Not gonna lie I learned SQL much faster on the job than practicing on my own.
But if you really want to learn SQL on your own, then push beyond writing basic SQL queries (SELECT/JOIN/WHERE/HAVING/Window functions). Do the following:
1. Set up your own sql database
2. Learn procedural extensions to SQL (for Oracle, learn PL/SQL. For Microsoft learn T-SQL. And so on).
3. Learn to use those procedural extensions to create/update tables, write functions in SQL, check for errors, and log events that happen in your database
4. Learn query optimization for your specific SQL type. Memory management, making queries efficient and faster, and so on
5. Integration with other languages: can you feed your database off of web data? Can you have your database push daily files out for further data processing in Python or another language?
6. Traditional data engineering works off of the ETL model (extraction of data from original source, transformation of raw data into needed information, loading information into database for storage and use). But with the explosion of big data and the availability of cheap storage, people are moving to the ELT model. Can you use tools within and without your database to be comfortable in the new model?
If you learn all that on your own, you'll be sorted for any basic SQL job. If you just focus on the basics on your own, you'll be overwhelmed in the real world. Just my opinion.
