Struct FragmentDef 

pub struct FragmentDef {
    pub name: String,
    pub file: String,
    pub span: Option<Span>,
    pub source: Arc<str>,
    pub placeholders: Vec<String>,
    pub supplied_variables: Vec<String>,
}

A fragment definition — the go-to-definition target for a ref:, and the vocabulary a ref: line completes against.

Fields

name: String

Fragment name (globally unique across the scanned files).

file: String

Source name of the file declaring it.

span: Option<Span>

Byte span of its # @proef annotation line, when locatable — the landing anchor.

source: Arc<str>

The exact text span was measured against. Carried rather than re-read: a consumer converting the span needs a line index built from these bytes, and a fresh read could observe a newer edit and mis-anchor.

placeholders: Vec<String>

Every variable the entry reads, in first-seen order — exactly the names a bind: in scope has to supply (ADR-0018).

Read off the engine’s own AST at scan time, so an editor offering them is offering the file’s real interface rather than a second description of it that could disagree. Without this the only way to learn a foreign corpus’s variable names is to run the suite and read proef::lower::unbound_placeholder.

Faithful to what the entry reads, so subtract Self::supplied_variables before offering these as bind: keys.

supplied_variables: Vec<String>

Every variable the entry supplies to itself ([Options] variable:).

A name here needs no bind: and may not have one (proef::pack::option_declared_twice), so an editor offering it as a completion would be proposing an edit its own diagnostics then reject.

Trait Implementations

impl Clone for FragmentDef

fn clone(&self) -> FragmentDef

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for FragmentDef

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

Formats the value using the given formatter.