dataknobs-kv 0.1.0

Key/value store abstraction with hierarchical keys, pattern matching, and metadata support

DataKnobs KV Package

A powerful key/value store abstraction with hierarchical keys, pattern matching, and rich metadata support, built on top of the dataknobs-data package.

Overview

The dataknobs-kv package provides a simple yet powerful interface for managing key/value pairs across multiple storage backends. It features hierarchical key structures, advanced pattern matching, metadata management, and seamless integration with various storage technologies through the dataknobs-data abstraction layer.

Features

• Hierarchical Keys: Dot-notation key paths for organized data structure
• Pattern Matching: Powerful glob-style patterns for key operations
• Rich Metadata: Associate custom metadata with every key/value pair
• Multiple Backends: Memory, File, PostgreSQL, Elasticsearch, S3, and more
• TTL Support: Automatic expiration for time-sensitive data
• Atomic Operations: Compare-and-swap, increment, append operations
• Namespace Isolation: Multi-tenant support with isolated namespaces
• Type Safety: Strong typing with automatic serialization/deserialization

Installation

# Basic installation
pip install dataknobs-kv

# With caching support
pip install dataknobs-kv[cache]

# With all features
pip install dataknobs-kv[all]

Quick Start

from dataknobs_kv import KVStore

# Create a store with memory backend
store = KVStore(backend="memory")

# Basic operations
await store.set("app.name", "MyApplication")
await store.set("app.version", "1.0.0")
await store.set("app.config.debug", True)

# Get values
name = await store.get("app.name")
debug = await store.get("app.config.debug")

# Pattern matching
app_keys = await store.keys("app.*")
all_config = await store.get_pattern("app.config.*")

# Delete keys
await store.delete("app.config.debug")
await store.delete_pattern("app.temp.*")

Hierarchical Key Structure

# Organize data with dot-notation paths
await store.set("users.123.name", "John Doe")
await store.set("users.123.email", "[e-mail]")
await store.set("users.123.settings.theme", "dark")
await store.set("users.123.settings.notifications", True)

# Get all user data
user_data = await store.get_pattern("users.123.*")

# Get all user settings
user_settings = await store.get_pattern("users.123.settings.*")

# Delete entire user
await store.delete_pattern("users.123.**")

Pattern Matching

# Wildcards
await store.keys("users.*.name")        # All user names
await store.keys("*.config.*")          # All config at any level
await store.keys("logs.2024-*")         # All 2024 logs

# Recursive wildcards
await store.keys("app.**")              # Everything under app
await store.keys("**.error")            # All error keys at any level

# Single character and sets
await store.keys("user.?.name")         # user.1.name, user.a.name, etc.
await store.keys("log.[0-9].txt")       # log.0.txt through log.9.txt

# Alternatives
await store.keys("env.{dev,prod}.config") # dev or prod config

Metadata Management

# Set with metadata
await store.set(
    "document.report.pdf",
    document_bytes,
    metadata={
        "content_type": "application/pdf",
        "author": "John Doe",
        "created": "2024-01-15",
        "tags": ["quarterly", "finance"],
        "ttl": 86400  # Expire in 24 hours
    }
)

# Get metadata
metadata = await store.get_metadata("document.report.pdf")
print(f"Author: {metadata['author']}")
print(f"Tags: {metadata['tags']}")

# Update metadata
await store.set_metadata("document.report.pdf", {
    "reviewed": True,
    "reviewer": "Jane Smith"
})

TTL and Expiration

# Set with TTL (time to live)
await store.set("session.abc123", session_data, ttl=3600)  # 1 hour
await store.set("cache.results", results, ttl=300)         # 5 minutes

# Check if key exists (returns False if expired)
if await store.exists("session.abc123"):
    data = await store.get("session.abc123")

# Manual cleanup of expired keys
expired_count = await store.cleanup_expired()
print(f"Removed {expired_count} expired keys")

