Challenge 3: CRUD in the Django Shell — Possible Solution
====================================================================

$ python manage.py shell

>>> from catalog.models import Book

# CREATE
>>> book = Book.objects.create(
...     title="The Left Hand of Darkness",
...     description="An envoy visits a wintry planet whose inhabitants have no fixed gender.",
...     price=12.99,
...     in_stock=True,
... )
>>> book.pk
1

# READ
>>> fetched = Book.objects.get(pk=1)
>>> fetched.title
'The Left Hand of Darkness'

# UPDATE
>>> fetched.price = 9.99
>>> fetched.save()
>>> Book.objects.get(pk=1).price
Decimal('9.99')

# DELETE
>>> fetched.delete()
(1, {'catalog.Book': 1})

>>> Book.objects.filter(pk=1).exists()
False

WHY THIS WORKS
--------------
- Book.objects.create(...) is a shortcut that both builds a new Book
  instance AND saves it to the database in one call — equivalent to
  writing `book = Book(...)` followed by `book.save()`, just more concise
  for the common "create and immediately persist" case.
- Book.objects.get(pk=1) is safe here (per this chapter's gotcha) because
  a primary key lookup is guaranteed to return at most one row — no risk
  of DoesNotExist or MultipleObjectsReturned as long as that pk exists.
- Updating a field on an already-fetched instance and calling .save()
  is how the ORM handles updates — there's no separate "update" method;
  you modify the Python object's attributes and persist the whole object
  back with .save(). Django only sends the changed field(s) as part of
  the underlying UPDATE statement.
- fetched.delete() removes the row from the database and returns a tuple
  summarizing what was deleted (useful when a delete cascades to related
  objects, covered properly in the next chapter on relationships) — after
  this call, the Python object still exists in memory but no longer
  corresponds to any row.
- Book.objects.filter(pk=1).exists() is a cheap way to confirm deletion
  without needing to catch a DoesNotExist exception from .get() —
  .exists() runs an efficient existence check in SQL rather than
  fetching the whole row.