recipe-scrapers-js 0.1.0-alpha.5 → 0.1.0-alpha.6

package/README.md

@@ -19,14 +19,16 @@ A TypeScript/JavaScript library for scraping recipe data from various cooking we
 
 ## Installation
 
+Add the `recipe-scrapers-js` package and its peer dependencies.
+
 ```bash
-npm install recipe-scrapers-js
+npm install recipe-scrapers-js cheerio zod
 # or
-yarn add recipe-scrapers-js
+yarn add recipe-scrapers-js cheerio zod
 # or
-pnpm add recipe-scrapers-js
+pnpm add recipe-scrapers-js cheerio zod
 # or
-bun add recipe-scrapers-js
+bun add recipe-scrapers-js cheerio zod
 ```
 
 ## Usage
@@ -195,6 +197,7 @@ This project is licensed under the MIT License - see the [LICENSE](LICENSE) file
 - Original [recipe-scrapers](https://github.com/hhursev/recipe-scrapers) Python library by [hhursev](https://github.com/hhursev)
 - [Schema.org Recipe specification](https://schema.org/Recipe)
 - [Cheerio](https://cheerio.js.org/) for HTML parsing
+- [Zod](https://zod.dev/) for schema validation
 
 ## Copyright and Usage
 

package/dist/index.js

@@ -1,5 +1,6 @@
 import * as cheerio from "cheerio";
 import { parse, toSeconds } from "iso8601-duration";
+import z from "zod/v4";
 
 //#region src/utils/index.ts
 function isDefined(value) {
@@ -145,7 +146,7 @@ function parseMinutes(value) {
 
 //#endregion
 //#region src/utils/ingredients.ts
-const DEFAULT_GROUP_NAME = "Ingredients";
+const DEFAULT_INGREDIENTS_GROUP_NAME = "Ingredients";
 const DEFAULT_GROUPING_SELECTORS = {
 	wprm: {
 		headingSelectors: [".wprm-recipe-ingredient-group h4", ".wprm-recipe-group-name"],
@@ -227,13 +228,13 @@ function groupIngredients($, ingredientsList, headingSelector, itemSelector) {
 		const $el = $(el);
 		if ($el.is(groupNameSelector)) {
 			const headingText = normalizeString($el.text());
-			currentHeading = headingText || DEFAULT_GROUP_NAME;
+			currentHeading = headingText || DEFAULT_INGREDIENTS_GROUP_NAME;
 			if (!groupings.has(currentHeading)) groupings.set(currentHeading, /* @__PURE__ */ new Set());
 		} else if ($el.is(ingredientSelector)) {
 			const text = normalizeString($el.text());
 			if (!text) continue;
 			const matched = bestMatch(text, ingredients);
-			const heading = currentHeading || DEFAULT_GROUP_NAME;
+			const heading = currentHeading || DEFAULT_INGREDIENTS_GROUP_NAME;
 			if (!groupings.has(heading)) groupings.set(heading, /* @__PURE__ */ new Set());
 			groupings.get(heading)?.add(matched);
 		}
@@ -1076,6 +1077,127 @@ var AllRecipes = class extends AbstractScraper {
 	extractors = {};
 };
 
