Struct Selection 

pub struct Selection {

    pub include: Vec<Pattern>,
    pub exclude: Vec<Pattern>,
    pub min_size: Option<u64>,
    pub kinds: Vec<EntryKind>,
    pub modified: ModifiedWindow,
    pub ignored: IgnoredEntries,
    pub depth: Option<Bound>,
    pub min_share: Option<ShareThreshold>,
    pub breadth: Option<Bound>,
    pub limit: Option<Bound>,
    pub sort: Option<SortKey>,
    pub reverse: bool,
    pub size: SizeMetric,
}

Which retained entries a query considers, and how its results are shaped.

Fields

include: Vec<Pattern>

Patterns an entry must match at least one of, when non-empty.

exclude: Vec<Pattern>

Patterns that exclude an entry outright; exclusion wins over inclusion.

min_size: Option<u64>

Smallest size, in the selected metric, an entry may have.

kinds: Vec<EntryKind>

Entry kinds to consider; empty means every kind.

modified: ModifiedWindow

Modification-time window.

ignored: IgnoredEntries

Entries to consider by .gitignore classification.

Anything but IgnoredEntries::Include needs an index that observed control state; crate::query::Request::validate refuses it otherwise.

depth: Option<Bound>

How deep a rendered tree descends, or None to let each view apply its own.

Optional for the same reason limit and sort are. The depth that suits a tree is not the depth that suits a flat enumeration, and while this was a plain Bound the only default the library could offer was “unbounded” – so the CLI declared default_value = "2" itself and every other caller silently got a different report for the same request.

min_share: Option<ShareThreshold>

Minimum displayed share of the selected root, as an exact percentage.

breadth: Option<Bound>

Maximum immediate child rows displayed per directory.

limit: Option<Bound>

How many entries a view reports. Rows to keep, or None to let each view apply its own bound.

Optional for the same reason sort is: a bound that suits a per-directory tree is not the bound that suits a complete enumeration, and a single shared default produced “the ten alphabetically-first entries” of a 192,871-entry tree.

sort: Option<SortKey>

Ordering key, or None to let each view apply its own default.

Optional rather than defaulted here because the sensible default differs by view: a tree and a type breakdown rank by size, while a flat file listing reads in name order. One shared default would be wrong for one of them.

reverse: bool

Whether the ordering is reversed.

size: SizeMetric

Which size metric the report answers in.

Implementations

impl Selection

pub fn is_unfiltered(&self) -> bool

Whether this selection constrains which entries are considered.

An unconstrained selection lets a view read pre-computed roll-up state directly instead of traversing entries, which is the difference between the two performance tiers a report can run in.

pub fn admits(&self, candidate: &Candidate<'_>) -> bool

Whether an entry passes every filter.

pub fn size_of(&self, candidate: &Candidate<'_>) -> u64

The size of an entry in the selected metric.

Trait Implementations

impl Clone for Selection

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 Selection

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

Formats the value using the given formatter.

impl Default for Selection

fn default() -> Self

Returns the “default value” for a type.

impl From<Selection> for EntrySelection

fn from(query: Selection) -> Self

Converts to this type from the input type.