Two Scoops of Django Study - 5

We studied Django Queries and the Database Layer, focusing on Chapter 7 of Two Scoops of Django. Hyunju summarized and explained the core of the chapter, and Hoseong prepared and presented advanced material.

The core content covered the roles of the Django ORM and Model Manager, and the point that for detail pages we should use get_object_or_404() instead of get(), and only in views. Queries that could raise exceptions should be handled with try-except, and using lazy evaluation—deferring SQL execution until the data is actually needed—improves code readability and maintainability.

As advanced query tools, we introduced Query Expressions and Database Functions, and concluded that they can be used to improve both safety and performance. Conversely, unless there's a specific reason, raw SQL should be avoided, and applying an index as needed is recommended.

In the transactions section, we looked at the approach of treating each HTTP request as a transaction.

• Explicit transaction declarations
• The relationship between django.http.StreamingHttpResponse and transactions
• How to use the django.db.transaction.non_atomic_requests decorator to take specific views out of the transaction

In the deep-dive session, we covered ObjectDoesNotExist vs DoesNotExist, iterator() vs all(), how DB indexes work, checking query time with EXPLAIN, and Transaction ACID. Since we've usually validated and extracted values using IPython notebook and the Django ORM, some of this was familiar, but it was especially helpful to pin down precisely what we'd only vaguely understood about db_index and Transaction ACID.

Overall, we wrapped up by reorganizing our understanding of Django DB features commonly used in practice, resolving to use this as a chance to accelerate our own growth.