First build the web app by running

Now, there are two options to run the application

Make sure that you have Docker installed on your machine then run:

After the image is built, and the server is running, the site can be accessed at:

Make sure that you have MongoDB available on your machine then start the database server in a separated terminal:

Then, start the server using the command

Now, the site ca be accessed at:

To test the server simply run

Please note that the criterias are taken from problem statement with serveral added tests.

Since in the real life situation, the graph (the delivery map) would be stored on the server not the client, it is important to have a database to store these structures. Thus, apart from the 3 use cases, the api must also support the get query to obtain all the graph options as well as the ability to post new graph. Since the web app does not require constant connection, simple API calls are utilized.

• GET the delivery cost of a specified route
• URL: /api/getRouteCost
• params:
  • graphId: the _id of the graph in the database
  • route: the nodes separated by a dash (e.g. A-B-E)
• returns:
  • The cost of following the given route. "No Such Route" if the route does not exists.

• GET the number of possible delivery routes with given conditions
• URL: /api/getNumberOfRoutes
• params:
  • graphId: the _id of the graph in the database
  • origin: the start node
  • endpoint: the destination
  • noRepeat: whether going through the same path twice is allowed
  • maxStop: (optional) the maximum number of stop allowed
  • maxCost: (optional) the maximum cost allowed
• returns:
  • The number of possible routes

• GET the cheapest delivery route between two towns
• URL: /api/getCheapestCost
• params:
  • graphId: the _id of the graph in the database
  • origin: the start node
  • endpoint: the destination
• returns:
  • The cost of the cheapest route. "No Such Route" if the route does not exists.
• Notes:
  • The cheapest cost is calculated using Djikstra's algorithm. To increase the efficiency, the cheapest cost to each node from the origin are stored after it is calculated. Thus, if the origin has already been used in a calculation, the server will simply respond with the stored cheapest cost.

• POST create a new graph
• URL: /api/createNewGraph
• params:
  • graph: the string representing the graph (e.g. AB1, AC4, AD10, BE3, CD4, CF2, DE1, EB3, EA2, FD1)
• returns:
  • The _id of the graph in the database (Create one if not already existed. If already exists, simply returns the _id)
• Notes:
  • The order of edges in the input does not matter.

• GET get all graphs
• URL: /api/getAllGraphs
• returns:
  • The array of _id and the string representing each graph.

If the case of more heavy traffics, caching responses could help improve the response time.