling_terminal, so tests do not depend on whether the suite happens to run from a terminal.

pub fn stdout(&self) -> Vec<u8>

Returns everything written to stdout so far. Only meaningful for a Runner::new_captured instance; always empty otherwise.

pub fn stderr(&self) -> Vec<u8>

Returns everything written to stderr so far. Only meaningful for a Runner::new_captured instance; always empty otherwise.

pub fn run( &self, mgr: &dyn Driver, profile_name: &str, tool: &Tool, args: &[String], ) -> Result<i32, ClusterToolError>

Port of (*Runner).Run. Resolves the cluster kubeconfig from mgr, writes it to a temporary file, and runs tool with args attached to the caller’s streams. Returns the tool’s exit status, which is only meaningful when the returned error is Ok; a tool that could not be started at all yields Err.

See the module doc comment for why this takes mgr and profile_name directly rather than loading them from an ambient config context, as the Go original does.

Trait Implementations

impl Default for Runner

fn default() -> Self

Returns the “default value” for a type.