bat-bit 1.2.1

The Dark Knight of Data Structures - High-Performance C++ Sparse Bitset for Python.

🦇 Bat-Bit: The Dark Knight of Data Structures

"It's not who I am underneath, but what I do that defines me." High-Performance C++ Sparse Bitset Engine for Python.

---

🌃 The Problem

Python's built-in set() and list are powerful, but they are memory-hungry. Trying to store 10 Million integers in a Python set can consume over 600 MB of RAM. If you work with Big Data, High-Frequency Trading, or massive IoT logs, you will hit MemoryError fast.

🦇 The Solution

Bat-Bit is a blazing fast C++ extension wrapped for Python. It uses Sparse Paging Bitset architecture to store data.

• It allocates memory only when needed (Sparse).
• It performs operations at the Bit-level (Processor friendly).
• It releases the GIL (Global Interpreter Lock) for True Multi-Core Parallelism.

---

⚡ Benchmarks (The Shocking Truth)

Tested on 10 Million Records (Integers):

Metric	Standard Python set()	Bat-Bit 🦇	Improvement
Memory Usage	~600 MB	~120 MB	80% RAM Savings 💾
Insertion Time	~12.0s	~1.2s	10x Faster 🚀
Ops / Second	~800k	7.7 Million	Hypersonic Speed 🔥

---

🛠 Installation

Get it directly from PyPI:

pip install bat-bit

Requires: Python 3.7+

💻 Usage

1. Basic Usage (Like a Set)

import bat_bit

# Initialize the BatCave
bruce = bat_bit.BatCave()

# Deploy single gadget (Insert number)
bruce.deploy(500)
bruce.deploy(1000000000) # Handles massive gaps effortlessly!

# Signal check (Search / Lookup) - O(1) Time Complexity
if bruce.signal(500):
    print("Target Found!") # Prints: Target Found!

if not bruce.signal(999):
    print("Target Not Found!")

2. 🆕 New in v1.2.0: Arrays & Maps

Bat-Bit now replaces Python Lists and Dictionaries with high-performance C++ alternatives.

📏 BatVector (The Fast Array)

A contiguous memory array for high-speed numeric processing.

import bat_bit

# Create Vector
vec = bat_bit.BatVector()

# Batch Push (Hypersonic Speed)
data = list(range(1000000))
vec.push_batch(data)

# Access (O(1))
print(vec[500]) 

🗺️ BatMap (The Efficient Dictionary)

A typed hash map that avoids Python object overhead.

m = bat_bit.BatMap()

# Batch Insert
keys = [1, 2, 3]
values = [10.5, 20.5, 30.5]
m.put_batch(keys, values)

print(m[1]) # Output: 10.5

3. High-Speed Batch Processing

Avoid Python overhead by sending data in bulk.

data = [x for x in range(1000000)] # 1 Million items
bruce.deploy_batch(data) # Super Fast

4. 🚀 Hypersonic Parallel Mode (Multi-Core)

Unleash the full power of your CPU. This method releases the GIL and uses all available CPU cores to insert data in parallel.

import bat_bit
import random

# Generate 10 Million numbers
data_stream = [random.randint(0, 1000000000) for _ in range(10000000)]

bruce = bat_bit.BatCave()

# Fires on ALL Cores (e.g., 8 Threads)
bruce.deploy_batch_parallel(data_stream)

print(f"Memory Used: {bruce.memory_usage() / 1024 / 1024:.2f} MB")

🧠 How it Works

Unlike a Hash Map that stores the number itself (32/64 bits) + pointers + overhead, Bat-Bit treats memory as a Bit-Map.

The number 5 is just the 5th bit in a page set to 1.

Sparse Paging: If you store numbers 1 and 1,000,000,000, Bat-Bit only creates 2 tiny pages of memory. The empty space in between takes 0 RAM.

5. 🏛️ BatStore (OOP / Entity Component System)

Store millions of Objects (like Users, Products) with 80% less RAM using Columnar Storage.

# Define Schema once
store = bat_bit.BatStore()
store.add_str_col("username")
store.add_int_col("age")
store.add_float_col("balance")

# Create Rows (Objects)
uid = store.new_row()
store.set_str("username", uid, "Bruce Wayne")
store.set_int("age", uid, 35)
store.set_float("balance", uid, 999999.99)

# Retrieve Data
print(store.get_str("username", uid)) # Output: Bruce Wayne

Memory Benchmark (1 Million Objects):

• Python List of Dicts: ~280 MB
• BatStore: ~62 MB (Huge Savings!)

🤝 Contributing

We welcome fellow vigilantes!

1. Fork the repo.
2. Create your feature branch (git checkout -b feature/NewGadget).
3. Commit your changes.
4. Push to the branch.
5. Open a Pull Request.

📜 License

Distributed under the MIT License. See LICENSE for more information.

Built with 🖤 using C++ & Pybind11