Fix BookListSerializer update example logic to support creation

[JPisOP007]

Description

This PR fixes a bug in the documentation example for BookListSerializer.update() under the ListSerializer section of the serializers API guide.

Problem

The current documentation example crashes with a KeyError if list items are submitted without an id field. This occurs when trying to perform creations alongside updates via a ListSerializer.

Solution

1. Added list comprehensions to separate the list items:
   • Filtered data_mapping to only include validated data items containing an 'id' key (updates/deletes).
   • Filtered out new_items that do not contain an 'id' key (creations).
2. Iterated over new_items and called .create(data) on the child serializer (self.child.create(data)).
3. Updated the BookSerializer definition in the example to use id = serializers.IntegerField(required=False). This ensures validation succeeds when creating new items that do not yet have an ID.

refs #9469

[ulugbekbackend]

Logic looks correct overall — nice catch on the KeyError that happens when creating list items without an id.
A couple of small suggestions:

Avoid iterating validated_data twice. Right now data_mapping and new_items are built with two separate comprehensions, each looping over the full list. This could be combined into a single loop:

data_mapping = {}
new_items = []
for item in validated_data:
    if 'id' in item:
        data_mapping[item['id']] = item
    else:
        new_items.append(item)

This keeps the same result but only goes through validated_data once.

Is there a docs test covering this example? DRF sometimes has tests that verify code examples in the docs actually run correctly. If one exists for this section, it might need updating too — worth checking.

Otherwise this is a solid fix for a real bug in the documentation. Thanks for working on it!

[ulugbekbackend]

Overall logic is solid — good fix for the KeyError on creation without an id. Left a couple of inline notes.

[ulugbekbackend]

Nit: this could be a single loop instead of two separate comprehensions, to avoid iterating validated_data twice — e.g.:

data_mapping = {}
new_items = []
for item in validated_data:
    if 'id' in item:
        data_mapping[item['id']] = item
    else:
        new_items.append(item)

[ulugbekbackend]

Makes sense given the new create path.