Zendesk Search App

This app was built using Ruby 2.6.3.

Usage

Clone the repo:

foo@bar:~$ git clone https://github.com/factcondenser/zendesk_search.git

cd into the app's root directory:

foo@bar:~$ cd zendesk_search

Run the tests to make sure everything's working correctly:

foo@bar:~/zendesk_search$ rake test

Run the app:

foo@bar:~/zendesk_search$ ruby zendesk_search.rb

The usage should be fairly straightforward from there.

Additional Notes

It's possible to make additional data searchable by this app. users.json, tickets.json, and organizations.json are the three datafiles searchable by default, but more can be added. As an example, I've included movies.json in the db folder under the app's root directory. To make the data in movies.json searchable, make the following changes to zd_search_policy.rb:

module ZdSearchPolicy
  INPUT_MAP = {
    'main' => {
      '1' => 'search',
      '2' => 'list_fields'
    },
    'categories' => {
      '1' => 'users',
      '2' => 'tickets',
-     '3' => 'organizations'
+ .   '3' => 'organizations',
+     '4' => 'movies'
    }
  }.freeze
  
  ...
end

The data in movies.json was pulled from Wikipedia.

Known Issues

Arrays and nested objects

The app is naive in its treatment of arrays and nested objects in the JSON it searches. Searchable fields are only those properties located on the first level of the JSON objects provided by the db folder's JSON files. The app assumes that all objects in a given JSON file have the same properties. Searching on a field that has an array as a value is possible, but, as with every other searchable field, only exact matches will turn up results (e.g. an organization for which 'tags' equals '["Fulton", "West", "Rodriguez", "Farley"]' will match a search value of exactly '["Fulton", "West", "Rodriguez", "Farley"]', but will not match 'Fulton' or '["West", "Fulton", "Rodriguez", "Farley"]').

Coupling between category names and JSON file names

The app expects each category name to match the name of a JSON file in db (e.g. 'movies' matches 'db/movies.json'). Not only is this brittle, but due to the way category names currently get displayed, it's also pretty easy to end up with some ugly category options (e.g. 'db/pull_requests.json' would need a matching 'pull_requests' category name. In the app, this would get displayed as 'Pull_requests').

Odd output format in test fixtures

The test fixtures containing the output to STDOUT from running through various scenarios with the app (test/fixtures/*_output.txt) all list the values of the 'user' inputs (mocked by piping in strings from one of test/fixtures/*_input.txt) before everything else. For some reason I haven't figured out yet, that's the way the output for those scenarios consistently looks. For example, running

foo@bar:~/zendesk_search$ cat test/fixtures/quit_input.txt | ruby zendesk_search.rb --environment test

outputs

Welcome to Zendesk Search!
Type 'q' or 'quit' to exit at any time

========== SEARCH OPTIONS ==========
1) Search Zendesk
2) View a list of searchable fields

Please enter your choice: quit

Thank you for using Zendesk Search!

but running that same line in the test file using backticks

`cat test/fixtures/quit_input.txt | ruby zendesk_search.rb --environment test`

outputs

Please enter your choice: quit
Welcome to Zendesk Search!
Type 'q' or 'quit' to exit at any time

========== SEARCH OPTIONS ==========
1) Search Zendesk
2) View a list of searchable fields

Thank you for using Zendesk Search!

Notice how 'Please enter your choice: quit' appears at the top of the latter output.

Feedback

I'm always looking for ways to improve, so feel free to email me. Any feedback is welcome!