Entity parsing service for Stride/Hipchat.

This is a REST API that takes text messages, and parses out and returns emoticons, user mentions, and links.

It is designed with the assumption that most chat messages are relatively small, and uses regexes to do the parsing.
Scaling horizontally should be trivial: instances have no shared state (or state at all), so they can be spun up/down as needed.
I briefly thought about using AppEngine (or similar) to make scaling even simpler, but didn't want to burn too much of my 2 hours on configuration.

Future TODOs (ideally this would most be integrated with existing system):
  - A load balancer would need to handle routing between instances, and implement detection for things like queries of death, DDOS, etc.
  - Monitoring and logging:
    - QPS, Errors, Latency, Detection rate for each entity, Resource Usage, Uptime, Restarts, Binary Age
  - End-to-end Integration tests
  - Deployment configs
  - Load testing
  - A few smaller scale TODOs are scattered in the code


Usage:
  go build github.com/JarekSed/stride/server
  ./server


Example test request and response:
$  curl -X POST 'http://localhost:8000/entities' -d '{"Message": "@bob @john (success) such a cool feature; https://twitter.com/jdorfman/status/430511497475670016"}'
{"emoticons":["success"],"links":[{"url":"https://twitter.com/jdorfman/status/430511497475670016","title":"Justin Dorfman; on Twitter: \"nice @littlebigdetail from @HipChat (shows hex colors when pasted in chat). http://t.co/7cI6Gjy5pq\""}],"mentions":["bob","john"]}


Dependencies:
  - github.com/gorilla/mux: http routing
  - golang.org/x/net/html: html parsing for finding title of pages linked to
  - mvdan.cc/xurls: URL parsing