Link Shortener

This is a link shortener application.

Getting Started

Start the application using one of the following methods.

Virtual Environment

1. Create virtual environment (python -m venv .venv) and activate it.
2. Install required packages (pip install requirements.txt).
3. Start the application (uvicorn app.main:app --reload --port 8080).

Docker

1. Build container (docker build . -t shortener)
2. Run contianer (docker run -p 8080:8080 shortener)

Routes

Base URL: http://localhost:8080.

Documentation

• /docs

Encode endpoint

• /v0/encode
• Example request:

curl -i -X POST \
   -H "Content-Type:application/json" \
   -d \
'{
  "url": "/some/url"
}' \
 'http://0.0.0.0:8080/v0/encode'

This is the endpoint for encoding a url. Provide a url to receive an encoded url and an optional value check_encoding_exists in the body to not store a URL twice.

Decode endpoint

• /v0/decode
• Example request:

curl -i -X POST \
   -H "Content-Type:application/json" \
   -d \
'{
  "encoded_url": "A",
  "base": "http://somebase"
}' \
 'http://localhost:8080/v0/decode'

This is endpoint for decoding a url. Provide an encoded_url to receive a decoded url and an optional value base to prepend to the decoded url.

Decisions

This section describes some architectural choices and mentions shortcuts.

Shortening algorithm

• Urls are stored with inrecementing ids.
• Ids are then encoded to base 66 (not my idea, based on some research on the internet)
• 66 Chars are used, which are either alphanumeric or unreserved for URLs according to RFC 3986

Shortcuts

• This application does not include logging and authentication/authorization.
• The encoding algorithm is not very fast, but can be improved by using extra space for quotient/remainder
• check_encoding_exists is an optional parameter, as looking for existing values is a slow operation

Testing

For running tests have a look at the "Getting started" section and either

• Run pip install requirements-test.txt to install additional test dependencies or
• Build a testing container with docker build -f Dockerfile.tests . -t shortener-tests
• Run the tests with docker run shortener-tests

Linting

Ruff was used for linting. To lint the project:

1. Install ruff pip install ruff
2. Lint, f.e. with ruff check --fix .