scripts/sample_lines.py

#!/usr/bin/env python3
"""Print and save the first N lines of a file (safe for very large files).
Usage: python scripts/sample_lines.py <input-file> <num_lines>
Writes a sample file to prepared_data/samples/<basename>.sample.txt
"""
import sys
from pathlib import Path

def sample_lines(path: Path, n: int = 20):
    out_dir = Path('prepared_data') / 'samples'
    out_dir.mkdir(parents=True, exist_ok=True)
    out_file = out_dir / (path.name + '.sample.txt')
    printed = []
    with path.open('r', errors='ignore') as fh, out_file.open('w', errors='ignore') as oh:
        for i, line in enumerate(fh):
            if i >= n:
                break
            printed.append(line.rstrip('\n'))
            oh.write(line)
    print(f"Wrote sample {out_file} ({len(printed)} lines)")
    print('\n'.join(printed))

if __name__ == '__main__':
    if len(sys.argv) < 2:
        print('Usage: sample_lines.py <input-file> [num_lines]')
        sys.exit(1)
    p = Path(sys.argv[1])
    n = int(sys.argv[2]) if len(sys.argv) > 2 else 20
    if not p.exists():
        print('File not found:', p)
        sys.exit(2)
    sample_lines(p, n)