Struct CompletionEngine 

pub struct CompletionEngine { /* private fields */ }

High-level entry point for parsing and completing command lines.

Reuse one engine across many completion requests for the same tree. The constructor precomputes global context-only flags so each request only pays for parsing, context resolution, and ranking.

Implementations

impl CompletionEngine

pub fn new(tree: CompletionTree) -> Self

Creates an engine for a prebuilt completion tree.

The engine keeps a clone of the tree for suggestion ranking and caches global context-only flags up front.

pub fn complete( &self, line: &str, cursor: usize, ) -> (CursorState, Vec<SuggestionOutput>)

Parses line at cursor and returns the cursor state with suggestions.

This is the canonical one-shot completion entrypoint when callers do not need the intermediate CompletionAnalysis.

Examples

use osp_cli::completion::{
    CommandSpec, CompletionEngine, CompletionTreeBuilder, SuggestionOutput,
};

let tree = CompletionTreeBuilder
    .build_from_specs(&[CommandSpec::new("ldap")], [])
    .expect("tree should build");
let engine = CompletionEngine::new(tree);

let (cursor, suggestions) = engine.complete("ld", 2);

assert_eq!(cursor.token_stub, "ld");
assert!(matches!(
    suggestions.first(),
    Some(SuggestionOutput::Item(item)) if item.text == "ldap"
));

pub fn suggestions_for_analysis( &self, analysis: &CompletionAnalysis, ) -> Vec<SuggestionOutput>

Generates suggestions for a previously computed completion analysis.

pub fn analyze(&self, line: &str, cursor: usize) -> CompletionAnalysis

Parses line and resolves completion context at cursor.

cursor is clamped to the line length and to a valid UTF-8 character boundary before parsing.

pub fn analyze_command( &self, full_cmd: CommandLine, cursor_cmd: CommandLine, cursor: CursorState, ) -> CompletionAnalysis

Resolves completion state from pre-parsed command representations.

This is mainly useful for tests or callers that already have split command state for the full line and the cursor-local prefix.

pub fn tokenize(&self, line: &str) -> Vec<String>

Tokenizes a shell-like command line using the parser’s permissive rules.

pub fn matched_command_len_tokens(&self, tokens: &[String]) -> usize

Returns how many leading tokens resolve to a command path in the tree.

pub fn classify_match( &self, analysis: &CompletionAnalysis, value: &str, ) -> MatchKind

Classifies a candidate value relative to the current completion analysis.

Trait Implementations

impl Clone for CompletionEngine

fn clone(&self) -> CompletionEngine

Returns a duplicate of the value.

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

Performs copy-assignment from source.

impl Debug for CompletionEngine

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

Formats the value using the given formatter.