URL Shortener built with flask & redis.
Project was done in 3 hours, but it still have improvements to be done
- Flask setup project
- Redis integration
- Allow client to short a URL
- Same long URL always is shortened to the same unique short URL
- Shortener algorithm (alphanumeric 8 digits, could be sha512)
- Allow client to send shortened URL and obtain the original long URL
- Redirection of the shortened URL to long URL
- Counting visits mechanism
- Swagger integration
- Fix redirect endpoint (not working in Swagger integration)
- Counting visits mechanism
- Better frontend with bootstrap UI + LOGO
- Workflow diagram
- Tests
- Deploy heroku
Before you start, check if your machine have the redis. If not you can pull redis through docker. You should have python 3.6+ installed.
To install the project, follow these steps:
- Setup redis locally on docker.
docker run -d -p 6379:6379 redis
If you have redis locally, this will conflict with. If you already have redis locally in your machine, please move to next step.
- Install dependencies
pip install -r requirements.txt
- Run server
python runserver.py
or you can run manually at
The server should be active on "localhost:5000"
They are available at (localhost:5000/apidocs)[localhost:5000/apidocs] as well.
| Description | Method | URL | Parameters |
|---|---|---|---|
| Shorten a long URL | GET | /shorten | url |
| Access a long URL from a short URL if exists | GET | /{short_id} | short_id |
| Access url details, long url, visits, etc | GET | /{short_id}/detail | short_id |
- Only alphanumeric characters are allowed when creating the short url, ie 62 characters. ("abcdefghijklmnopqrstuvwxyzABCDEFGHIJKLMNOPQRSTUVWXYZ0123456789")
- A pseudorandom number between 0 and 61 (total length) is generated which is used to get a character from the valid characters in step 1.
For example:
| Random | Character |
|---|---|
| 5 | strings[5] = "f" |
Such algorithm is repeated 8 times. Resulting in possible (62 * 8) unique shorten URL. Resulting in 496 unique URL coexisting inside the db.
--
Observations:
-
If a collision happens, the algorithm will try again. For the purpose of this challenge I have defined the length to be 8 characters, since I was not focused on collisions or performance here, but in a production environment we might use a different approach like a sha512 hash generatin for example.
-
I definitily would add some tests if this would go to production.
-
A health check endpoint would be nice too.
-
In a production environment, probably would add a dockerfile for the application. I did not focused on adding this since it was not a requirement.