Please do not fork this repository.

C# Rocks!

Welcome to the Wizkids C# test! The test is inspired by one of our products (AppWriter) which helps people with reading/writing problems by enabling them to read text out loud and get word suggestions (predictions) while typing.

Predictions are word completions based on the input text, for example: If the input text is I like ca the predictions could be cars, cats and cake.

The test requires you to create a small WPF project that implements the word prediction functionality with results from a custom dictionary and our word prediction web service.

If you encounter any unexpected issues please contact us at srs AT wizkids.dk and we will get back to you ASAP.

Scope

We expect a simply structured C# WPF application with good readability. Design and UX is not a major consideration - keeping it simple is fine.

We expect a well-versed developer to be able to complete the tasks within a few hours.

While a simple test project you should structure it as a project meant for a business production environment.

Code-behind tasks

1. Write a client for the word predictions web service.

2. Write a service for fetching words from the custom dictionary.

UI/XAML tasks

1. Write XAML code for displaying a text box and two lists of word prediction results (word prediction web service and custom dictionary).

2. Write handlers for fetching word predictions using both the web service client and custom dictionary service.

3. On keystrokes in the text box: Fetch word predictions based on the text value and display the results in the two lists.

• Display the first 10 results from each prediction source.
• The text value used for predictions text should be based on the position of the caret inside the text box, for example:

Hello, my name is J|ohn Doe and I like pie! 

The extracted text should be from the start of the text box up to the caret, which in this example would be Hello, my name is J.

4. Make your word predictions results clickable and insert the clicked prediction into the text box on click.

• The word prediction should be inserted into the text box at the location of the caret.

5. (OPTIONAL) Emphasize the part of each word prediction that has already been written.

For example, using the text from step #3 Hello, my name is J the letter J should be emphasized in the predictions John, James and June:

John
James
June

Miscellaneous tasks

1. Generate a token for use with the word predictions web service at https://services.lingapps.dk/misc/CSharpJSRocks.

2. Add your code in a Git project on GitHub and submit the link at https://services.lingapps.dk/misc/CSharpJSRocks.

(OPTIONAL, but strongly recommended) Add a link to one of your C# projects (in which you've been a major contributor) - Git repository preferred.

Custom dictionary description

The custom dictionary is implemented as an SQLite 3 database (Dictionary.db). It contains a Wordstable with the columns Id and Value.

The database is not encrypted/does not require a password.

If you have issues downloading the database file directly (GitHub some times display a 404 error), try cloning the project instead.

Word predictions web service description

URL

https://services.lingapps.dk/misc/getPredictions

Authorization

Header: Authorization: Bearer [TOKEN] (token generated in Miscellaneous task #1)

Parameters

• locale: String, the language in which the predictions should be computed. Valid values are da-DK and en-GB.
• text: String, the text from which to compute the predictions.

Response

JSON encoded array of strings ordered by prediction probability (descending), for example: ["cars", "cats", "cake"]

Example

Request

GET /misc/getPredictions?locale=en-GB&text=I%20like%20ca HTTP/1.1
Host: services.lingapps.dk
Authorization: Bearer MjAxOS0wMS0wMQ==.dGVzdEBleGFtcGxlLmNvbQ==.MjExMWMyYjdjZGY3YTU3MmU4MTA5OWY0MDgyMmM0OTk=
Connection: Close

Response

HTTP/1.1 200 OK
Content-Length: 34
Connection: close
Content-Type: application/json

[
    "cars",
    "cats",
    "cake"
]