iFoodie (愛食記) Taiwan Restaurants Scraper

Extracts restaurant listings from ifoodie.tw (愛食記), a major Taiwanese restaurant review and ranking site (≈15,000 restaurants in Taipei alone, every city and county covered). Filter by city, district, keyword (cuisine, dish or restaurant name) and sort order, and get one clean JSON record per restaurant.

No login, no API key, no proxy needed. Pay-per-event: one restaurant event per restaurant saved. List pages and restaurant pages that are fetched but not saved cost nothing.

What you get

Each record (list data, always present):

Field	Example
id, url	559d314fc03a103ee86c4e16, https://ifoodie.tw/restaurant/559d314fc03a103ee86c4e16-松滿樓
name, storeName	松滿樓
city, district, address, postalCode	台北市, 大安區, 臺北市大安區濟南路三段26號, 106
lat, lng	25.040058, 121.5349414
phone, website, bookingUrl, menuUrl	0227518479, http://www.smlts.com.tw/, OpenTable / inline links
rating, ifoodieRating, ratingDistribution	user rating 4 (0–5), editorial score 3.2, { "1": 0, …, "5": 4 }
reviewCnt, blogCnt	1942, 7
avgPrice, priceLevel	670 (TWD per person), 2 (1 cheap … 4 expensive)
categories	["中式料理", "合菜", "午餐", "晚餐"]
openNow, openingHours, openingHoursList	false, 今日營業: 11:30-14:00, 17:30-20:50, one line per weekday. openNow and the 現正營業: / 今日營業: prefix reflect the moment of the scrape (Taiwan time).
coverUrl, summary, intro, description	cover photo, one-line tagline, short and long editorial blurbs
lastVisit, scrapedAt	ISO timestamps

With Fetch restaurant pages on (default), each record also has: facilities (包廂, online booking…), notes (service charge, cards, reservation advice, minimum spend), dishes (signature dishes), priceRange, priceCategory, suitableGroups, mealTypes, landmarks (nearest MRT station etc.), businessDistricts, tags, portionSize, environment (ambience description).

Input

Example: ramen in Da'an, best-rated first —

{ "city": "台北市", "district": "大安區", "keyword": "拉麵", "sortby": "rating", "maxItems": 100 }

Field	Meaning
city (required)	City or county as ifoodie spells it: 台北市, 新北市, 桃園市, 台中市, 台南市, 高雄市, 基隆市, 新竹市, 新竹縣, 宜蘭縣, 花蓮縣, 台東縣, 屏東縣, 嘉義市, 彰化縣, 南投縣, 苗栗縣, 雲林縣, 澎湖縣, 金門縣. An unknown city fails the run immediately (nothing is charged).
district	Optional district, e.g. 大安區, 信義區. A misspelt district is not an error on ifoodie's side: the run ends with zero restaurants and a warning in the log.
keyword	Optional Chinese keyword: cuisine (拉麵, 火鍋, 燒肉), meal (早午餐, 宵夜), venue (咖啡廳) or a restaurant name.
sortby	default (ifoodie's own ranking: by rating on a plain city list, by recent activity on keyword lists), recent (recently reviewed), popular (most visited) or rating.
maxItems	Stop after this many restaurants. Default 100.
fetchDetails	Open each restaurant page for the extra fields (default on). Off = list data only, ~15× fewer requests.
maxConcurrency	Parallel fetches, default 5.
proxyConfiguration	Optional; not needed for ifoodie.tw.

Reach and limits

• ifoodie shows at most 67 pages × 15 = ~1000 restaurants per query. To cover a whole city, run once per district (台北市 has 12) or per keyword; the actor de-duplicates within a run, so combine outputs across runs by id.
• Sorting by recent / popular (and default on keyword lists, which resolves to recent) is not perfectly stable between page requests; a few restaurants may shift pages and be skipped on deep runs. rating is stable.
• Text is Traditional Chinese as published on the site. address uses the official 臺北市 spelling while city uses ifoodie's 台北市.

Pricing

Pay-per-event with a single event, restaurant, charged once per restaurant pushed to the dataset. If your run's maximum total charge cannot pay for all of maxItems, the actor stops at the budget and says so in the status message; nothing is ever charged beyond it. A run that finds no restaurants (bad keyword, empty district) costs nothing.

Data source and terms

Only public pages of ifoodie.tw are read, at a modest rate (5 parallel requests by default). Reviews, user profiles and blogger check-ins are deliberately not extracted — only the restaurant's own business data and ifoodie's editorial summaries. Please respect ifoodie's terms of use and Taiwan's Personal Data Protection Act when you use the data.