batchCooking/apps/api/test/ingredient-matcher.test.ts

import { INGREDIENT_LABEL_SYNONYMS_EN } from "@batch-cooking/shared";
import { expect } from "chai";
import { prisma } from "../src/db/prisma.js";
import {
  type IngredientMatchEntry,
  type UnitMatchEntry,
  extractQuantity,
  loadIngredientCatalog,
  loadUnitCatalog,
  matchIngredientName,
  matchUnit,
} from "../src/lib/ingredient-matcher.js";
import { resetDatabase } from "../test-support/reset-db.js";

describe("ingredient-matcher", () => {
  describe("matchIngredientName", () => {
    const tomato: IngredientMatchEntry = { ingredientId: 1, label: "Tomato" };
    const chicken: IngredientMatchEntry = { ingredientId: 2, label: "Chicken" };
    const chickenBreast: IngredientMatchEntry = { ingredientId: 3, label: "Chicken breast" };
    const onion: IngredientMatchEntry = { ingredientId: 4, label: "Onion" };
    const allPurposeFlour: IngredientMatchEntry = { ingredientId: 5, label: "All-purpose flour" };
    const catalog = [tomato, chicken, chickenBreast, onion, allPurposeFlour];

    it("matches an exact single-word label", () => {
      expect(matchIngredientName("tomato", catalog)).to.equal(tomato.ingredientId);
    });

    it("is case- and accent-insensitive", () => {
      expect(matchIngredientName("TOMATO", catalog)).to.equal(tomato.ingredientId);
      expect(matchIngredientName("Tömato", catalog)).to.equal(tomato.ingredientId);
    });

    it("tolerates a regular plural", () => {
      expect(matchIngredientName("tomatoes", catalog)).to.equal(tomato.ingredientId);
      expect(matchIngredientName("onions", catalog)).to.equal(onion.ingredientId);
    });

    it("tolerates extra descriptive words around the match", () => {
      expect(matchIngredientName("2 large diced yellow onions", catalog)).to.equal(
        onion.ingredientId,
      );
    });

    it("prefers the more specific multi-word label over a shorter one it contains", () => {
      expect(matchIngredientName("boneless skinless chicken breasts", catalog)).to.equal(
        chickenBreast.ingredientId,
      );
    });

    it("still matches the shorter label when the more specific one isn't mentioned", () => {
      expect(matchIngredientName("diced chicken thighs", catalog)).to.equal(chicken.ingredientId);
    });

    it("matches a hyphenated multi-word label", () => {
      expect(matchIngredientName("2 cups all-purpose flour", catalog)).to.equal(
        allPurposeFlour.ingredientId,
      );
    });

    it("doesn't false-positive a short label inside an unrelated longer word", () => {
      // "egg" must not match inside "eggplant" — whole-token comparison, not substring.
      const eggplant: IngredientMatchEntry = { ingredientId: 6, label: "Eggplant" };
      const egg: IngredientMatchEntry = { ingredientId: 7, label: "Egg" };
      expect(matchIngredientName("eggplant", [egg, eggplant])).to.equal(eggplant.ingredientId);
    });

    it("returns null when nothing matches", () => {
      expect(matchIngredientName("mango", catalog)).to.equal(null);
    });

    it("returns null for an empty catalog", () => {
      expect(matchIngredientName("tomato", [])).to.equal(null);
    });

    it("returns null for an empty name", () => {
      expect(matchIngredientName("", catalog)).to.equal(null);
    });

    it("matches an alternate wording of the same ingredient via a second catalog entry sharing its ingredientId (issue #54)", () => {
      const vanillaBean: IngredientMatchEntry = { ingredientId: 8, label: "Vanilla bean" };
      const vanillaBeanSynonym: IngredientMatchEntry = { ingredientId: 8, label: "Vanilla pod" };
      const synonymCatalog = [vanillaBean, vanillaBeanSynonym];

      expect(matchIngredientName("1 vanilla pod", synonymCatalog)).to.equal(8);
      expect(matchIngredientName("1 vanilla bean", synonymCatalog)).to.equal(8);
    });

    it("breaks a same-specificity tie by the lowest ingredientId", () => {
      const onionA: IngredientMatchEntry = { ingredientId: 20, label: "Onion" };
      const onionB: IngredientMatchEntry = { ingredientId: 21, label: "Onion" };
      expect(matchIngredientName("onion", [onionB, onionA])).to.equal(20);
    });
  });

  describe("matchUnit", () => {
    const gram: UnitMatchEntry = { unitId: 1, synonyms: ["g", "gram", "grams"] };
    const tablespoon: UnitMatchEntry = {
      unitId: 2,
      synonyms: ["tbsp", "tbs", "tablespoon", "tablespoons"],
    };
    const cup: UnitMatchEntry = { unitId: 3, synonyms: ["cup", "cups"] };
    const catalog = [gram, tablespoon, cup];

