chetnam / recipes Goto Github PK

Intro

Note: API is currently shut-down, so none of the links below will work.

Cooking is fun, and websites that share recipes are great. Detailed stories are fantastic, but maybe not if you have to scroll a mile to get to the actual recipe.

Here's a list of recipe-sharing websites ranked by how many words are used before you can get to the recipe: https://api-get-recipes.azurewebsites.net/api/sources.

Check out API Details below to see what other data this project has collected.

Summary

This project collects information about recipes from various recipe-sharing websites, stores this information, and provides access to this information through an API. There are two main components of this project, currently deployed through Microsoft Azure Functions:

• scrape-recipes: Scripts that collects information about recipes from sources via web-scraping and stores in a database; runs once a day; written in Python
• get-recipes: Provides endpoints to access different types of information from the data collected; written in NodeJS

API Details From the API, you can

• Determine the verbosity of a source before you get to the recipe:

  https://api-get-recipes.azurewebsites.net/api/sources

• Get five recipes from a given source. Use the query paramter 'source' to specify your requested source. For example, if you wanted recipes from Delish:

  https://api-get-recipes.azurewebsites.net/api/sources?source=Delish

• Get all recipes collected for a certain keyword/food. Use the query parameter 'topic' to specify your keyword or food of choice. For example, if you wanted all recipes related to tacos, try

  https://api-get-recipes.azurewebsites.net/api/recipes?topic=taco

• Get a random list of five recipes. If you're feeling spontaneous or indecisive, try

  https://api-get-recipes.azurewebsites.net/api/recipes/random

Other Details

• This project currently collects recipes from AllRecipes, Delish, and Love&OliveOil. It collects no more than 10 recipes a day from each of these sites.
• Both the scraping script and the API are hosted on Azure Functions. Python on the Consumption Plan is a preview offering in Azure, so it is quite possible that the scraper could be unavailable without warning. The API is in NodeJS on the Consumption Plan, and should be stable; however, it is currently hosted on a Windows machine and a serverless service, which means the cold start time can be non-trivial -- i.e. you might have to wait a couple seconds when you click on a link from the API Details section.
• More recipe-collection sources might be added in the future.