Post Processing

中文版

resolve_* loads missing data. post_* runs after the current subtree is fully assembled — it transforms, aggregates, or formats data that is already in place.

Extend the Same Sprint Example

Now that Sprint -> Task -> User can already be resolved, we can derive two fields from the finished subtree:

task_count
contributor_names

import asyncio
from typing import Optional

from pydantic import BaseModel
from pydantic_resolve import Loader, Resolver, build_list, build_object


# --- Fake database ---
USERS = {
    7: {"id": 7, "name": "Ada"},
    8: {"id": 8, "name": "Bob"},
    9: {"id": 9, "name": "Cara"},
}

TASKS = [
    {"id": 10, "title": "Design docs", "sprint_id": 1, "owner_id": 7},
    {"id": 11, "title": "Refine examples", "sprint_id": 1, "owner_id": 8},
    {"id": 12, "title": "Write tests", "sprint_id": 1, "owner_id": 7},
]


async def user_loader(user_ids: list[int]):
    users = [USERS.get(uid) for uid in user_ids]
    return build_object(users, user_ids, lambda u: u.id)


async def task_loader(sprint_ids: list[int]):
    tasks = [t for t in TASKS if t["sprint_id"] in sprint_ids]
    return build_list(tasks, sprint_ids, lambda t: t["sprint_id"])


class UserView(BaseModel):
    id: int
    name: str


class TaskView(BaseModel):
    id: int
    title: str
    owner_id: int
    owner: Optional[UserView] = None

    def resolve_owner(self, loader=Loader(user_loader)):
        return loader.load(self.owner_id)


class SprintView(BaseModel):
    id: int
    name: str
    tasks: list[TaskView] = []
    task_count: int = 0
    contributor_names: list[str] = []

    def resolve_tasks(self, loader=Loader(task_loader)):
        return loader.load(self.id)

    def post_task_count(self):
        return len(self.tasks)

    def post_contributor_names(self):
        return sorted({task.owner.name for task in self.tasks if task.owner})


# --- Resolve ---
raw_sprints = [{"id": 1, "name": "Sprint 24"}]
sprints = [SprintView.model_validate(s) for s in raw_sprints]
sprints = await Resolver().resolve(sprints)

print(sprints[0].model_dump())
# {'id': 1, 'name': 'Sprint 24',
#  'tasks': [
#      {'id': 10, 'title': 'Design docs', 'owner_id': 7, 'owner': {'id': 7, 'name': 'Ada'}},
#      {'id': 11, 'title': 'Refine examples', 'owner_id': 8, 'owner': {'id': 8, 'name': 'Bob'}},
#      {'id': 12, 'title': 'Write tests', 'owner_id': 7, 'owner': {'id': 7, 'name': 'Ada'}},
#  ],
#  'task_count': 3,
#  'contributor_names': ['Ada', 'Bob']}

Execution Order

For one sprint, the lifecycle looks like this:

flowchart LR
    a["resolve_tasks"] --> b["TaskView.resolve_owner"]
    b --> c["post_task_count"]
    c --> d["post_contributor_names"]

The exact implementation can be async under the hood, but the mental model is simple:

Load descendants first (all resolve_* methods).
Run post_* only after descendant data is ready.

That timing is why post_* is ideal for summary fields, formatting, and business-specific derived values.

Rule of Thumb

Question	`resolve_*`	`post_*`
Needs external IO?	Yes	Usually no
Runs before descendants are ready?	Yes	No
Good for counts, labels, formatting?	Sometimes	Yes
Return value is resolved again?	Yes	No

What `post_*` Is Good At

Typical uses include:

Counts and totals: task_count, total_price, unread_count
Sorted display fields: contributor_names, sorted_tags
Deduplicated labels: unique_categories
String formatting: full_title, display_name
Business summaries that should not require another query

Example: Formatting

class TaskView(BaseModel):
    id: int
    title: str
    priority: int
    priority_label: str = ""

    def post_priority_label(self):
        labels = {1: "Low", 2: "Medium", 3: "High"}
        return labels.get(self.priority, "Unknown")

