A web app that serves as an example of structuring an application with mount where business logic is separated from app's external resources.
$ cd neo
$ boot repl
boot.user=>
boot.user=> (dev)
#object[clojure.lang.Namespace 0x371d0df1 "dev"]
dev=> (reset)
INFO utils.logging - >> starting.. #'neo.conf/config
INFO utils.logging - >> starting.. #'neo.db/conn
INFO neo.db - creating a connection to datomic: datomic:mem://neo
INFO utils.logging - >> starting.. #'neo.www/neo-app
INFO utils.logging - >> starting.. #'neo.app/nrepl
{:started ["#'neo.conf/config" "#'neo.db/conn" "#'neo.www/neo-app" "#'neo.app/nrepl"]}
dev=>dev=> (stage-order-book conn orders)
({:db-before datomic.db.Db@3ff443a5, :db-after datomic.db.Db@dfffb0d... })via REPL:
dev=> (find-orders conn "GOOG")
({:db/id 17592186045418, :order/symbol "GOOG", :order/bid 665.51M, :order/qty 100, :order/offer 665.59M}
{:db/id 17592186045420, :order/symbol "GOOG", :order/bid 665.50M, :order/qty 300, :order/offer 665.58M}
{:db/id 17592186045422, :order/symbol "GOOG", :order/bid 665.48M, :order/qty 100, :order/offer 665.53M}
{:db/id 17592186045424, :order/symbol "GOOG", :order/bid 665.49M, :order/qty 200, :order/offer 665.52M})
dev=> (find-orders conn "TSLA")
({:db/id 17592186045426, :order/symbol "TSLA", :order/bid 232.38M, :order/qty 200, :order/offer 232.43M}
{:db/id 17592186045428, :order/symbol "TSLA", :order/bid 232.41M, :order/qty 100, :order/offer 232.46M})
dev=>via HTTP GET http://localhost:4242/neo/orders/GOOG
As of right now Neo is undergoing his initial training, and it will only return the total order quantity that was matched accross all the orders in the book.
All match requests do hit Neo's ultra low speed matching engine.
via REPL:
dev=> (def quote {:ticker "GOOG", :bid 665.54M, :offer 665.58M :qty 400})
#'dev/quote
dev=> (match-quote quote (find-orders conn "GOOG"))
300The quote's offer matched nothing, the bid matched two of the orders (their offers <= quote's bid) in the book:
{:db/id 17592186045422, :order/symbol "GOOG", :order/bid 665.48M, :order/qty 100, :order/offer 665.53M}
{:db/id 17592186045424, :order/symbol "GOOG", :order/bid 665.49M, :order/qty 200, :order/offer 665.52M})The quote's quantity is 400, 300 of which matched, hence total matched quantity is 300.
via HTTP GET http://localhost:4242/neo/match-quote?ticker=GOOG&qty=500&bid=665.54&offer=665.58:
At any given point Neo can be reloaded from the REPL:
dev=> (reset)
INFO utils.logging - << stopping.. #'neo.app/nrepl
INFO utils.logging - << stopping.. #'neo.www/neo-app
INFO utils.logging - << stopping.. #'neo.db/conn
INFO neo.db - disconnecting from datomic:mem://neo
INFO utils.logging - << stopping.. #'neo.conf/config
:reloading ()
INFO utils.logging - >> starting.. #'neo.conf/config
INFO neo.conf - loading config from dev/resources/config.edn
INFO utils.logging - >> starting.. #'neo.db/conn
INFO neo.db - conf: {:datomic {:uri datomic:mem://neo}, :www {:port 4242}, :h2 {:classname org.h2.Driver, :subprotocol h2, :subname jdbc:h2:mem:mount, :user sa, :password }, :rabbit {:api-port 15672, :password guest, :queue r-queue, :username guest, :port 5672, :node jabit, :exchange-type direct, :host 192.168.1.1, :vhost /captoman, :auto-delete-q? true, :routing-key , :exchange foo}, :nrepl {:host 0.0.0.0, :port 7878}}
INFO neo.db - creating a connection to datomic: datomic:mem://neo
INFO utils.logging - >> starting.. #'neo.www/neo-app
INFO utils.logging - >> starting.. #'neo.app/nrepl
{:started ["#'neo.conf/config" "#'neo.db/conn" "#'neo.www/neo-app" "#'neo.app/nrepl"]}
dev=>In case a namespace that was recompiled has a defstate in it, mount will clean it up and restart after the namespace was recompiled:
(more details on ns recompilation here)
Neo business logic does not have defstates and lives in its own gated neighborhood:
options/ book, engine and order live a full, happy referentially transparent life.
When you call (mount/start) or (reset), mount tells you what states were started:
dev=> (mount/start)
{:started ["#'neo.conf/config" "#'neo.db/conn" "#'neo.www/neo-app" "#'neo.app/nrepl"]}
dev=>You can easily correlate all these states to their homes:
It does not mean that each state should live in its separate namespaces, but it's always a good idea to start with that, since, at least in this example, those are different external resources.