kochwas/src/lib/server/recipes/importer.ts

import type Database from 'better-sqlite3';
import type { Recipe } from '$lib/types';
import { fetchText } from '../http';
import { extractRecipeFromHtml } from '../parsers/json-ld-recipe';
import { downloadImage } from '../images/image-downloader';
import {
  getRecipeById,
  getRecipeIdBySourceUrl,
  insertRecipe
} from './repository';

export class ImporterError extends Error {
  constructor(
    public readonly code:
      | 'INVALID_URL'
      | 'FETCH_FAILED'
      | 'NO_RECIPE_FOUND',
    message: string
  ) {
    super(message);
    this.name = 'ImporterError';
  }
}

function hostnameOrThrow(url: string): string {
  try {
    return new URL(url).hostname.toLowerCase();
  } catch {
    throw new ImporterError('INVALID_URL', `Not a valid URL: ${url}`);
  }
}

// Manuelle URL-Importe sind absichtlich NICHT mehr auf die allowed_domain-
// Whitelist beschränkt — der User pastet bewusst eine URL und erwartet,
// dass der Import klappt. Die Whitelist bleibt für die Web-Suche (searxng)
// relevant, weil dort ein breites Crawl-Feld eingeschränkt werden soll.
export async function previewRecipe(_db: Database.Database, url: string): Promise<Recipe> {
  const host = hostnameOrThrow(url);
  let html: string;
  try {
    html = await fetchText(url);
  } catch (e) {
    throw new ImporterError('FETCH_FAILED', (e as Error).message);
  }
  const recipe = extractRecipeFromHtml(html);
  if (!recipe) {
    throw new ImporterError('NO_RECIPE_FOUND', 'No schema.org/Recipe JSON-LD on page');
  }
  recipe.source_url = url;
  recipe.source_domain = host.replace(/^www\./, '');
  return recipe;
}

export async function importRecipe(
  db: Database.Database,
  imageDir: string,
  url: string
): Promise<{ id: number; duplicate: boolean; recipe: Recipe }> {
  const existingId = getRecipeIdBySourceUrl(db, url);
  if (existingId !== null) {
    const recipe = getRecipeById(db, existingId);
    if (recipe) return { id: existingId, duplicate: true, recipe };
  }
  const recipe = await previewRecipe(db, url);
  let imageFilename: string | null = null;
  if (recipe.image_path) {
    imageFilename = await downloadImage(recipe.image_path, imageDir);
  }
  recipe.image_path = imageFilename;
  const id = insertRecipe(db, recipe);
  const persisted = getRecipeById(db, id);
  return { id, duplicate: false, recipe: persisted ?? recipe };
}