Struct Job 

pub struct Job {
    pub id: JobId,
    pub command: String,
    /* private fields */
}

A background job.

Fields

id: JobId

Job ID.

command: String

Command description.

Implementations

impl Job

pub fn new( id: JobId, session_id: u64, command: String, handle: JoinHandle<ExecResult>, ) -> Self

Create a new job from a task handle.

pub fn from_channel( id: JobId, session_id: u64, command: String, rx: Receiver<ExecResult>, ) -> Self

Create a new job from a result channel.

pub fn stopped( id: JobId, session_id: u64, command: String, pid: u32, pgid: u32, ) -> Self

Create a stopped job (from Ctrl-Z on a foreground process).

pub fn output_file(&self) -> Option<&PathBuf>

Get the output file path (if available).

pub fn streams(&self) -> JobStreams

This job’s live output streams (see JobStreams).

pub fn is_done(&mut self) -> bool

Check if the job has completed.

Stopped jobs are not considered done.

pub fn status(&mut self) -> JobStatus

Get the job’s status.

pub fn status_string(&mut self) -> String

Get the job’s status as a string suitable for /v/jobs/{id}/status.

Returns:

• "running" if the job is still running
• "stopped" if the job is stopped (Ctrl-Z / SIGTSTP)
• "done:0" if the job completed successfully
• "killed:{code}" if the job was terminated by kill %N
• "failed:{code}" if the job failed with an exit code

The vocabulary must stay in step with Self::status — GH #252 was exactly this pair drifting: status() learned the stopped check and this string twin didn’t, so /v/jobs/N/status reported a Ctrl-Z’d job as "running" forever (a stopped job has no result channel, so try_poll can never produce a result).

pub fn cleanup_files(&mut self)

Remove any temp files associated with this job.

pub fn try_result(&self) -> Option<&ExecResult>

Get the result if completed, without waiting.

pub fn try_poll(&mut self) -> bool

Try to poll the result channel and update status.

This is a non-blocking check that updates self.result if the job has completed. Returns true if the job is now done.