Bite

Bite is an app for food lovers and organization junkies. It offers a convenient way to save, view, and recommend restaurants to friends.

##Technology Stack

Python, Flask, SQLAlchemy, SQLite, Javascript, jQuery, AJAX, Google Maps API, Foursquare API, HTML5 Geolocation API, HTML5, Jinja, CSS3

##Features

• Login
• Restaurant search by location and restaurant name or cuisine
• Bookmarks
• Bookmark deletion
• User-to-user recommendations
• Proximity-based restaurant mapping
• Detection of user location
• Client-side validation

##Technologies

###Foursquare API To get up-to-date and crowd sourced restaurant information, Bite uses the Foursquare venues API. When a user searches for a restaurant or cuisine in a given location, a function is called which dynamically creates an HTTP GET request to the Foursquare Venues API. This request includes the user’s query information as parameters, as well as a category id limiting the response to only include venues in the Foursquare “food” category. The response is then decoded from JSON to a python dictionary and sent to the search results page.

###Google Maps API On load of the map page, a SQLAlchemy query runs which queries the database for the user’s bookmarks and the associated restaurant information, including latitude and longitude. This information is sent over as JSON to the route, which loops through all the bookmarks, and places pins on the map using the Google Maps API.

###HTML5 Geolocation API Bite integrates the HTML5 Geolocation API with the Google Maps API, the browser asks the user whether it can access its current location. If so, the location information is sent over to the function that creates the Google Map, and the map is centered around the user’s current location. If geolocation fails or the user does not authorize it, the map defaults to San Francisco.

This functionality allows users visiting neighborhoods where they have bookmarked restaurants the flexibility of opening the app and seeing bookmarked restaurants in their proximity.

###SQLAlchemy ORM #####Bookmarks After getting back a list of search results from the Foursquare API, users select the restaurant they would like to save as a “bookmark.” When a user saves a bookmark, there is a server side check with SQLAlchemy for whether it has already been bookmarked for that user and whether the restaurant is already in the database of restaurants. If the restaurant and bookmark are not already in database yet, their API information is cached in order to optimize load time.

#####Recommendations From the search results page that users see when they submit a restaurant search, users also have the options to recommend restaurants to other users registered with Bite. Once a user selects the user to send the recommendation to, SQLAlchemy queries check whether the recipient already has that restaurant bookmarked and whether the recommender has already recommended that restaurant to the recipient. If both of those conditions are not true, a new recommendation is created in the database.

Upon login, users who have pending recommendations are prompted to view and respond to recommendations by accepting or passing on specific recommendations. When a user responds to a recommendation, the pending attribute of the recommendation is changed from True to False. If they accept a recommendation, new bookmark and bookmarkrecommendation records are created with the restaurant and user information from that recommendation.

##Data Model

Below is a visual representation of the Bite data model:

#####Table descriptions:

#####User The users table is used to represent each user of the Bite app and holds login information.

#####Restaurant The restaurants table holds information about the restaurants that users have bookmarked or recommended.

#####Bookmark A bookmark record is created in the bookmarks table when a user selects a restaurant to save as a bookmark. It links the user to a specific restaurant. A user may have many bookmarks.

#####Recommendation A recommendation record in the recommendations table is created when a user selects a restaurant to recommend to another user. This table binds a recommender user id and recipient user id to a specific restaurant id. To keep track of when the recommendation has been responded to by the recipient user, each recommendation record has a “pending” attribute that is initially set to the Boolean “True” when the recommendation is first created, and is changed to “False” when the user approves or passes on the request. A user may have many recommendations.

#####BookmarkRecommendation If and when user accepts a recommendation, a new bookmarkrecommendation record is created in this table. The bookmarkrecommendation table is used to represent the recommendation(s) associated with each bookmark. Each time a user accepts a recommendation from another user, a new bookmarkrecommendation is created.

For example, if two users recommend the same restaurant to a third user and that third user accepts the recommendation, a) one new bookmark is created with the restaurant id of the recommendation, and b) two new records are created in the bookmarkrecommendation table with the same bookmark id but two different recommendation id’s.

A user may have many bookmarkrecommendations.

##Installation

• Clone the repo
• Set up a virtual environment
• Pip install -r requirements.txt
• At the command line:

python -i db/model.py

Then:

create_db()

• In a new shell use ls command in project folder and verify main.db exists in the “db” repository
• To see the structure of the database, at the command line:

sqlite3 db/main.db 

Then:

sqlite> .mode columns
sqlite> .tables on
sqlite> .tables 
sqlite> .schema