Struct GroupFacts 

pub struct GroupFacts {

    pub clone_type: CloneClass,
    pub scope: CloneScope,
    pub instances: u64,
    pub smallest_member_tokens: u64,
    pub largest_member_tokens: u64,
    pub min_pairwise: f64,
    pub files: u64,
    pub directories: u64,
    pub languages: u64,
    pub min_clone_tokens: u64,
    pub identifier_jaccard: Option<f64>,
    pub api_similarity: Option<f64>,
    pub has_loop: Option<bool>,
    pub has_dynamic_allocation: Option<bool>,
    pub call_count: Option<u64>,
    pub churn: Option<f64>,
    pub ownership_spread: Option<f64>,
}

What the ranking reads about one clone group.

Every field is a fact the scan established, not a judgement: the judgements are what rank derives from them. The reserved fields are inputs the requirements name that no analysis mode can supply yet; they are declared here and reported as absent rather than defaulted, so that the day a backend supplies one, nothing has to be told the difference between a missing value and a zero.

Fields

clone_type: CloneClass

How closely the members match.

scope: CloneScope

Whether the members are whole units or runs inside them.

instances: u64

Occurrences in the group.

smallest_member_tokens: u64

Token count of the smallest occurrence.

The smallest rather than the largest: a group is only as convincing as its least substantial member, since that is the one that could most easily have matched by coincidence.

largest_member_tokens: u64

Token count of the largest occurrence, which is what a reader would have to read, keep in step, or lift out.

min_pairwise: f64

Weakest pairwise similarity across the group. Exactly 1 for a group matched on identical content.

files: u64

Distinct files the occurrences sit in.

directories: u64

Distinct directories the occurrences sit in.

languages: u64

Distinct languages the occurrences are written in.

One, in every mode that exists today: content fingerprints are computed per language, so no group can span two. The input is read anyway, so that a cross-language frontend starts affecting the ranking by being implemented rather than by also being wired in here.

min_clone_tokens: u64

The run’s minimum clone length, which the sizes above are read against.

identifier_jaccard: Option<f64>

Weakest raw identifier-set agreement against the canonical member.

Structural whole-unit analysis supplies this before normalization loses the spelling. Other modes leave it absent rather than treating an unavailable measurement as disagreement.

api_similarity: Option<f64>

Weakest call-surface agreement the analysis measured.

None means neither side offered a call surface to compare, not that their call surfaces agree.

has_loop: Option<bool>

Whether every member has a loop, when Structural mode measured bodies.

has_dynamic_allocation: Option<bool>

Whether every member calls a recognised allocation API.

call_count: Option<u64>

Fewest call sites in any member, when Structural mode measured bodies.

churn: Option<f64>

How often the duplicated code changed. Reserved: this needs repository history, which no analysis mode reads yet.

ownership_spread: Option<f64>

How many people own the copies. Reserved, on the same footing as Self::churn.

Trait Implementations

impl Clone for GroupFacts

fn clone(&self) -> GroupFacts

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Copy for GroupFacts

impl Debug for GroupFacts

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

Formats the value using the given formatter.

impl PartialEq for GroupFacts

fn eq(&self, other: &GroupFacts) -> bool

Equality operator ==.

fn ne(&self, other: &Rhs) -> bool

Inequality operator !=.

impl StructuralPartialEq for GroupFacts