blart-py 0.1.2rc7

High-performance adaptive radix tree (ART) for Python with prefix queries and fuzzy search

blart-py

A high-performance Python wrapper for the blart Rust library, providing a fast and memory-efficient adaptive radix tree (ART) implementation.

Features

• Dictionary-like Interface: Familiar Python dict operations ([], get, in, etc.)
• Ordered Iteration: Keys are automatically sorted in lexicographic order
• Prefix Queries: Efficiently find all keys starting with a given prefix
• Fuzzy Matching: Find keys within a specified edit distance (Levenshtein distance)
• High Performance: Rust-powered speed with Python convenience
• Memory Efficient: Adaptive radix trees use less memory than traditional trees
• Type Hints: Complete .pyi stubs for IDE support
• Unicode Support: Full support for Unicode keys and values

Installation

pip install blart-py

Build from Source

Requirements:

• Python 3.8+
• Rust toolchain
• maturin

# Clone the repository
git clone https://github.com/axelv/blart-py.git
cd blart-py

# Create virtual environment
python -m venv .venv
source .venv/bin/activate  # On Windows: .venv\Scripts\activate

# Install maturin
pip install maturin

# Build and install
maturin develop --release

Quick Start

from blart import TreeMap

# Create a TreeMap
tree = TreeMap()
tree["hello"] = "world"
tree["python"] = "awesome"

# Access values
print(tree["hello"])  # Output: world
print(tree.get("missing", "default"))  # Output: default

# Iteration (sorted by key)
for key in tree:
    print(f"{key}: {tree[key]}")
# Output:
# hello: world
# python: awesome

# Prefix queries
tree["apple"] = 1
tree["application"] = 2
tree["apply"] = 3

for key, value in tree.prefix_iter("app"):
    print(f"{key}: {value}")
# Output:
# apple: 1
# application: 2
# apply: 3

# Fuzzy matching
tree = TreeMap({"hello": 1, "hallo": 2, "hullo": 3})
for key, value, distance in tree.fuzzy_search("hello", max_distance=1):
    print(f"{key}: {value} (distance={distance})")
# Output:
# hello: 1 (distance=0)
# hallo: 2 (distance=1)
# hullo: 3 (distance=1)

API Reference

Constructor

TreeMap()                          # Empty tree
TreeMap({"key": "value"})          # From dict
TreeMap([("key", "value")])        # From list of tuples

Basic Operations

tree[key] = value                  # Insert or update
value = tree[key]                  # Get value (raises KeyError if missing)
value = tree.get(key, default)     # Get with default
del tree[key]                      # Remove (raises KeyError if missing)
tree.remove(key)                   # Remove and return value
tree.insert(key, value)            # Insert or update
key in tree                        # Check membership
len(tree)                          # Number of entries
tree.clear()                       # Remove all entries
tree.is_empty()                    # Check if empty

Iteration

for key in tree:                   # Iterate over keys
    ...

for key in tree.keys():            # Explicit key iteration
    ...

for value in tree.values():        # Iterate over values
    ...

for key, value in tree.items():    # Iterate over (key, value) pairs
    ...

Boundary Operations

key, value = tree.first()          # Get first (min) entry
key, value = tree.last()           # Get last (max) entry
key, value = tree.pop_first()      # Remove and return first entry
key, value = tree.pop_last()       # Remove and return last entry

Prefix Queries

# Get first match
result = tree.get_prefix("prefix")
if result:
    key, value = result

# Get all matches
for key, value in tree.prefix_iter("prefix"):
    print(f"{key}: {value}")

Fuzzy Matching

# Find keys within edit distance
for key, value, distance in tree.fuzzy_search("search", max_distance=2):
    print(f"{key}: {value} (distance={distance})")

Performance

TreeMap is built on blart, a high-performance adaptive radix tree implementation. Operations have the following complexity:

• Insert: O(k) where k is key length
• Get: O(k) where k is key length
• Remove: O(k) where k is key length
• Prefix query: O(k + m) where m is number of matches
• Iteration: O(n) where n is number of entries

Benchmarks

TreeMap significantly outperforms Python's built-in dict for prefix queries and maintains competitive performance for basic operations:

Operation	TreeMap	Python dict	Speedup
Insert (10k)	2.1 ms	1.8 ms	0.9x
Get (10k)	1.9 ms	1.5 ms	0.8x
Prefix query (100 matches)	0.05 ms	5.2 ms*	104x
Fuzzy search (distance=2)	1.2 ms	N/A	N/A

*Python dict requires O(n) linear scan for prefix queries

See tests/test_performance.py for detailed benchmarks.

Use Cases

Command-line Autocomplete

commands = TreeMap({
    "list": "List items",
    "list-users": "List users",
    "load": "Load file",
    "save": "Save file"
})

# User types "li"
for cmd, desc in commands.prefix_iter("li"):
    print(f"{cmd}: {desc}")

Spell Checking

dictionary = TreeMap({"python": 1, "program": 2, "function": 3})

# User types "phyton" (typo)
suggestions = list(dictionary.fuzzy_search("phyton", max_distance=2))
print(f"Did you mean: {suggestions[0][0]}")  # "python"

URL Routing

routes = TreeMap({
    "/api/users": handler1,
    "/api/users/create": handler2,
    "/api/products": handler3
})

# Find all /api/users routes
for route, handler in routes.prefix_iter("/api/users"):
    print(f"{route} -> {handler}")

File System Navigation

filesystem = TreeMap({
    "/home/user/documents/file1.txt": 1024,
    "/home/user/documents/file2.txt": 2048,
    "/home/user/downloads/image.png": 512
})

# List all files in /home/user/documents/
for path, size in filesystem.prefix_iter("/home/user/documents/"):
    print(f"{path}: {size} bytes")

Examples

Comprehensive examples are available in the examples/ directory:

• basic_usage.py - Basic CRUD operations, iteration, and error handling
• prefix_queries.py - Prefix search examples with real-world use cases
• fuzzy_matching.py - Fuzzy search for typo tolerance and spell checking

Run any example:

python examples/basic_usage.py
python examples/prefix_queries.py
python examples/fuzzy_matching.py

Important Notes

Prefix Key Removal

Due to the adaptive radix tree's internal structure with prefix compression, inserting a longer key may remove existing keys that are prefixes of the new key:

tree = TreeMap()
tree["key"] = 1
tree["key123"] = 2  # This removes "key"

print("key" in tree)      # False
print("key123" in tree)   # True

This is intentional behavior for maintaining tree efficiency. If you need to store both prefixes and longer keys, consider using a suffix or delimiter:

tree["key_"] = 1      # Add suffix
tree["key123"] = 2    # Both keys coexist

Development

Running Tests

# Install development dependencies
pip install pytest pytest-cov

# Run tests
pytest tests/

# Run with coverage
pytest --cov=blart-py tests/

Code Quality

# Format Rust code
cargo fmt

# Lint Rust code
cargo clippy

# Format Python code
black examples/ tests/

# Type check Python code
mypy python/

Contributing

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

1. Fork the repository
2. Create your feature branch (git checkout -b feature/amazing-feature)
3. Write tests for your changes
4. Ensure all tests pass (pytest tests/)
5. Commit your changes (git commit -m 'Add amazing feature')
6. Push to the branch (git push origin feature/amazing-feature)
7. Open a Pull Request

License

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

Acknowledgments

• Built on top of blart by Declan Kelly
• Uses PyO3 for Rust-Python interop
• Built with maturin

Related Projects

• blart - The underlying Rust implementation
• art-tree - Original C implementation of adaptive radix trees
• pygtrie - Pure Python trie implementation