mongospec 0.1.2

Async MongoDB ODM with msgspec integration and automatic collection binding

mongospec

Minimal async MongoDB ODM built for speed and simplicity, featuring automatic collection binding and msgspec integration.

Table of Contents

• Installation
• Key Features
• Quick Start
• Core Concepts
  • Document Models
  • Connection Management
  • Collection Binding
  • Index Creation
• CRUD Operations
  • Creating Documents
  • Reading Documents
  • Updating Documents
  • Deleting Documents
  • Counting Documents
• Advanced Usage
  • Working with Cursors
  • Batch Operations
  • Atomic Updates
  • Upsert Operations
• Performance Considerations
• API Reference
• Development Status

Installation

You can install mongospec from PyPI:

pip install mongospec

Dependencies:

• Python 3.13+
• mongojet ~= 0.3.1
• msgspec ~= 0.19.0

Key Features

• ⚡ Blazing fast - Uses msgspec (fastest Python serialization) and mongojet (fastest async MongoDB wrapper)
• 🧩 Dead simple - No complex abstractions, just clean document handling
• 🏎️ Zero overhead - Optimized for performance-critical applications
• 🔄 Async first - Built from the ground up for asyncio
• 🧬 Type-safe - Full typing support for better IDE integration and code safety

Quick Start

Here's a minimal example to get you started:

import asyncio
from datetime import datetime

import mongojet
import msgspec

import mongospec
from mongospec import MongoDocument


class User(MongoDocument):
    __collection_name__ = "users"
    __indexes__ = [{"keys": [("email", 1)], "options": {"unique": True}}]

    name: str
    email: str
    created_at: datetime = msgspec.field(default_factory=datetime.now)


async def main():
    # Connect to MongoDB
    mongo_client = await mongojet.create_client("mongodb://localhost:27017")
    try:
        # Initialize mongospec with our models
        await mongospec.init(mongo_client.get_database("db"), document_types=[User])

        # Create a new user
        user = User(name="Alice", email="[e-mail]")
        await user.insert()
        print(f"Created user with ID: {user._id}")

        # Find the user
        found_user = await User.find_one({"email": "[e-mail]"})
        print(f"Found user: {found_user.name} ({found_user.email})")

        # Update the user
        found_user.name = "Alice Smith"
        await found_user.save()
        print(f"Updated user name to: {found_user.name}")

        # Delete the user
        await found_user.delete()
        print("User deleted")
    finally:
        # Always close the connection when done
        await mongospec.close()


asyncio.run(main())

Core Concepts

Document Models

Document models in mongospec are based on msgspec.Struct and define the schema for your MongoDB collections. Each model automatically maps to a collection and provides methods for CRUD operations.

from datetime import datetime
from typing import Optional, ClassVar, List, Dict, Any

import msgspec
from bson import ObjectId
from mongojet import IndexModel

from mongospec import MongoDocument


class Product(MongoDocument):
    # Custom collection name (optional, defaults to class name)
    __collection_name__ = "products"

    # MongoDB indexes to create
    __indexes__: ClassVar[List[Dict[str, Any]]] = [
        {"keys": [("sku", 1)], "options": {"unique": True}}
    ]

    # Document fields (all typed)
    name: str
    price: float
    description: Optional[str] = None
    sku: str
    in_stock: bool = True
    created_at: datetime = msgspec.field(default_factory=datetime.now)
    updated_at: Optional[datetime] = None

    # The _id field is already defined in MongoDocument

Connection Management

mongospec handles MongoDB connections through a simple interface:

import mongojet
import mongospec

# Create a mongojet client
mongo_client = await mongojet.create_client("mongodb://localhost:27017")
db = mongo_client.get_database("your_database")

# Initialize mongospec with your database
await mongospec.init(db, document_types=[User, Product, Order])

# Always close connections when shutting down
await mongospec.close()

Collection Binding

When you initialize mongospec with document types, it:

1. Binds each document type to its corresponding collection
2. Creates any defined indexes
3. Makes all CRUD operations immediately available

