Home

Awesome

The internet is full of recipes. A recipe is really only the directions and the ingredients, but for any given recipe website you always get a lot more information thatn you want (comments, links, advertisements). For instance the page for Alton Brown's waffle recipe looks like this:

Website example

How can you just pull out the directions and ingredients, in a tag-independent way that can be used across any website?

Recipe extractor

How it works

First it grabs the Markdown-formatted text of the html page using html2text. It evalutes the number of 'cooking' related words in each line and pulls out the peak.

Website example

It does the same for extracting the ingredient list. Then it scans the directions and the ingredient list to determine what the actual baking time would be and gives that estimate. Finally, it parses the measurements and foods in the ingredient list and cross-references the USDA SR27 food databse to get accurate estimations of the nutrition content.

To Do

Installation

Requirements

sudo apt-get install python-nltk
sudo pip install pint

Then goto python console

>>> import nltk
>>> nltk.download('brown')
>>> nltk.download('maxent_treebank_pos_tagger')
>>> nltk.download('wordnet')

Clone the SR27 database downloader and run the makefile to generate the database. Copy the database into the same folder. Then go into the database and make sqlite3 fts4 virtual table

drop table data;
drop table nutrition_date;
drop table nutrition_def;

create virtual table data using fts4(ndb_no,shrt_desc,long_desc,com_desc);
insert into data(ndb_no,shrt_desc,long_desc,com_desc) select ndb_no,shrt_desc,long_desc,com_desc from food_des;

create virtual table nutrition_data using fts4(ndb_no,nutr_no,nutr_val);
insert into nutrition_data(ndb_no,nutr_no,nutr_val) select ndb_no,nutr_no,nutr_val from nut_data;

create virtual table nutrition_def using fts4(nutr_no,units,tagname);
insert into nutrition_def(nutr_no,units,tagname) select nutr_no,units,tagname from nutr_def;

Example Output

Visit the current incantation at http://ips.colab.duke.edu:8081/extractor.html and run which shows the extraction of the following recipe: http://www.foodnetwork.com/recipes/alton-brown/baked-macaroni-and-cheese-recipe.html.

Ingredients

Directions

In a large pot of boiling, salted water cook the pasta to al dente. While the pasta is cooking, in a separate pot, melt the butter. Whisk in the flour and mustard and keep it moving for about five minutes. Make sure it's free of lumps. Stir in the milk, onion, bay leaf, and paprika. Simmer for ten minutes and remove the bay leaf. Temper in the egg. Stir in 3/4 of the cheese. Season with salt and pepper. Fold the macaroni into the mix and pour into a 2-quart casserole dish. Top with remaining cheese. Melt the butter in a saute pan and toss the bread crumbs to coat. Top the macaroni with the bread crumbs. Bake for 30 minutes. Remove from oven and rest for five minutes before serving. Remember to save leftovers for fried Macaroni and Cheese. Recipe courtesy Alton Brown

+1 minute for tossing.

Calculated time: 51 minute

Calculated cost: $24.64

Serving size is about 13.0

Nutrition data (ALL)

Main

Sugars

Other

Metals

Vitamins

Amino acids

Steroids

Fatty Acids