Created
May 11, 2020 11:15
-
-
Save dave-shawley/a46146d954d65763ba0a9cd4c669045b to your computer and use it in GitHub Desktop.
Quick & dirty scraper for recipes from Milk Street
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
| #!/usr/bin/env python | |
| # | |
| # Requirements: requests, beautifulsoup4 | |
| # | |
| import argparse | |
| import dataclasses | |
| import typing | |
| import bs4 | |
| import requests | |
| @dataclasses.dataclass | |
| class Ingredient: | |
| quantity: str | |
| label: str | |
| @dataclasses.dataclass | |
| class Recipe: | |
| title: str | |
| url: str | |
| ingredients: typing.List[Ingredient] = dataclasses.field( | |
| default_factory=list) | |
| instructions: typing.List[str] = dataclasses.field(default_factory=list) | |
| parser = argparse.ArgumentParser() | |
| parser.add_argument('url', metavar='URL', type=str) | |
| parser.add_argument('output', metavar='OUTPUT', type=argparse.FileType('w')) | |
| opts = parser.parse_args() | |
| rsp = requests.get(opts.url) | |
| rsp.raise_for_status() | |
| soup = bs4.BeautifulSoup(rsp.text, features='html.parser') | |
| content = soup.find(class_='recipe__text__content') | |
| recipe = Recipe( | |
| title=soup.find('h1', class_='recipe-header__title').text, | |
| url=opts.url, | |
| ) | |
| parent = content.find('ul', class_='ingredients-list') | |
| for ingredient in parent.find_all('li', class_='ingredient'): | |
| recipe.ingredients.append( | |
| Ingredient(ingredient.find(class_='ingredient__quantity').text, | |
| ingredient.find(class_='ingredient__label').text)) | |
| parent = content.find('ol', class_='recipe__directions__list') | |
| for direction in parent.find_all('div', class_='recipe__direction__text'): | |
| recipe.instructions.append(direction.text) | |
| opts.output.write('<html><head><title>') | |
| opts.output.write(recipe.title) | |
| opts.output.write('</title><meta charset=utf-8></head><body>') | |
| opts.output.write('<h1>') | |
| opts.output.write(recipe.title) | |
| opts.output.write('</h1>') | |
| opts.output.write('<h2>Ingredients</h2><table>') | |
| for ingredient in recipe.ingredients: | |
| opts.output.write('<tr><td>') | |
| opts.output.write(ingredient.quantity) | |
| opts.output.write('</td><td>') | |
| opts.output.write(ingredient.label) | |
| opts.output.write('</td></tr>') | |
| opts.output.write('</table><h2>Directions</h2><ol>') | |
| for instruction in recipe.instructions: | |
| opts.output.write('<li>') | |
| opts.output.write(instruction) | |
| opts.output.write('</li>') | |
| opts.output.write('</ol><p><i>Extracted from ') | |
| opts.output.write(opts.url) | |
| opts.output.write('</i></p></body></html>') |
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment