Travis Optimiser

Deployed via google cloud functions

Endpoint: (GET) https://asia-east2-travis-mvp-v2.cloudfunctions.net/travis-recommender

To make a request: attach requested google places id in a header, accepts either 1 or 2 headers in a list: ids:["ChIJczgQh8lC1moR9r9gP44FRvY", "ChIJdedaLk5d1moRQOX0CXZWBB0"]

• These are Chinatown and Southern Cross respectively, so search should return shops in the middle

Response format (json): {"33":"ChIJ9RT5wLRC1moRpIkhaxraJMc","37":"ChIJcxVwKrVC1moRcHnRJ0dR5vM"}

• responses generally should be list of 5, unless places cannot be found
• to check a id, use https://maps.googleapis.com/maps/api/place/details/json?key=<>&place_id=<>
  • key=<> add your gmaps api key
  • place_id=<> insert the id that was returned

Curl test (status test): curl -X GET -k -H 'ids: ["ChIJczgQh8lC1moR9r9gP44FRvY", "ChIJdedaLk5d1moRQOX0CXZWBB0"]' -i 'https://asia-east2-travis-mvp-v2.cloudfunctions.net/travis-recommender'

DONE list

• Flat file location: Stored on a GCP Cloud Storage bucket, see config.yml (not published to github)

• Add google search for different types of establishments, in an area

• Add single radius search

• Add controller for search, that

  • check if search is single radius or between two locations
  • makes sure there are 5 results, and saves any additional searches to the csv file

• Deploy to google cloud functions - DONE

  • Modify storage location to a GCP cloud storage filepath - DONE

• Update IAM Authorisation - Should be Automatic when called from within gcp

To do list:

• Update data to cloud when user requests new data
• Content based recommender that classifies restaurants by type
• Create a list of types of restaurants, for recommendation
• System that links to accounts
• convert data management system to an object