pocketbase-orm 0.10.0

Add your description here

PocketBase ORM

A Python ORM (Object-Relational Mapper) for PocketBase that provides Pydantic model integration and automatic schema synchronization.

Features

• 🚀 Pydantic model integration for data validation and serialization
• 🔄 Automatic schema synchronization with PocketBase collections
• 📦 Support for most PocketBase field types including relations and file uploads
• 🛠️ Simple and intuitive API for CRUD operations

Installation

uv install pocketbase-orm

Quick Start

from pocketbase import PocketBase
from pocketbase_orm import PBModel
from pydantic import EmailStr, AnyUrl, Field
from datetime import datetime, timezone
from pocketbase.client import FileUpload

# Define your models
class RelatedModel(PBModel):
    name: str

    class Meta:
        collection_name = "related_models" # Optional

class Example(PBModel):
    text_field: str
    number_field: int
    is_active: bool
    url_field: AnyUrl
    created_at: datetime
    options: list[str]
    email_field: EmailStr | None = None
    related_model: RelatedModel | str = Field(..., description="Related model reference")
    image: FileUpload | str = Field(default=None, description="Image file upload")

# Initialize PocketBase client and bind it to the ORM
client = PocketBase("YOUR_POCKETBASE_URL")
client.admins.auth_with_password("[e-mail]", "password")
PBModel.bind_client(client)

# Sync collection schemas
RelatedModel.sync_collection()
Example.sync_collection()

# Create and save records
related_model = RelatedModel(name="Related Model")
related_model.save()

# Create a new record with file upload
with open("image.png", "rb") as f:
    example = Example(
        text_field="Test with image",
        number_field=123,
        is_active=True,
        email_field="[e-mail]",
        url_field="http://example.com",
        created_at=datetime.now(timezone.utc),
        options=["option1", "option2"],
        related_model=related_model.id,
        image=FileUpload(("image.png", f))
    )
    example.save()

# Query records
examples = Example.get_full_list()
single_example = Example.get_one("RECORD_ID")
filtered_example = Example.get_first_list_item(filter='email_field = "[e-mail]"')

# Get file contents
image_bytes = example.get_file_contents("image")

Model Definition

Models inherit from PBModel and use Pydantic field types:

from enum import Enum

class UserType(str, Enum):
    ADMIN = "admin"
    REGULAR = "regular"
    GUEST = "guest"

class MyModel(PBModel):
    name: str
    age: int
    email: EmailStr | None = None
    user_type: UserType  # Will be created as a select field in PocketBase

    class Meta:
        collection_name = "custom_collection_name"  # Optional

The collection name will be automatically derived from the class name (pluralized) if not specified in the Meta class.

Supported Field Types

• Text: str
• Number: int, float
• Boolean: bool
• Email: EmailStr
• URL: AnyUrl
• DateTime: datetime
• JSON: List, Dict
• File: FileUpload | str
• Relation: Union[RelatedModel, str]
• Select: Enum

API Reference

Class Methods

• bind_client(client: PocketBase): Bind PocketBase client to the model class
• sync_collection(): Create or update the collection schema in PocketBase
• delete_collection(): Delete the entire collection from PocketBase
• get_one(id: str, **kwargs) -> T: Get a single record by ID
• get_list(*args, **kwargs) -> List[T]: Get a paginated list of records
• get_full_list(*args, **kwargs) -> List[T]: Get all records
• get_first_list_item(*args, **kwargs) -> T: Get the first matching record
• delete_by_id(id: str, **kwargs): Delete a record by ID

Instance Methods

• save() -> T: Create or update the record
• delete(): Delete the current record
• get_file_contents(field: str) -> bytes: Get the contents of a file field

Limitations

• The ORM currently supports basic CRUD operations and schema synchronization
• Complex queries should use the PocketBase client directly
• Relationship handling is limited to single relations
• Indexes must be created manually

Contributing

Contributions are welcome! Please feel free to submit a Pull Request.

License

This project is licensed under the MIT License - see the LICENSE file for details.