Wagtail：如何通过子页面字段筛选父页面模型？

Hey there! I’ve run into this exact scenario before with Wagtail’s page hierarchy. The key issue here is that Wagtail’s base Page model doesn’t know about your ChildPage-specific fields like is_completed—so we need to adjust our query to target the specific page type. Here are two solid approaches:

1. Using the children Relation with specific #

Wagtail provides a specific relation that lets you access fields from your custom page subclasses. We’ll combine this with filtering to target only ChildPage instances where is_completed is True:

# Get all ParentPages with at least one completed ChildPage child
parent_pages = ParentPage.objects.filter(
    children__model='your_app.childpage',  # Replace with your app + model name
    children__specific__is_completed=True
).distinct()

• The children__model filter ensures we’re only looking at child pages of type ChildPage (use the lowercase app_label.model_name format—you can get this from ChildPage._meta.label).
• distinct() is important because a single ParentPage might have multiple completed ChildPages, which would otherwise return duplicate ParentPage entries.

2. Using Subqueries (More Efficient) #

For larger datasets, subqueries are often faster because they avoid joining tables and deduplicating results. We’ll use Django’s Exists and OuterRef to check if a ParentPage has any matching ChildPages:

from django.db.models import Exists, OuterRef

parent_pages = ParentPage.objects.filter(
    Exists(
        ChildPage.objects.filter(
            parent=OuterRef('pk'),
            is_completed=True
        )
    )
)

This query directly checks for the existence of at least one completed ChildPage under each ParentPage, which is both clean and performant.

Why Your Initial Attempts Didn’t Work #

• The standard ParentPage.objects.filter(child_page__is_completed=True) fails because there’s no direct child_page foreign key relation—Wagtail uses a tree structure via Treebeard, so we need to use the children relation instead.
• ChildPage.objects.children_of([parent_ids]) returns ChildPages, not ParentPages, so we need to reverse the query to start from ParentPage and check their children.

内容的提问来源于stack exchange，提问作者MDalt