"""List .vbp files in a given analyzer-output dir that are newer than a cutoff date,
filtering out Crystal Reports SDK samples and common third-party library bundles.

Usage: py find_newer_vbp.py <analyzer-output-dir> [cutoff_date]
       cutoff_date format: YYYY/MM/DD HH:MM:SS  (default 2020/06/09 16:13:00)
"""
import csv, os, sys

OUT_DIR = sys.argv[1]
CUTOFF = sys.argv[2] if len(sys.argv) > 2 else '2020/06/09 16:13:00'

NOISE_SUBSTRINGS = (
    'seagat', 'crw\\samples', 'samples\\code\\vb',
    '\\tools\\', '\\plug-ins\\', '\\plugins\\',
    '\\program files\\microsoft visual studio\\',
    '\\program files (x86)\\microsoft visual studio\\',
)

newer = []
csv_path = os.path.join(OUT_DIR, 'cat_vb6-project.csv')
with open(csv_path, encoding='utf-8') as f:
    r = csv.reader(f); next(r)
    for size, sizemb, mod, path in r:
        if mod <= CUTOFF:
            continue
        pl = path.lower()
        if any(s in pl for s in NOISE_SUBSTRINGS):
            continue
        newer.append((mod, int(size), path))

newer.sort(reverse=True)
print(f'Found {len(newer)} .vbp files newer than {CUTOFF} (filtered)\n')
for mod, size, path in newer[:50]:
    print(f'{mod}  {size:>6}b  {path}')
if len(newer) > 50:
    print(f'\n... and {len(newer)-50} more')