+//#endregion
+//#region src/scrapers/americastestkitchen.ts
+const recipeIngredientItemSchema = z.object({ fields: z.object({
+	qty: z.string(),
+	preText: z.string(),
+	postText: z.string(),
+	measurement: z.string().nullable(),
+	pluralIngredient: z.boolean(),
+	ingredient: z.object({
+		contentType: z.string(),
+		fields: z.object({
+			title: z.string(),
+			pluralTitle: z.string(),
+			kind: z.string()
+		})
+	})
+}) });
+const recipeIngredientGroupSchema = z.object({ fields: z.object({
+	title: z.string(),
+	recipeIngredientItems: z.array(recipeIngredientItemSchema)
+}) });
+const recipeInstructionSchema = z.object({ fields: z.object({ content: z.string() }) });
+const recipeDataSchema = z.object({
+	totalCookTime: z.number(),
+	recipeTimeNote: z.string().optional(),
+	ingredientGroups: z.array(recipeIngredientGroupSchema),
+	headnote: z.string().optional(),
+	instructions: z.array(recipeInstructionSchema)
+});
+const pagePropsDataSchema = z.object({ props: z.object({ pageProps: z.object({ data: recipeDataSchema }) }) });
+var AmericasTestKitchen = class extends AbstractScraper {
+	data = null;
+	static host() {
+		return "americastestkitchen.com";
+	}
+	extractors = {
+		ingredients: this.ingredients.bind(this),
+		instructions: this.instructions.bind(this),
+		siteName: this.siteName.bind(this)
+	};
+	siteName() {
+		return "America's Test Kitchen";
+	}
+	ingredients(prevValue) {
+		let ingredients = this.parseIngredients();
+		if (!ingredients) ingredients = this.parseHtmlIngredients(prevValue);
+		if (!ingredients) throw new Error("Failed to extract ingredients");
+		return ingredients;
+	}
+	instructions(prevValue) {
+		const data = this.getRecipeData();
+		if (!data) {
+			if (prevValue) return prevValue;
+			throw new Error("Failed to extract instructions");
+		}
+		const { headnote } = data;
+		let headnoteText = "";
+		if (headnote) headnoteText = `Note: ${normalizeString(headnote)}`;
+		const instructionTexts = [];
+		for (const instruction of data.instructions) instructionTexts.push(normalizeString(instruction.fields.content));
+		return new Set([headnoteText, ...instructionTexts]);
+	}
+	parseHtmlIngredients(prevValue) {
+		const headingSelector = "[class*=\"RecipeIngredientGroups_group\"] > span";
+		const ingredientSelector = "[class*=\"RecipeIngredient\"] label";
+		if (isList(prevValue) && prevValue.size > 0) {
+			const result = groupIngredients(this.$, prevValue, headingSelector, ingredientSelector);
+			return result;
+		}
+		return null;
+	}
+	getRecipeData() {
+		if (this.data === null) {
+			const jsonElement = this.$("script[type=\"application/json\"]");
+			const jsonString = jsonElement.html();
+			if (!jsonString) {
+				this.logger.warn("Could not find JSON data script tag");
+				return null;
+			}
+			try {
+				const parsed = pagePropsDataSchema.parse(JSON.parse(jsonString));
+				this.data = parsed.props.pageProps.data;
+			} catch (error) {
+				this.logger.error("Failed to parse JSON data:", error);
+				return null;
+			}
+		}
+		return this.data;
+	}
+	parseIngredientItem(ingredientItem) {
+		const { fields } = ingredientItem;
+		const fragments = [
+			fields.qty || "",
+			fields.measurement || "",
+			fields.ingredient.fields.title || "",
+			fields.postText || ""
+		];
+		const filteredFragments = [];
+		for (const fragment of fragments) if (fragment) filteredFragments.push(fragment.trimEnd());
+		return filteredFragments.join(" ").trimEnd().replace(" ,", ",");
+	}
+	parseIngredients() {
+		const data = this.getRecipeData();
+		if (!data) return null;
+		const { ingredientGroups } = data;
+		if (ingredientGroups.length === 1) {
+			const ingredientSet = /* @__PURE__ */ new Set();
+			for (const item of ingredientGroups[0].fields.recipeIngredientItems) ingredientSet.add(this.parseIngredientItem(item));
+			return ingredientSet;
+		}
+		const ingredientMap = /* @__PURE__ */ new Map();
+		for (const group of ingredientGroups) {
+			const groupTitle = group.fields.title || DEFAULT_INGREDIENTS_GROUP_NAME;
+			const ingredientSet = /* @__PURE__ */ new Set();
+			for (const item of group.fields.recipeIngredientItems) ingredientSet.add(this.parseIngredientItem(item));
+			ingredientMap.set(groupTitle, ingredientSet);
+		}
+		return ingredientMap;
+	}
+};
+
 //#endregion
 //#region src/scrapers/bbcgoodfood.ts
 var BBCGoodFood = class extends AbstractScraper {
@@ -1083,16 +1205,6 @@ var BBCGoodFood = class extends AbstractScraper {
 		return "bbcgoodfood.com";
 	}
 	extractors = { ingredients: this.ingredients.bind(this) };
-	/**
-	* The NYTimes website appears to auto generate it's CSS class names,
-	* which results in them ending with a string a random characters.
-	* Matching the exact class name is likely to break fairly quickly
-	* so instead we are going to match on a partial class name.
-	* For example, h3[class*='ingredientgroup_name'] matches an h3 element
-	* with a class that contains the value 'ingredient_groupname' at least once
-	* anywhere in the element class attribute.
-	* @link https://developer.mozilla.org/en-US/docs/Web/CSS/Attribute_selectors
-	*/
 	ingredients(prevValue) {
 		const headingSelector = ".recipe__ingredients h3";
 		const ingredientSelector = ".recipe__ingredients li";
@@ -1104,6 +1216,15 @@ var BBCGoodFood = class extends AbstractScraper {
 	}
 };
 
+//#endregion
+//#region src/scrapers/eatingwell.ts
+var EatingWell = class extends AbstractScraper {
+	static host() {
+		return "eatingwell.com";
+	}
+	extractors = {};
+};
+
 //#endregion
 //#region src/scrapers/epicurious.ts
 var Epicurious = class extends AbstractScraper {
@@ -1124,16 +1245,6 @@ var NYTimes = class extends AbstractScraper {
 		return "cooking.nytimes.com";
 	}
 	extractors = { ingredients: this.ingredients.bind(this) };
-	/**
-	* The NYTimes website appears to auto generate it's CSS class names,
-	* which results in them ending with a string a random characters.
-	* Matching the exact class name is likely to break fairly quickly
-	* so instead we are going to match on a partial class name.
-	* For example, h3[class*='ingredientgroup_name'] matches an h3 element
-	* with a class that contains the value 'ingredient_groupname' at least once
-	* anywhere in the element class attribute.
-	* @link https://developer.mozilla.org/en-US/docs/Web/CSS/Attribute_selectors
-	*/
 	ingredients(prevValue) {
 		const headingSelector = "h3[class*=\"ingredientgroup_name\"]";
 		const ingredientSelector = "li[class*=\"ingredient\"]";
@@ -1184,7 +1295,9 @@ var SimplyRecipes = class extends AbstractScraper {
 */
 const scrapers = {
 	[AllRecipes.host()]: AllRecipes,
+	[AmericasTestKitchen.host()]: AmericasTestKitchen,
 	[BBCGoodFood.host()]: BBCGoodFood,
+	[EatingWell.host()]: EatingWell,
 	[Epicurious.host()]: Epicurious,
 	[SeriousEats.host()]: SeriousEats,
 	[SimplyRecipes.host()]: SimplyRecipes,

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "recipe-scrapers-js",
-  "version": "0.1.0-alpha.5",
+  "version": "0.1.0-alpha.6",
   "license": "MIT",
   "description": "A recipe scrapers library",
   "author": {
@@ -39,7 +39,8 @@
     "prepublishOnly": "bun run lint && bun run build"
   },
   "peerDependencies": {
-    "cheerio": "^1.1.0"
+    "cheerio": "^1.1.0",
+    "zod": "^3.25.74"
   },
   "dependencies": {
     "iso8601-duration": "^2.1.2",