Struct ExecutionResult 

pub struct ExecutionResult {
    pub messages: Vec<String>,
    pub protocol_results: Vec<ActionResult>,
    pub protocol_result_actions: Vec<String>,
    pub raw_actions: Vec<Value>,
    pub failures: Vec<ActionFailure>,
}

Result of executing all actions

Fields

messages: Vec<String>

Messages to display to the user

protocol_results: Vec<ActionResult>

Protocol-specific action results

protocol_result_actions: Vec<String>

The action name that produced each entry of protocol_results, index for index.

A protocol that owes exactly one reply per request uses it to pick the reply when a batch carries several (see ExecutionResult::chosen_reply).

raw_actions: Vec<Value>

Raw action JSON (for protocols that need to manually process actions) This is used by protocols like mDNS and NFS that have special manual processing

failures: Vec<ActionFailure>

Actions in this batch that failed to execute, in submission order.

Empty on a fully successful batch. A non-empty list does not mean the batch was aborted: execution continues past a failed action (see execute_actions).

Implementations

impl ExecutionResult

pub fn new() -> Self

pub fn add_message(&mut self, message: String)

pub fn add_protocol_result(&mut self, result: ActionResult)

pub fn add_named_protocol_result( &mut self, action_name: String, result: ActionResult, )

Record a protocol result together with the name of the action that produced it.

pub fn chosen_reply(&self, preferred: &[&str]) -> Option<usize>

Which entry of protocol_results is the reply, for a protocol that owes exactly one per request: the first that writes bytes and was produced by an action in preferred, otherwise the first that writes bytes at all. None when nothing writes.

Why not simply the first: a small model often answers with a generic acknowledgement and then the answer the request needed - +OK then the STAT line, measured in the real-model eval - and the specific answer is the one the client is waiting for.

pub fn add_failure( &mut self, index: usize, action: impl Into<String>, error: impl Into<String>, )

Record an action that could not be executed.

pub fn has_failures(&self) -> bool

Whether any action in the batch failed.

pub fn failure_summary(&self) -> Option<String>

One-line summary of the failures, or None if the batch was clean.

pub fn access_log_actions(&self) -> Vec<Value>

The action array to record in the access log.

Successful actions are recorded verbatim. A failed action keeps its original JSON under action but is wrapped in an envelope whose type is FAILED: <action name> and which carries the executor’s error, so list_access_logs cannot show a failed action as though it had run.

Trait Implementations

impl Default for ExecutionResult

fn default() -> Self

Returns the “default value” for a type.