NOTE

Written for a coding exercise. So kinda verbose in the comments and what not.  

MessageProcessor

Example code/app that will consume a chat message string and emit a json string containing info about specific content of the app.

It looks for

• mentions - Of the form @{userid}
• emoticons - Of the form ({emoticonname})
• http links - http://cnn.com

Output of MessageProcessorTests

Input: @bob @john (success) such a cool feature; https://twitter.com/jdorfman/status/430511497475670016
Output:
{
  "emoticons": [
    "success"
  ],
  "links": [
    {
      "title": "Justin Dorfman on Twitter: "nice @l",
      "url": "https://twitter.com/jdorfman/status/430511497475670016"
    }
  ],
  "mentions": [
    "bob",
    "john"
  ]
}

Architecture

• MessageProcessor - Main entry point. Uses a list of detectors to locate the items of interest.
• MPRegDetector - Base detector class, factory for creating simple regex detectors. Defines Detector protocol.
• LinkDetector - Special detector that uses the built in NSTextCheckingTypeLink detector. Conforms to our Detector protocol.
• String+MPUtils - Utils for the String class.
• HtmlPageParser - Only looks for titles. Not very fancy.
• SyncLinkExternalMapper - Partner class for the link detector that maps a list of links to link, title pairs.

Please see the source code for more detail. A good starting place is MessageProcessor.swift.

Building

1. install the latest version of cocoa pods that can handle Swift projects (currently this is 0.34.1).

2. install then into the project.

   pod install

3. open the workspace (not the project).

4. run the tests.

5. run the app.

App

The app is a universal iOS 8 project.

Note that the results come in at the top of the view.

Spec Notes

• All errors are suppressed and result in empty strings and arrays rather than nils.
• URLs with no title are returned with an empty string for the title.
• Titles are trimmed at 40 chars. Was not clear how to handle this from the spec so picked something.
• All json object entries are sorted.
• Close to i18n ready, but no tests for it, therefore likely bugs to be found.

Enhancements

• Emoticons and mentions likely need their own mappers to retrieve a resource ref.
• Incremental parsing could be provided with Visitor style blocks/closures that could be used while the user is entering the string.
• Mappers would use local caching/data stores to avoid latencies while user is entering text.
• Detectors and Mappers could be made concurrent. However given the input size and on-device usage it's debatable how useful this would be. Some testing would be useful here.

Spec

Take a chat message string and returns a JSON string containing information about its contents. Special content to look for includes:
1. @mentions - A way to mention a user. Always starts with an '@' and ends when hitting a non-word character. (http://help.hipchat.com/knowledgebase/articles/64429-how-do-mentions-work-)
2. Emoticons - For this exercise, you only need to consider 'custom' emoticons which are ASCII strings, no longer than 15 characters, contained in parenthesis. You can assume that anything matching this format is an emoticon. (http://hipchat-emoticons.nyh.name)
3. Links - Any URLs contained in the message, along with the page's title.

For example, calling your function with the following inputs should result in the corresponding return values.
Input: "@chris you around?"
Return (string):
{
  "mentions": [
    "chris"
  ]
}

Input: "Good morning! (megusta) (coffee)"
Return (string):
{
  "emoticons": [
    "megusta",
    "coffee"
  ]
}


Input: "Olympics are starting soon; http://www.nbcolympics.com"
Return (string):
{
  "links": [
    {
      "url": "http://www.nbcolympics.com",
      "title": "NBC Olympics | 2014 NBC Olympics in Sochi Russia"
    }
  ]
}


Input: "@bob @john (success) such a cool feature; https://twitter.com/jdorfman/status/430511497475670016"
Return (string):
{
  "mentions": [
    "bob",
    "john"
  ],
  "emoticons": [
    "success"
  ]
  "links": [
    {
      "url": "https://twitter.com/jdorfman/status/430511497475670016",
      "title": "Twitter / jdorfman: nice @littlebigdetail from ..."
    }
  ]
}