gender-detect 0.1.0

​⚡ Ultra-fast name-to-gender prediction engine. Uses mmap and binary search for <1ms lookups. Features a pre-compiled 4.6MB binary database covering 700k+ global entries. Built for scale.

gender-detect

A high-performance, binary-search based gender detection library and CLI tool. This project uses a pre-compiled binary database to predict gender and country of origin based on first names with sub-millisecond latency using memory mapping (mmap).

Features

• Extreme Speed: Uses binary search (O(log n)) on a packed binary database with mmap for zero-copy lookups.
• Zero-Dependency: Built entirely using Python standard libraries.
• CLI Ready: Includes a built-in table-formatted command line interface.
• Privacy Focused: 100% local; no external API calls or data tracking.

Installation

pip install gender-detect

CLI Usage

After installation, you can use the gender-detect command directly from your terminal:

gender-detect John

For automation, you can output the result in raw JSON:

gender-detect John --json

Library Usage

Simple Prediction

Input a name to get a statistical analysis of the likely gender and primary origin.

from gender_detect import GenderDetector

gd = GenderDetector()
result = gd.predict("John")

print(result)

Response Format

The gender_probability represents the likelihood of the gender being correct based on total global samples.

{
  "name": "john",
  "likely_gender": "male",
  "gender_probability": 0.83,
  "top_reported_country": "US",
  "data_breakdown": [
    {
      "country": "US",
      "male_samples": 4,
      "female_samples": 1
    },
    {
      "country": "GB",
      "male_samples": 1,
      "female_samples": 0
    }
  ]
}

How it Works

The library utilizes a custom packed binary format (4sHBB):

• 4 bytes: BLAKE2b hash prefix of the name.
• 2 bytes: ISO-3166-1 numeric country code.
• 1 byte: Male sample count.
• 1 byte: Female sample count.

By sorting these 8-byte entries by their hash, the library performs a binary search directly on the file disk/memory, ensuring a tiny memory footprint regardless of database size.

Contribution

Data contributions are managed through contribute.json in the main repository.

1. Add your name data to the JSON list.
2. Ensure country_code is the numeric ISO-3166-1 value.
3. Submit a Pull Request.

The CI/CD pipeline automatically validates the JSON and recompiles the names.bin database upon merging.

License

MIT - See LICENSE file for details.