To run app on your local machine, make sure the you have installed the following software:
Start app
cd <project_dir>
docker-compose up
The API is available on port 8000.
You should wait a little bit (depends on inet speed and PC performance) until the cache fills up (you'll see the log message).
You can use Postman, curl in your terminal or any other HTTP client to test search endpoint
Example:
curl localhost:8000/search/Canon
- Flask: API
- Celery (beat mode): periodic tasks
- Redis: cache storage and broker for celery-beat.
The API returns list of image objects. Fields are:
- 'id' and 'cropped_picture': in any case;
- 'tags', 'file_name', 'author', 'camera': if there is a match.
- If the "schedule" time of the "reload_cache" task is less than the execution time of this task, the tasks will accumulate in the execution stack
This app describes only the simple process of caching data from external API and searching data using this cache. Some cases wasn't handled such as:
- merging several matching tags for one image in API response (if such matching occurred);
- controlling performance in case of large amount of data (hundreds of millions records) in cache
- sensitive data encryption/hiding
- running services as root etc
In addition:
- only basic API error handling are implemented (by means of logs)