Add nested paths

[mirza99]

Add the possibility to have nested paths like :

// Routes config
const routes = {
    persons: {
        host: 'https://wwww.google.com',
        path: '/persons',
        findById: {
            path: "/{id}",
            getSettings: {
              path: '/settings'
            }
        }
    }
}

Use
routing.generate('persons.findById.getSettings', {id: xxx})

Generates
https://www.google.com/persons/xxx/settings

[ArvsIndrarys]

This solution would consist of two steps :

• create a pre-cache that would generate a map of all possible routes and store it permanently in a file. The update would be done at a specific call (parse? update? generate?)

routing.generate()
[users.generateKey] => users.{userId}.crypto.generateKey

• At each call in the program (generate? apply? getRoute?), routing would read in the file to get the correct path, then apply the parameters

routing.getRoute(users.generateKey, {userId: xxx})
users.{userId}.crypto.generateKey => http://www.example.com/users/xxx/crypto/generateKey

[mirza99]

Yes, good idea. It would improve performance.

By nested, we meant for example the following route file:

// route file
module.exports = {
    host: "http://google.com",

    persons: {
        path: '/persons',

        findAll: {
            host: 'http://bing.com',
            path: "/list",
           findAllByAge: {
             path: "/age/{age}",
           }
        },
        findById: {
            path: "/{id}",
        }
    }
}

routing.getRoute('persons.findAllByAge', {params: {age: 20}}) // creates http://bing.com/persons/list/age/20

[ArvsIndrarys]

We could avoid making a breaking change in our code by just keeping the generate() function name 'client-side'.

As the user already setups the routing with const routing = new Routing();, the constructor now generates a cache with the listing of all routes following this form :

{ 
  persons: 'persons',
  'persons.findById': 'persons.findById',
  'persons.getName': 'persons.findById.getName',
  'persons.getSecondLetter': 'persons.findById.getName.getSecondLetter',
  'persons.findEverywhere': 'persons.findEverywhere',
  notPersons: 'notPersons',
  'notPersons.findById': 'notPersons.findById',
  'notPersons.getName': 'notPersons.findById.getName',
  'notPersons.getSecondLetter': 'notPersons.findById.getName.getSecondLetter',
  'notPersons.testDoubleSlashes': 'notPersons.findById.getName.testDoubleSlashes',
  'notPersons.findAny': 'notPersons.findAny' 
}

The user can use it as is, or adds the path for the cache (the docs will be updaded too):
const routing = new Routing('/temp/cache.dat');

This allows generate to work as the initial version did, by parsing from the reduced given path (e.g: notPersons.getSecondLetter) the full path (notPersons.findById.getName.getSecondLetter), generating and replacing the parameters if needed.

The best way of enhancing it now would be to already register the path so as the file would be only parsed once :
'persons.getSecondLetter': 'http://yahoo.com/persons/{id}/name/{name}/secondLetter'