v1.0.2

Release v1.0.2

✨ New Features

• Early detection for header anomalies: Pre-validation mechanism to detect title rows before DuckDB sniffing

  • Analyzes first 5 lines to detect separator pattern anomalies
  • Automatically suggests delimiter + skip=1 configuration

• Automatic fallback for non-standard CSV: Files with title rows or non-standard structures

  • No new CLI options (fully automatic)
  • Tries common delimiter/skip combinations
  • Produces clean output + reject_errors.csv

🐛 Bug Fixes

• Improved handling of CSV files with title rows
• Better DuckDB dialect detection fallback

📚 Documentation

• Added technical documentation for early detection mechanism

Installation

pip install --upgrade csvnorm
csvnorm --version

Full diff: v1.0.1...v1.0.2