Makes long url ... short
- Create a randomly generated short ID for a given URL.
- If the same URL are provided twice, the short ID is not reused.
- The tool do not use the URL as a seed for the short ID creation.
- The short ID length is 7 and a lookup is done at the creation time to avoid collision, you can look at https://zelark.github.io/nano-id-cc/ using the alphabet with no look alike
23456789ABCDEFGHJKLMNPQRSTUVWXYZabcdefghijkmnpqrstwxyzto see the chance of ID collision based on utilisation. - The Short ID is generated from alphabetical and numeric characters only and look alike characters like
0/OorI/1are prohibited, see https://github.com/CyberAP/nanoid-dictionary/blob/master/nolookalikes.js - You can't short an URL from the tool domain (you can try to short https://shorter.now.sh/whatever to see the form error state).
- Clone this repo and run
make startat the root - Open http://localhost:9556
- Clone this repo and run
make startat the root - Run
yarn install - Run
yarn dev - Open http://localhost:3000
- use the serverless service now.sh to run the api.
- use the AWS free tiers for the mysql database.
- https://shorter.now.sh/
- 2 Lambda functions:
- Api endpoint to create and save the short URL
- Handler to redirect the short url to the original url
- Node.js
- Mysql
- Nanoid - https://github.com/ai/nanoid/
- Next.js
- React
- Robot - https://thisrobot.life/
- CSS - https://github.com/zeit/styled-jsx
- Environment and library security updates
- One could monitor security channels and put alert in place to update the libraries / environment.
- API floods
- One could implement some Rate limiting or captcha
- Database overflow.
- One could use database clusters.
- Too many request to the API (memory / CPU issue)
- The api functions are designed to be used in a serverless configuration, or in a cluster, so one could configure the infrastructure accordingly.