Speller

Speller is a .NET Core application that implements symmetric spelling algorithm.

Installation

Use the .NET Core 2.2+ to execute this application.

First you should restore all packages which are used by application. Go to the application root and execute.

dotnet restore

Now, you should build your application.

dotnet build

If you would like to execute the Web API application, go to the Speller.Presentation.Web.Api project and execute. It will create a self-hosted application.

dotnet run

Usage

The console application is static. To use it, just run the application.

If you would like to use the Web API application, execute it (check the installation section). Using Postman you can send data to the API.

By default, the API is on /api/speller.

Using REST, the request will be something like below.

POST /api/speller HTTP/1.1
Host: localhost:32768
Content-Type: application/json
{
    "dictionary": [
        "love",
        "kid",
        "card"
    ],
    "words": [
        "ove",
        "car"
    ]
}

In dictionary array, you are supposed to send the destination words, that is, the words the algorithm will "search".

In words array, you are supposed to send the origin words, that is, the words the algorithm will check and validate.

The response for that request will be...

[
    "love",
    "card"
]

Azure Machine Learning support

This application is prepared to be used side-by-side with a Azure Machine Learning application. When the algorithm do not know the correction of a specific word, it will add the word in a list and, after the execution, it will send the unknown words to the Machine Learning. The return of Machine Learning Endpoint will replace the words in the original list.

You can have many Machine Learning Endpoints in one application. To add them, use the appsettings.json file. In the MachineLearningEndpointSettings section, there are two sample objects that you must add endpoint URL and API key.

To use your Azure Machine Learning Endpoint, you must set the machineLearning parameter in the request's URL. The value of this parameter must be the name of endpoint configured in appsettings.json.

For example, we have an API called GenreByName configured.

"MachineLearningEndpointSettings": {
    "GenreByName": {
        "Url":  "https://machinelearning.micro$oft.com/api/endpoint",
        "ApiKey":  "1234567abcd=="
    }
}

So, if you would like to use the Machine Learning like I said, you should call the API in the following manner.

/api/speller?machineLearning=genrebyname

If you do not want to use Machine Learning, do not put the machineLearning parameter in URL.

Docker support

This project is ready for be executed in (Linux) Docker containers. The provided Dockerfile will create the machine and exposes it by port 80.

Contributing

Pull requests are welcome. For major changes, please open an issue first to discuss what you would like to change.

Please make sure to update tests as appropriate.

License

MIT