Struct UnifiedExtractor 

pub struct UnifiedExtractor { /* private fields */ }

Unified extractor that extracts all information in a single AI call.

This replaces the separate AiMemoryExtractor and FocusExtractor, reducing API calls and providing consistent extraction.

Implementations

impl UnifiedExtractor

pub fn new(provider: Box<dyn Provider>, model: String) -> Self

Create a new unified extractor.

pub fn new_minimal(model: String) -> Self

Create a minimal unified extractor for background tasks.

pub async fn extract_unified( &self, text: &str, session_id: Option<&str>, project_path: Option<&str>, ) -> Result<UnifiedExtractionResult>

Extract all information from conversation text in a single AI call.

pub async fn extract_unified_with_foci( &self, text: &str, existing_foci: &[(&str, &str, &[String])], session_id: Option<&str>, project_path: Option<&str>, ) -> Result<UnifiedExtractionResult>

Extract all information WITH focus selection in a single AI call.

This method receives existing focuses and asks AI to select the best match or create a new focus if none matches. This ensures focus continuity.

Arguments

• text - Conversation text to analyze
• existing_foci - Current focus points from FocusManager (id, topic, keywords)
• session_id - Optional session ID
• project_path - Optional project path

Returns

UnifiedExtractionResult with focus_decision field populated

pub fn model_name(&self) -> &str

Get the model name used for extraction.