# Initialize mongospec with your models
await mongospec.init(db, document_types=[User, Product])

# After initialization, you can use all CRUD methods
user = User(name="John", email="[e-mail]")
await user.insert()  # Collection is already bound

Index Creation

Define indexes on your model using the __indexes__ class variable:

class User(MongoDocument):
    __collection_name__ = "users"
    __indexes__ = [
        # Simple unique index
        {"keys": [("email", 1)], "options": {"unique": True}},

        # Compound index
        {"keys": [("last_name", 1), ("first_name", 1)], "options": {}},

        # Text index
        {"keys": [("description", "text")], "options": {"weights": {"title": 10, "description": 5}}}
    ]

    first_name: str
    last_name: str
    email: str
    description: str = ""

The indexes are automatically created when you call mongospec.init() with your document types.

CRUD Operations

Creating Documents

mongospec provides several ways to create documents:

# Create a single document
user = User(name="Alice", email="[e-mail]")
await user.insert()

# Alternative approach using class method
user = User(name="Bob", email="[e-mail]")
await User.insert_one(user)

# Batch insert multiple documents
users = [
    User(name="Charlie", email="[e-mail]"),
    User(name="Dave", email="[e-mail]")
]
inserted_ids = await User.insert_many(users)

# Conditional insert (only if not exists)
user = User(name="Eve", email="[e-mail]")
result = await User.insert_if_not_exists(
    user,
    filter={"email": "[e-mail]"}
)
if result:
    print("User was inserted")
else:
    print("User with this email already exists")

Reading Documents

Find documents with various query options:

# Find by ID
user_id = "662a3b4c1f94c72a88123456"  # Can be string or ObjectId
user = await User.find_by_id(user_id)

# Find by query
user = await User.find_one({"email": "[e-mail]"})

# Find multiple documents
cursor = await User.find({"age": {"$gt": 30}})
async for user in cursor:
    print(f"Found user: {user.name}")

# Get all documents in a collection (use with caution for large collections)
all_users = await User.find_all()

# Check if document exists
if await User.exists({"email": "[e-mail]"}):
    print("User exists")

Updating Documents

Update documents with various strategies:

# Find and modify approach
user = await User.find_one({"email": "[e-mail]"})
if user:
    user.name = "Alice Jones"
    await user.save()  # Full document replacement

# Direct update with operators
await User.update_one(
    {"email": "[e-mail]"},
    {"$set": {"status": "active"}, "$inc": {"login_count": 1}}
)

# Update multiple documents
modified_count = await User.update_many(
    {"status": "pending"},
    {"$set": {"status": "active"}}
)
print(f"Activated {modified_count} users")

# Update by ID with operators
await User.update_by_id(
    "662a3b4c1f94c72a88123456",
    {"$set": {"verified": True}}
)

# Atomic find-and-modify operation
updated_user = await User.find_one_and_update(
    {"_id": user_id, "version": current_version},
    {"$set": {"data": new_data}, "$inc": {"version": 1}},
    return_updated=True
)

Deleting Documents

Delete documents with various methods:

# Delete a document instance
user = await User.find_one({"email": "[e-mail]"})
if user:
    await user.delete()

# Delete by query
deleted_count = await User.delete_one({"email": "[e-mail]"})

# Delete by ID
deleted_count = await User.delete_by_id("662a3b4c1f94c72a88123456")

# Bulk delete
deleted_count = await User.delete_many({"status": "inactive"})
print(f"Deleted {deleted_count} inactive users")

Counting Documents

Count documents with different methods:

# Count with filter
active_count = await User.count_documents({"status": "active"})

# Estimated count (faster but approximate)
total_count = await User.estimated_document_count()

# Alternative count method
pending_count = await User.count({"status": "pending"})

Advanced Usage

Working with Cursors

For large result sets, use cursors for memory-efficient processing:

# Create a cursor with optional batch size
cursor = await User.find(
    {"department": "engineering"},
    batch_size=100  # Process in batches of 100
)

