Struct EvalReport 

pub struct EvalReport {

    pub claimed: usize,
    pub skipped_not_due: usize,
    pub skipped_locked: usize,
    pub skipped_paused: usize,
    pub items: usize,
    pub duplicates: usize,
    pub runs_started: usize,
    pub ingested: usize,
    pub unidentifiable: usize,
    pub unusable: usize,
    pub errors: Vec<String>,
    pub draining: Vec<String>,
    pub cursor_held: Vec<String>,
}

What one evaluation pass did. Shaped like loop run’s report for symmetry.

Fields

claimed: usize

skipped_not_due: usize

skipped_locked: usize

skipped_paused: usize

items: usize

duplicates: usize

runs_started: usize

ingested: usize

Items recorded without starting a workflow, because no runtime is wired in. Counted apart from runs_started so the report never claims a run happened when none did.

unidentifiable: usize

Items the connector returned that carry no usable identity, so no run could be minted for them. Reported rather than silently dropped: an item nobody can name is a connector bug, and it should be visible as one.

unusable: usize

Declarations that cannot fire as written — a cron that does not parse, a timezone this build refuses, a composite gate naming a member the declaration does not carry.

Counted APART from skipped_not_due on purpose. Folded in there, an unusable trigger is indistinguishable from a healthy one waiting its turn, which is the failure mode with no symptom: the work simply never happens and every report looks green. The reason lands in errors.

errors: Vec<String>

Not fatal to the pass: one broken trigger must not stop the others.

draining: Vec<String>

A connector reported a backlog, so this trigger is due again at once rather than after its interval.

cursor_held: Vec<String>

Triggers whose cursor was deliberately left unmoved this pass because at least one item in the firing failed to start a run (#129) — the Event add failed, declared context was unavailable, or the run itself refused. The same page is offered again next tick; whatever already started comes back as a duplicates count rather than double-processing, which is what makes re-polling safe — see #128’s stable run ids, a prerequisite for this.

Trait Implementations

impl Clone for EvalReport

fn clone(&self) -> Self

Returns a duplicate of the value.

fn clone_from(&mut self, source: &Self)

Performs copy-assignment from source.

impl Debug for EvalReport

fn fmt(&self, f: &mut Formatter<'_>) -> Result

Formats the value using the given formatter.

impl Default for EvalReport

fn default() -> Self

Returns the “default value” for a type.

impl Serialize for EvalReport

fn serialize<__S>(&self, __serializer: __S) -> Result<__S::Ok, __S::Error>

where __S: Serializer,

Serialize this value into the given Serde serializer.