jubins / Spellchecker

Spellchecker is a custom API that helps determine if a word is spelled correctly. If not, the API returns some autocomplete suggestions of similar words. Built using Python Flask and Trie data structure.

Spellchecker

Spellchecker is a custom API that helps determine if a word is spelled correctly. If not, the API returns some autocomplete suggestions of similar words. Built using Python Flask and Trie data structure.

Project Setup

Dependencies

1. Python 3. Any version of Python 3 should work.
2. Docker/Docker-compose. If you do not have docker installed, follow instructions in Non-dockerized: Setting up the server.

Docker: Setting up the server

1. Make sure have docker/docker-compose installed and running, otherwise follow non-dockerized setup.
2. From your terminal go to main project directory i.e. the Spellchecker project directory that contains the docker-compose.yml file.
3. Type docker-compose up --build to start the server. You will see docker building device registry and starting the Flask server.
4. Go to your browser and copy-paste http://localhost:31337. If you see a welcome message then you've successfully setup the server and ready to start using the API.

Non-dockerized: Setting up the server

1. Create a virtual environment using below command.

   $ python3 -m venv venv

2. Activate your virtual environment.

   $ source/venv/bin/active

3. If you are on Windows or need more help with virtual environments follow this link.
4. From your terminal go to main project directory i.e. the Spellchecker project directory that contains the docker-compose.yml file.
5. Then go the spellchecker directory that contains requirements.txt.
6. Install all the requirements using below command and make sure your virtual environment is active.

   $ pip install requirements.txt

7. Start the Flask server using below command. You should see

   $ python3 api.py

8. Go to your browser and copy-paste http://localhost:31337. If you see a welcome message then you've successfully setup the server and ready to start using the API.

API Usage

As mentioned in the client requirements the endpoint:

• Is /spellcheck/{word}
• Runs on port 31337.
• Returns 404 if the word is not found
• Returns 200 if the word is found
  • The body of the response includes acceptable spelling suggestions if the word is misspelled.

Below are some examples of implementation, go to your browser and copy-paste below URL to give it a try or perform GET calls using Postman or any programming language.

 URL: http://localhost:31337/spellcheck/car
 
 Response Code: 200
 Response Body:
 {"suggestions": [], "correct": true}

 URL: http://localhost:31337/spell/caR
 
 Response Code: 200
 Response Body:
 {
     "suggestions": [
         "car",
         "carabao",
         "carabaos",
         "carabineer",
         ...
     ],
     "correct": false
 }

 URL: http://localhost:31337/spell/karaoke
 Response Code: 404
 Response Body:
 {"message": "Word: karaoke not found."}

Test cases

To run the test cases go to Spellchecker/spellchecker in a different terminal window and type below command. Make sure your Flask server is running.

 $ python3 tests.py 

If you get requests module import error make sure you install it using below command.

 $ pip3 install requests==2.25.0

Client requirements

Data

Client has provided a dictionary of acceptable words in dictionary.txt contained in this archive at Spellchecker/spellchecker/dictionary.txt.

Details on word conformity

The client has specified that word does not conform if it:

• It has repeating characters (i.e. balllooooon)
• It is missing one or more vowels (i.e. balln)
• It has mixed-casing (i.e. BAllOOn)
  • Note: “Hello” and “HELLO” are correct and should not be considered mixed-casing.
• It has a combination of a, b and c type errors (i.e. bllllLLlln)

Project structure

• Spellchecker/spellchecker/models.py file contains the code for checking word conformity and provides suggestions using Prefix Trie data structure.
• Spellchecker/spellchecker/validators.py handles API validation or any user errors.
• Spellchecker/spellchecker/api.py is the main entry point where API endpoints are defined and server can be started.

Contact

• Jubin Soni
• [e-mail]