# Process documents one at a time (memory efficient)
async for user in cursor:
    process_user(user)

# Or convert a limited number to a list
first_20_users = await cursor.to_list(20)

Batch Operations

Perform operations on multiple documents:

# Bulk insert
users = [User(name=f"User_{i}", email=f"user{i}@example.com") for i in range(100)]
ids = await User.insert_many(users)

# Bulk update
modified = await User.update_many(
    {"status": "trial"},
    {"$set": {"status": "active"}, "$unset": {"trial_ends_at": ""}}
)

# Bulk delete
deleted = await User.delete_many({"last_login": {"$lt": one_year_ago}})

Atomic Updates

Use atomic operations for concurrency control:

# Implement optimistic concurrency control
updated = await User.find_one_and_update(
    {"_id": user_id, "version": 5},  # Only update if version matches
    {
        "$set": {"profile": new_profile},
        "$inc": {"version": 1}  # Increment version number
    },
    return_updated=True  # Get the updated document
)

if not updated:
    print("Update failed - document was modified by another process")

Upsert Operations

Use upsert to insert or update as needed:

# Save with upsert
user = User(name="Alice", email="[e-mail]")
await user.save(upsert=True)  # Insert if not exists, update if exists

# Update with upsert
await User.update_one(
    {"email": "[e-mail]"},
    {"$set": {"name": "Bob Smith", "status": "active"}},
    upsert=True  # Create if not exists
)

Performance Considerations

mongospec is designed for high performance:

• Batch Operations: Use insert_many, update_many, and delete_many for better performance with multiple documents
• Projected Queries: Limit returned fields when possible using projection
• Cursor Batching: Set appropriate batch_size for large result sets
• Indexes: Ensure proper indexes are defined for your query patterns
• Connection Reuse: Initialize mongospec once and reuse for multiple operations

Example of projected query:

# Only retrieve needed fields
users = await User.find(
    {"department": "sales"},
    projection={"name": 1, "email": 1, "_id": 1}  # Only these fields
).to_list(50)

API Reference

Module Functions

• mongospec.init(db, document_types=None) - Initialize connection and bind document types
• mongospec.close() - Close database connection
• mongospec.get_collection(name) - Get a collection by name
• mongospec.is_connected() - Check connection status

MongoDocument

Base class for all document models with the following class variables:

• __collection_name__ - Custom collection name (optional)
• __preserve_types__ - Types to preserve during serialization
• __indexes__ - MongoDB indexes to create
• __collection__ - Runtime collection reference (set by mongospec.init)

Instance methods:

• insert() - Insert document
• save(upsert=False) - Save document changes
• delete() - Delete document
• dump() - Serialize to dict

Class methods:

• Find Operations

  • find_one(filter) - Find single document
  • find_by_id(document_id) - Find by ID
  • find(filter, batch_size=None) - Create cursor for query
  • find_all() - Get all documents
  • exists(filter) - Check if documents exist

• Insert Operations

  • insert_one(document) - Insert single document
  • insert_many(documents, ordered=True) - Insert multiple documents
  • insert_if_not_exists(document, filter=None) - Conditional insert

• Update Operations

  • update_one(filter, update) - Update single document
  • update_many(filter, update) - Update multiple documents
  • update_by_id(document_id, update) - Update by ID
  • find_one_and_update(filter, update, return_updated=True) - Atomic update

• Delete Operations

  • delete_one(filter) - Delete single document
  • delete_many(filter) - Delete multiple documents
  • delete_by_id(document_id) - Delete by ID

• Count Operations

  • count_documents(filter=None) - Count documents
  • estimated_document_count() - Fast approximate count
  • count(filter=None) - Count documents

Development Status

mongospec is currently in beta stage. The core API is stable and all basic CRUD operations are fully supported. We're working on additional features and optimizations for future releases.

Designed for:

• Rapid prototyping
• Performance-critical applications
• Modern asyncio-based projects