Example: Enrichment from Nested Data

class SprintView(BaseModel):
    id: int
    name: str
    tasks: list[TaskView] = []
    has_overdue: bool = False

    def resolve_tasks(self, loader=Loader(task_loader)):
        return loader.load(self.id)

    def post_has_overdue(self):
        return any(t.due_date < date.today() for t in self.tasks)

Example: Aggregation

class OrderView(BaseModel):
    id: int
    items: list[OrderItem] = []
    total: float = 0.0

    def resolve_items(self, loader=Loader(item_loader)):
        return loader.load(self.id)

    def post_total(self):
        return sum(item.price * item.quantity for item in self.items)

What `post_*` Is Not For

Prefer keeping data loading in resolve_* and data transformation in post_*. That separation keeps the code readable:

resolve_* answers: where does the missing data come from?
post_* answers: what do we do with the data after it is ready?

# GOOD: load in resolve_*, transform in post_*
def resolve_owner(self, loader=Loader(user_loader)):
    return loader.load(self.owner_id)

def post_owner_display(self):
    return f"{self.owner.name} ({self.owner.email})"

If a field's primary purpose is to load external data, resolve_* is the right place — even though post_* technically supports loader (see below).

post_* Parameters

post_* methods can accept additional parameters beyond the basic form:

context

Access the global context dict passed to Resolver:

class SprintView(BaseModel):
    tasks: list[TaskView] = []
    visible_task_count: int = 0

    def resolve_tasks(self, loader=Loader(task_loader)):
        return loader.load(self.id)

    def post_visible_task_count(self, context):
        user_role = context.get('role', 'viewer')
        if user_role == 'admin':
            return len(self.tasks)
        return len([t for t in self.tasks if t.visible])

parent

Access the direct parent node. Useful for tree structures:

class TreeNode(BaseModel):
    name: str
    children: list[TreeNode] = []
    depth: int = 0

    def post_depth(self, parent):
        if parent is None:
            return 0
        return parent.depth + 1

ancestor_context

Access data exposed by ancestors via ExposeAs (covered in Cross-Layer Data Flow):

class TaskView(BaseModel):
    title: str
    full_title: str = ""

    def post_full_title(self, ancestor_context):
        sprint_name = ancestor_context.get('sprint_name', '')
        return f"{sprint_name} / {self.title}"

collector

Collect data from descendant nodes via SendTo (covered in Cross-Layer Data Flow):

class SprintView(BaseModel):
    tasks: list[TaskView] = []
    contributors: list[UserView] = []

    def resolve_tasks(self, loader=Loader(task_loader)):
        return loader.load(self.id)

    def post_contributors(self, collector=Collector('contributors')):
        return collector.values()

loader

post_* also accepts Loader — the same parameter used in resolve_*. This is a valid escape hatch, but avoid it unless you have a clear reason.

Two things to keep in mind:

Data loaded in post_* is not resolved recursively. If the loaded result contains resolve_* or post_* methods, they will not run.
The loaded data arrives late — other post_* methods on the same object cannot depend on it.

Typical legitimate use: loading supplemental data where the load key itself comes from a resolved field.

class TaskView(BaseModel):
    id: int
    title: str
    owner_id: int
    owner: Optional[UserView] = None
    department_name: str = ""

    def resolve_owner(self, loader=Loader(user_loader)):
        return loader.load(self.owner_id)

    def post_department_name(self, loader=Loader(department_loader)):
        # owner.department_id is only available after resolve_owner
        if self.owner:
            return loader.load(self.owner.department_id)

post_default_handler

A special post method that runs after all other post_* methods. It does not do automatic assignment — you must set fields manually:

class SprintView(BaseModel):
    tasks: list[TaskView] = []
    task_count: int = 0
    summary: str = ""

    def post_task_count(self):
        return len(self.tasks)

    def post_default_handler(self):
        # runs after post_task_count
        self.summary = f"{self.task_count} tasks in this sprint"

A Useful Boundary

post_* can accept advanced parameters such as context, parent, ancestor_context, and collector. But those features are easier to understand after the basic timing model is already clear.

When to Stop Here

The resolve_* + post_* combination covers the majority of data assembly needs. Most endpoints never need more than this.

Move on to Cross-Layer Data Flow only when you encounter parent-child coordination that manual traversal cannot handle cleanly.