dhi 1.5.0

Ultra-fast data validation for Python - 33M+ rows/sec, powered by Zig

dhi - Ultra-Fast Data Validation for Python

The fastest data validation library for Python. Powered by Zig for maximum performance.

🚀 Performance

33.18M rows/sec on name+email+age batches, 16.13M rows/sec on six-field batches, and 461.19M ints/sec on the direct integer-list batch path.

# Validate 10,000 users in about 0.30ms
from dhi import _dhi_native

users = [{"name": "Alice", "email": "[e-mail]", "age": 25}, ...]

field_specs = {
    'name': ('string', 2, 100),
    'email': ('email',),
    'age': ('int_positive',),
}

results, valid_count = _dhi_native.validate_batch_direct(users, field_specs)
# 33M rows/sec on name+email+age batches

✨ Features

• ⚡ Fastest: 33.18M rows/sec on name+email+age, 25.35M rows/sec with URL checks, and 461.19M ints/sec for direct integer-list validation
• 🎯 24+ Validators: Email, URL, UUID, IPv4, dates, numbers, strings
• 🔋 Zero Python Overhead: C extension extracts directly from dicts
• 🌍 General Purpose: Works with any dict structure
• 💪 Production Ready: Thoroughly tested and benchmarked

📦 Installation

pip install dhi

🎯 Quick Start

from dhi import _dhi_native

users = [
    {"name": "Alice", "email": "[e-mail]", "age": 25},
    {"name": "Bob", "email": "[e-mail]", "age": 30},
]

field_specs = {
    'name': ('string', 2, 100),
    'email': ('email',),
    'age': ('int_positive',),
}

results, valid_count = _dhi_native.validate_batch_direct(users, field_specs)
print(f"Valid: {valid_count}/{len(users)}")

�� Available Validators

String: email, url, uuid, ipv4, base64, iso_date, iso_datetime, string

Number: int, int_gt, int_gte, int_lt, int_lte, int_positive, int_non_negative, int_multiple_of

🏆 Benchmarks

Release 1.3.1 native batch results on 10,000-row batches:

name+email+age:             33.18M rows/sec  (~28% faster than the earlier 26M baseline)
name+email+age+url:         25.35M rows/sec  (~27% faster than the earlier 20M baseline)
name+email+age+url+uuid:    20.51M rows/sec  (~28% faster than the earlier 16M baseline)
all 6 (+ ipv4):             16.13M rows/sec  (~47% faster than the earlier 11M baseline)
uuid only:                 112.78M rows/sec
ipv4 only:                  68.98M rows/sec
direct int range list:     461.19M ints/sec

📝 License

MIT License - see LICENSE file

Links

• GitHub: https://github.com/justrach/dhi
• PyPI: https://pypi.org/project/dhi/