/**
 * Embedding utilities for note intelligence — duplicate detection, related notes, knowledge gaps.
 *
 * Uses @bytelyst/llm embed() when available, falls back gracefully.
 */

import { llm } from './llm.js';
import { config } from './config.js';

/**
 * Generate an embedding vector for a text string.
 * Returns null if the LLM provider doesn't support embeddings.
 */
export async function embedText(text: string): Promise<number[] | null> {
  const provider = llm();
  if (!provider.embed) return null;

  const trimmed = text.slice(0, 8000); // Most embedding models cap at ~8k tokens
  const res = await provider.embed({
    input: trimmed,
    model: config.LLM_EMBEDDING_MODEL,
  });
  return res.embeddings[0] ?? null;
}

/**
 * Compute cosine similarity between two embedding vectors.
 * Returns a value between -1 and 1 (1 = identical).
 */
export function cosineSimilarity(a: number[], b: number[]): number {
  if (a.length !== b.length || a.length === 0) return 0;
  let dot = 0;
  let magA = 0;
  let magB = 0;
  for (let i = 0; i < a.length; i++) {
    dot += a[i] * b[i];
    magA += a[i] * a[i];
    magB += b[i] * b[i];
  }
  const denom = Math.sqrt(magA) * Math.sqrt(magB);
  if (denom === 0) return 0;
  return dot / denom;
}

/**
 * Strip HTML and normalize whitespace for embedding input.
 */
export function stripHtmlForEmbedding(html: string): string {
  return html.replace(/<[^>]*>/g, ' ').replace(/\s+/g, ' ').trim();
}