# Automatic cleanup (runs in background)
store = KVStore(backend="memory", auto_cleanup=True, cleanup_interval=60)

Atomic Operations

# Compare and swap (CAS)
success = await store.compare_and_swap(
    "version",
    old_value="1.0.0",
    new_value="1.1.0"
)

# Increment/decrement counters
views = await store.increment("stats.page_views")
remaining = await store.increment("inventory.item_123", delta=-1)

# Append to strings
await store.append("logs.access", "\n2024-01-15 10:30 User login")
await store.append("notes", ", Remember to update docs")

Batch Operations

# Set multiple values
await store.set_many({
    "config.host": "localhost",
    "config.port": 8080,
    "config.ssl": True,
    "config.timeout": 30
})

# Get multiple values
values = await store.get_many([
    "config.host",
    "config.port",
    "config.ssl"
])

# Delete multiple keys
deleted = await store.delete_many([
    "temp.file1",
    "temp.file2",
    "cache.old"
])

Namespace Isolation

# Create isolated namespaces
user_store = KVStore(backend="postgres", namespace="user_data")
system_store = KVStore(backend="postgres", namespace="system")
cache_store = KVStore(backend="memory", namespace="cache")

# Same keys, different namespaces
await user_store.set("settings", {"theme": "dark"})
await system_store.set("settings", {"debug": True})

# Values are isolated
user_settings = await user_store.get("settings")   # {"theme": "dark"}
system_settings = await system_store.get("settings") # {"debug": True}

Backend Configuration

Memory Backend

store = KVStore(backend="memory")

File Backend

store = KVStore(backend="file", config={
    "path": "/data/kv_store.json",
    "format": "json",
    "compression": "gzip"
})

PostgreSQL Backend

store = KVStore(backend="postgres", config={
    "host": "localhost",
    "database": "myapp",
    "table": "kv_store",
    "user": "dbuser",
    "password": "dbpass"
})

Redis Backend

store = KVStore(backend="redis", config={
    "host": "localhost",
    "port": 6379,
    "db": 0,
    "password": "optional"
})

S3 Backend

store = KVStore(backend="s3", config={
    "bucket": "my-kv-store",
    "prefix": "data/",
    "region": "us-west-2"
})

Advanced Usage

Custom Serialization

import pickle

# Custom serializer for complex objects
class CustomStore(KVStore):
    def serialize(self, value):
        if isinstance(value, MyComplexClass):
            return pickle.dumps(value)
        return super().serialize(value)
    
    def deserialize(self, data, value_type):
        if value_type == "custom":
            return pickle.loads(data)
        return super().deserialize(data, value_type)

Caching Layer

# Add caching for performance
store = KVStore(
    backend="postgres",
    cache="memory",
    cache_ttl=60,  # Cache for 60 seconds
    cache_size=1000  # Max 1000 items in cache
)

Migration Between Backends

from dataknobs_kv import migrate_store

# Migrate from file to PostgreSQL
source = KVStore(backend="file", config={"path": "data.json"})
target = KVStore(backend="postgres", config=postgres_config)

await migrate_store(source, target, batch_size=100)

Development

# Install development dependencies
pip install -e ".[dev]"

# Run tests
pytest

# Run tests with coverage
pytest --cov=dataknobs_kv

# Type checking
mypy src/dataknobs_kv

# Linting
ruff check src/dataknobs_kv

# Format code
black src/dataknobs_kv

Architecture

The package is built on top of dataknobs-data, providing:

• Keys: Hierarchical path management with validation
• Values: Type-safe serialization/deserialization
• Patterns: Advanced pattern matching engine
• Metadata: Rich metadata with system and user fields
• Namespaces: Isolation for multi-tenant applications
• Backends: Leverages dataknobs-data storage backends

Performance

• Optimized pattern matching with compiled regex
• Efficient batch operations
• Optional caching layer
• Connection pooling for database backends
• Async/await support for concurrent operations

Contributing

Contributions are welcome! Please see our Contributing Guide for details.

License

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