    it("matches a full word synonym", () => {
      expect(matchUnit("tablespoon", catalog)).to.equal(tablespoon.unitId);
    });

    it("matches an abbreviation synonym", () => {
      expect(matchUnit("tbsp", catalog)).to.equal(tablespoon.unitId);
    });

    it("matches a plural synonym via the same stemming as ingredients", () => {
      expect(matchUnit("cups", catalog)).to.equal(cup.unitId);
    });

    it("is case-insensitive", () => {
      expect(matchUnit("TBSP", catalog)).to.equal(tablespoon.unitId);
    });

    it("only looks at the first word — ignores trailing text", () => {
      expect(matchUnit("cup flour", catalog)).to.equal(cup.unitId);
    });

    it("doesn't match a short abbreviation inside an unrelated word", () => {
      // "g" alone must not match "grated" — whole-token comparison.
      expect(matchUnit("grated", catalog)).to.equal(null);
    });

    it("returns null when nothing matches", () => {
      expect(matchUnit("pound", catalog)).to.equal(null);
    });

    it("returns null for an empty catalog", () => {
      expect(matchUnit("cup", [])).to.equal(null);
    });

    it("returns null for an empty string", () => {
      expect(matchUnit("", catalog)).to.equal(null);
    });
  });

  describe("extractQuantity", () => {
    it("extracts a plain integer", () => {
      expect(extractQuantity("2 onions")).to.deep.equal({ quantity: 2, remainder: "onions" });
    });

    it("extracts a decimal using a dot", () => {
      expect(extractQuantity("1.5 cups flour")).to.deep.equal({
        quantity: 1.5,
        remainder: "cups flour",
      });
    });

    it("extracts a decimal using a comma", () => {
      expect(extractQuantity("1,5 cups flour")).to.deep.equal({
        quantity: 1.5,
        remainder: "cups flour",
      });
    });

    it("extracts a simple fraction", () => {
      expect(extractQuantity("1/2 cup sugar")).to.deep.equal({
        quantity: 0.5,
        remainder: "cup sugar",
      });
    });

    it("extracts a mixed number", () => {
      expect(extractQuantity("1 1/2 cups sugar")).to.deep.equal({
        quantity: 1.5,
        remainder: "cups sugar",
      });
    });

    it("returns null quantity and the trimmed original text when there's no leading number", () => {
      expect(extractQuantity("salt to taste")).to.deep.equal({
        quantity: null,
        remainder: "salt to taste",
      });
    });

    it("trims surrounding whitespace", () => {
      expect(extractQuantity("  2 eggs  ")).to.deep.equal({ quantity: 2, remainder: "eggs" });
    });

    it("only takes the first number of a hyphenated range", () => {
      expect(extractQuantity("2-3 carrots")).to.deep.equal({
        quantity: 2,
        remainder: "-3 carrots",
      });
    });
  });

  describe("loadIngredientCatalog / loadUnitCatalog", () => {
    beforeEach(async () => {
      await resetDatabase();
    });

    after(async () => {
      await prisma.$disconnect();
    });

    it("loads one entry per Ingredient that has an English label, plus one per alternate wording (INGREDIENT_LABEL_SYNONYMS_EN), keyed by real ingredientId", async () => {
      const tomato = await prisma.ingredient.findFirstOrThrow({ where: { key: "tomato" } });
      const vanillaBean = await prisma.ingredient.findFirstOrThrow({
        where: { key: "vanillaBean" },
      });
      const ingredientCount = await prisma.ingredient.count();
      const synonymCount = Object.values(INGREDIENT_LABEL_SYNONYMS_EN).reduce(
        (sum, synonyms) => sum + synonyms.length,
        0,
      );

      const catalog = await loadIngredientCatalog();

      // Every seeded ingredient has an authored English label (verified at
      // generation time — see packages/shared/src/data/catalog-labels-en.ts),
      // so nothing should be silently skipped — plus one extra entry per
      // synonym (issue #54), sharing the same ingredientId as the primary
      // label's entry.
      expect(catalog).to.have.length(ingredientCount + synonymCount);
      const tomatoEntry = catalog.find((entry) => entry.ingredientId === tomato.id);
      expect(tomatoEntry?.label).to.equal("Tomato");

      const vanillaBeanEntries = catalog.filter((entry) => entry.ingredientId === vanillaBean.id);
      expect(vanillaBeanEntries.map((entry) => entry.label)).to.deep.equal([
        "Vanilla bean",
        "Vanilla pod",
      ]);
    });

    it("loads one entry per Unit that has English synonyms, keyed by real unitId", async () => {
      const cup = await prisma.unit.findFirstOrThrow({ where: { key: "cup" } });
      const unitCount = await prisma.unit.count();

      const catalog = await loadUnitCatalog();

      expect(catalog).to.have.length(unitCount);
      const cupEntry = catalog.find((entry) => entry.unitId === cup.id);
      expect(cupEntry?.synonyms).to.deep.equal(["cup", "cups"]);
    });
  });
});